Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Pharmacist response to alerts generated from medicaid pharmacy
Document related concepts
Transcript
volume eleven • number seven september 2005 Peer-Reviewed Journal of the Academy of Managed Care Pharmacy JMCP JOURNAL OF MANAGED CARE PHARMACY® Page 550 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs 559 Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes 565 Economic Burden of Anemia in an Insured Population 575 Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative AMCP HEADQUARTERS 100 North Pitt St., Suite 400 Alexandria, VA 22314 Tel: (703) 683-8416 • Fax: (703) 683-8417 Volume 11, No. 7 BOARD OF DIRECTORS C O N T E N T S ■ ORIGINAL RESEARCH 550 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs Karen Stockl, PharmD; Lori Cyprien, MS; and Eunice Y. Chang, PhD 559 Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes Sharashchandra Shetty, PhD; Kristina Secnik, RPh, MPH, PhD; and Alan K. Oglesby, MPH 565 Economic Burden of Anemia in an Insured Population ADVERTISING Allen R. Nissenson, MD, FACP; Sally Wade, MPH; L. Tim Goodnough, MD; Kevin Knight, MD, MPH; and Robert W. Dubois, MD, PhD Advertising for Journal of Managed Care Pharmacy is accepted in accordance with the advertising policy of the Academy of Managed Care Pharmacy. ■ CONTEMPORARY SUBJECT 575 Pharmacist Response to Alerts Generated From Medicaid For advertising information, contact: Professional Media Group, Inc., P.O. Box 189, 40 N. Woodbury Rd., Pitman, NJ 08071 Tel: (800) 486-5454 or (856) 589-5454 Fax: (856) 582-7611 E-mail: [email protected] Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative Troy K. Trygstad, PharmD, MBA; Dale Christensen, RPh, PhD; Jennifer Garmise, PharmD; Robert Sullivan, MD, MPH; and Steven E. Wegner, MD, JD ■ DEPARTMENTS 530 Cover Impressions Nashville Rocks! (2001) Dave Newman 590 Sheila Macho Cover Editor 584 Editorials • Fish Oil for Heart Disease— Happy to Be in Second Place Editorial Subjects—In This Issue and in Previous Issues • COX-2 Inhibitors: Little or No GI Protection, Increased Risk of Cardiovascular Events, High Cost, and Other Class-less Effects Frederic R. Curtiss, PhD, RPh, CEBS Editor-in-Chief Brian K. Crownover, MD, FAAFP, Lt. Col., MC, USAF Associate Editor • Medication Therapy Management Services for Long-term Care Patients: No Road Maps for Those Trying to Find Their Way Joshua J. Spooner, PharmD, MS 594 Letter 596 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference 620 Managed Care Pharmacy Residencies, Fellowships, and Other Programs Questions related to editorial content should be directed to JMCP Managing Editor Tamara C. Faggen: [email protected]; (703) 323-0170. Manuscripts should be submitted electronically at jmcp.msubmit.net. For questions about submission, contact Peer Review Administrator Jennifer A. Booker: [email protected]; (703) 317-0725. SUBSCRIPTIONS Annual subscription rates: USA, individuals, institutions–$60; other countries–$80. Single copies cost $10. Missing issues are replaced free of charge up to 6 months after date of issue. Send requests to AMCP headquarters. For article reprints, contact Diana Sholl, Reprint Management Services, (717) 560-2001, ext. 162; [email protected]. Microfilm and microfiche editions of Journal of Managed Care Pharmacy are available from University Microfilms, 300 N. Zeeb Rd., Ann Arbor, MI 48106. Reprint Guidelines are available at www.amcp.org. All articles published represent the opinions of the authors and do not reflect the official policy or views of the Academy of Managed Care Pharmacy or the authors’ institutions unless so specified. Joanne LaFleur, PharmD Associate Editor JMCP EDITORIAL REPRINTS • Evaluating the Relationship Between Diabetes Treatment and Health Care Costs—Measuring the Atom With a Yardstick? 524 Journal of Managed Care Pharmacy President: Dianne A. Kane Parker, PharmD, Amgen Inc., Fallbrook, CA President-Elect: Steven W. Gray, PharmD, JD, Kaiser Permanente, Downey, CA Past President: James R. (Rusty) Hailey, PharmD, DPh, MBA, Coventry Health Care, Inc., Franklin, TN Treasurer: Cathryn A. Carroll, PhD, Children’s Mercy Hospital & Clinics, Smithville, MO Secretary: Judith A. Cahill, CEBS, Academy of Managed Care Pharmacy, Alexandria, VA Director: Elaine Manieri, BSPharm, Caremark Rx, Inc., Hunt Valley, MD Director: Janeen McBride, RPh, MedImpact Healthcare Systems, Inc., San Diego, CA Director: Mark J. Rubino, RPh, MHA, Aetna, Inc., Hartford, CT Director: Brian Sweet, RPh, MBA, WellPoint Pharmacy Management, Grand Island, NY (JMCP liaison) Director: Richard A. Zabinski, PharmD, UnitedHealthcare Corporation, Edina, MN September 2005 Vol. 11, No. 7 www.amcp.org Copyright© 2005 Academy of Managed Care Pharmacy, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, without written permission from the Academy of Managed Care Pharmacy. J M C P E D I TO R I A L P O L I C Y ■■ Editorial Content and Peer Review All articles and editorials in JMCP undergo blinded peer review. Letters may be peer reviewed to ensure accuracy. The fundamental departments for manuscript submission are: • Original Research • Subject Reviews • Formulary Management • Contemporary Subjects • Editorials • Letters For manuscript preparation requirements, see “JMCP Author Guidelines” in this Journal or at www.amcp.org. ■■ Original Research These are well-referenced articles based on original research that has not been published elsewhere and reflects use of the scientific method. The research is guided by explicit hypotheses that are stated clearly by the authors. ■■ Subject Reviews These are well-referenced, comprehensive reviews of subjects relevant to managed care pharmacy. ■■ Formulary Management These are well-referenced, comprehensive reviews of subjects relevant to formulary management methods or procedures in the conduct of pharmacy and therapeutics (P&T) committees and may include description and interpretation of clinical evidence. ■■ Contemporary Subjects These are well-referenced submissions that describe pilot projects or other subjects that are not intended to be comprehensive reviews of the subject. ■■ Editorials These submissions should be relevant to managed care pharmacy and address a topic of contemporary interest. ■■ Letters If the letter addresses a previously published article, an author response may be appropriate. (See “Letter to the Editor” instructions at www.amcp.org.) EDITORIAL MISSION AND POLICIES JMCP publishes peer-reviewed original research manuscripts, subject reviews, and other content intended to advance the use of the scientific method, including the interpretation of research findings in managed care pharmacy. JMCP is dedicated to improving the quality of care delivered to patients served by managed care pharmacy by providing its readers with the results of scientific investigation and evaluation of clinical, health, service, and economic outcomes of pharmacy services and pharmaceutical interventions, including formulary management. JMCP strives to engage and serve professionals in pharmacy, medicine, nursing, and related fields to optimize the value of pharmaceutical products and pharmacy services delivered to patients. JMCP employs extensive bias-management procedures that include (a) full disclosure of all sources of potential bias and conflicts of interest, nonfinancial as well as financial; (b) full disclosure of potential conflicts of interest by reviewers as well as authors; and (c) accurate attribution of each author’s contribution to the article. Aggressive bias-management methods are necessary to ensure the integrity and reliability of published work. Editorial content is determined by the Editor-in-Chief with suggestions from the Editorial Advisory Board. The views and opinions expressed in JMCP do not necessarily reflect or represent official policy of the Academy of Managed Care Pharmacy or the authors’ institutions unless specifically stated. ■■ Advertising/Disclosure Policy A copy of the full advertising policy for JMCP is available from AMCP headquarters and the Advertising Representative. All aspects of the advertising sales and solicitation process are completely independent of the editorial process. Advertising is positioned either at the front or back of the Journal; it is not accepted for placement opposite or near subject-related editorial copy. Employees of advertisers may submit articles for publication in the editorial sections of JMCP, subject to the usual peer-review process. Financial disclosure, conflict-of-interest statements, and author attestations are required when manuscripts are submitted, and these disclosures generally accompany the article in abstracted form if the article is published. See “Advertising Opportunities” at www.amcp.org. Contact the Advertising Representative to receive a Media Kit. www.amcp.org Vol. 11, No. 7 September 2005 Editorial Office Academy of Managed Care Pharmacy 100 North Pitt St., Suite 400 Alexandria, VA 22314 Tel: (703) 683-8416 Fax: (703) 683-8417 E-mail: [email protected] [email protected] Advertising Sales Office Professional Media Group, Inc. 40 N. Woodbury Rd., Pitman, NJ 08071 Tel: (800) 486-5454 or (856) 589-5454 Fax: (856) 582-7611 E-mail: [email protected] JMCP Journal of Managed Care Pharmacy 527 EDITORIAL MISSION JMCP publishes peer-reviewed original research manuscripts, subject reviews, and other content intended to advance the use of the scientific method, including the interpretation of research findings in managed care pharmacy. JMCP is dedicated to improving the quality of care delivered to patients served by managed care pharmacy by providing its readers with the results of scientific investigation and evaluation of clinical, health, service, and economic outcomes of pharmacy services and pharmaceutical interventions, including formulary management. JMCP strives to engage and serve professionals in pharmacy, medicine, nursing, and related fields to optimize the value of pharmaceutical products and pharmacy services delivered to patients. JMCP employs extensive bias-management procedures intended to ensure the integrity and reliability of published work. EDITORIAL STAFF Editor-in-Chief Frederic R. Curtiss, PhD, RPh, CEBS (361) 749-0482, [email protected] Managing Editor, Tamara C. Faggen, (703) 323-0170, [email protected] Associate Editor, Brian K. Crownover, MD, FAAFP, Lt. Col., MC, USAF, (850) 883-8288, [email protected] Associate Editor, Lida R. Etemad, PharmD, MS, (952) 833-7036, [email protected] Associate Editor, Joanne LaFleur, PharmD, (801) 585-3794, [email protected] Peer Review Administrator, Jennifer A. Booker, (703) 317-0725, [email protected] Graphic Designer, Laura J. Mahoney, (703) 917-0737, ext.101, [email protected] Cover Editor, Sheila Macho, (952) 431-5993, [email protected] Publisher Judith A. Cahill, CEBS, Executive Director, Academy of Managed Care Pharmacy EDITORIAL ADVISORY BOARD The JMCP Editorial Advisory Board is chaired by Marvin D. Shepherd, PhD, Director of the Center for Pharmacoeconomic Studies of the College of Pharmacy at the University of Texas at Austin. Dr. Shepherd and the other advisers review manuscripts and assist in the determination of the value and accuracy of information provided to readers of JMCP. Emily Ann Baker, PharmD, Northstar Health Care Consulting, LLC, Cumming, Georgia John P. Barbuto, MD, HealthSouth Rehabilitation Hospital, Sandy, Utah Eliot Brinton, MD, School of Medicine, University of Utah, Salt Lake City Diana I. Brixner, RPh, PhD, Department of Pharmacotherapy, University of Utah, Salt Lake City Scott A. Bull, PharmD, ALZA Corporation, Mt. View, California Jeanne Carlson, CPA, Blue Care Network, BlueCross BlueShield of Michigan, Southfield Tara R. Cockerham, PharmD, Clinical Pharmacist, Atlanta, Georgia Eric J. Culley, PharmD, Highmark BlueShield, Pittsburgh, Pennsylvania Lisa A. Edwards, PharmD, Clinical Pharmacist, North Kingstown, Rhode Island Leslie Fish, PharmD, Fallon Community Health Plan, Worcester, Massachusetts Feride Frech, MPH, Novartis Pharmaceuticals Corp., East Hanover, New Jersey Zafar Hakim, PhD, Hoffman-La Roche, Nutley, New Jersey Joel Hay, PhD, School of Pharmacy, University of Southern California, Los Angeles Katherine Knapp, PhD, College of Pharmacy, Touro University, Vallejo, California Christina Meyer, MHS, Caremark, Hunt Valley, Maryland Robert P. Navarro, PharmD, Campbell Alliance, Raleigh, North Carolina Eduardo Ortiz, MD, MPH, VA Medical Center, Washington, DC Steven Pepin, PharmD, BCPS, MGI Pharma, Inc., Bloomington, Minnesota Steven R. Peskin, MD, MBA, Pharmaceutical Research Plus, Severna Park, Maryland Cathlene Richmond, PharmD, Drug Information Services, Kaiser Permanente, Oakland Regional Offices, California Madeline Ritchie, PharmD candidate, College of Pharmacy, Idaho State University, Boise Fred L. Sego, Jr., JD, RPh, Poulsbo, Washington Fadia T. Shaya, PhD, MPH, School of Pharmacy, University of Maryland, Baltimore Joshua Spooner, PharmD, MS, Advanced Concepts Institute, Philadelphia, Pennsylvania Marilyn Stebbins, PharmD, CHW Medical Foundation, Rancho Cordova, California; University of California, San Francisco Sean D. Sullivan, PhD, Pharmaceutical Outcomes Research and Policy Program, University of Washington, Seattle Kent H. Summers, PhD, School of Pharmacy, Purdue University, Lafayette, Indiana Sheryl L. Szeinbach, PhD, College of Pharmacy, Ohio State University, Columbus Robert J. Valuck, RPh, PhD, School of Pharmacy, University of Colorado Health Sciences Center, Denver Bill Yates, RPh, PhD, CaremarkPCS, Columbia, South Carolina Journal of Managed Care Pharmacy (ISSN 1083–4087) is published 9 times per year and is the official publication of the Academy of Managed Care Pharmacy (AMCP), 100 North Pitt St., Suite 400, Alexandria, VA 22314; (703) 683-8416; (800) TAP-AMCP; (703) 683-8417 (fax). The paper used by the Journal of Managed Care Pharmacy meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper) effective with Volume 7, Issue 5, 2001; prior to that issue, all paper was acid-free. Annual membership dues for AMCP include $60 allocated for the Journal of Managed Care Pharmacy. POSTMASTER: Send address changes to JMCP, 100 North Pitt St., Suite 400, Alexandria, VA 22314. www.amcp.org 528 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org C O V E R I M P R E S S I O N S About our cover artist Nashville Rocks! (2001) ■ Dave Newman D with an art agent who rented art for television ave Newman creates a fascinating mixedshows. Newman’s artwork has been used in the media work of art like a great chef concocts background sets for shows such as “Seinfeld,” a new dish. First, he chooses his basic “Caroline in the City,” and “Felicity.” ingredients: a piece of wood for the canvas and His work is in many private and corporate multiple hues of acrylic paint. Then he spices up collections such as Ajo Al’s Restaurants in the design with 3-dimensional objects culled from Scottsdale, Arizona; Pfizer, Inc. in San Diego, his studio filled with nostalgic memorabilia. California; and Fender Guitar in Corona, “I use photographs I’ve taken, metal parts, silkCalifornia. Newman recently painted two 10-feetscreened images, and ‘found objects’ of all kinds to high Fender Stratocaster guitar sculptures that will weave stories into my work,” Newman says. The be donated to a Big Brothers Big Sisters of Central finished collage is a delicious visual treat with the Arizona charity auction in February 2006. The perfect blend of color and texture. commission was sponsored by the Best Western Newman painted an acoustic guitar, musical “My paintings are created with an hotel chain and the Phoenix Airport Museum. His notes, and running cowboys to convey the “Music intentionally humorous, whimsical feel.” guitars are part of “GuitarMania,” a Greater Phoenix City” theme of his Nashville Rocks! collage. public art project presented by Big Brothers Big A 45-rpm record, dominoes, Scrabble tiles, scraps of sheet music, Tennessee state maps, photographs, and pieces of old Sisters of Central Arizona, the Rock and Roll Hall of Fame and license plates complete the composition. He created the artwork to be Museum, and Fender Guitar. The completed guitars will be displayed used as the image for the 2001 Country Music Marathon’s official at the Phoenix Airport and elsewhere throughout the Valley for a fiveposter. Held in Nashville in April, the annual race began in 2000 and month period beginning in September 2005. Newman’s Web site (www.davenewmanstudio.com) features is organized and managed by Elite Racing, Inc., which also commissioned Newman to do the inaugural race poster. Country music bands original mixed-media collages and acrylic paintings. He says that he play along the marathon route, and the participants are rewarded with enjoys using acrylic paint because it allows him to be impulsive and go with his instincts as he creates each piece. The Newman Gallery a star-studded postrace concert. The original Nashville Rocks! was auctioned off at the end of the (www.newmangallery.net) showcases Newman’s art and represents marathon, with proceeds going to the Leukemia & Lymphoma Society. 25 artists from around the country. It is located on historic Whiskey Newman has a history of contributing to communities through his art- Row in the 100-year-old St. Michael Hotel in downtown Prescott. work—he has also donated his art to the Heart Foundation, an Arizona Other galleries that feature his work are Atmosphere Art & Hair Salon in Scottsdale, Arizona; Culture Shop Gallery in Venice, California; and Montessori school, and museum auctions in Arizona and California. Born in 1956 in Long Beach, California, Newman was reared in Terzian Galleries in Park City, Utah. Newman says that one of his favorite themes is retro American Anaheim, a mile away from Disneyland. Growing up so close to Disneyland may have had an affect on his art, as his artist statement culture, especially Route 66. “Sometimes my art begins on a highway reveals: “My paintings are created with an intentionally humorous, where there is a spectacular view or where I’ve found something interwhimsical feel.” Newman’s vibrant work also reflects his lighthearted esting on the side of the road,” he says. One of his upcoming projects view of life. His collages are filled with so much detail that they almost involves the painting and collage of a 1977 Chevy pickup truck. He always garner a second look—which is just what Newman desires. already has the personalized Arizona license plates, “ROAD ART.” That “I want people to look again and again because there is always some- truck will undoubtedly blaze a colorful trail across the desert. thing else to see,” he says. Sheila Macho Newman is a self-taught artist, and he maintains a studio, gallery, Cover Editor and residence in Prescott, Arizona. His interest in art began in his early 30s, when he operated a screen-printing business and started reading about pop artists of the 1960s such as Andy Warhol and Jasper Johns. COVER CREDIT Warhol frequently utilized screen printing in his work, and Johns’s Dave Newman, Nashville Rocks!, mixed media on wood. Prescott, Arizona. Copyright© 2001. paintings often included images and objects from popular culture. By the age of 32, Newman had launched his own art career. “My acrylic paintings on canvas and wood are spontaneous commentaries SOURCES Interview with the artist. on pop culture: coffee cups, hearts, television, and the spaces www.davenewmanstudio.com. we inhabit,” he observed. When he lived in California, he worked www.newmangallery.net. 530 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org JMCP Author Guidelines JMCP accepts for consideration manuscripts prepared according to the Uniform Requirements for the Submission of Manuscripts to Biomedical Journals.1 ■■ Manuscript Preparation Manuscripts should include, in this order: title page, abstract, text, references, tables, and figures (see Submission Checklist for details). JMCP abstracts should be written narratives that contain the information described for each type of article shown below, where applicable. For descriptions of editorial content, see “JMCP Editorial Policy” in this Journal or at www.amcp.org. Original Research An abstract is required in the format of: • Keywords • Objective • Results • Methods • Conclusion Subject Reviews An abstract is required, generally in the format of: • Objective • Conclusion • Summary • Keywords Formulary Management An abstract is required in the format of Original Research or Subject Reviews, depending upon the subject matter. Contemporary Subjects An abstract is required in the format of Original Research or Subject Reviews, depending upon the subject matter. Editorials These submissions require no abstract but should include references. Letters These submissions require no abstract or title page. Please note: • A subsection in the Discussion labeled “Limitations” is generally appropriate for articles published in Original Research and sometimes appropriate for articles published in Subject Review, Formulary Management, or Contemporary Subject. • Most articles published in JMCP, particularly Subject Reviews, should incorporate or at least acknowledge the relevant work of others published previously in JMCP (see “Article Index by Subject Category” at www.amcp.org). • For articles in Original Research, a figure is recommended for making the effects of the inclusion and exclusion criteria clear to readers (see JMCP examples in 2003;9(4):320 [Figure 1] or 2003;9(3):258 [Table 1]). • Product trade names may be used only once, for the purpose of providing clarity for readers, generally at the first citation of the generic name but not in the Abstract. ■■ Reference Style References should be prepared following modified AMA style. All reference number in manuscript should be superscript (e.g., 1 ). See examples of common types of references below: The Journal of Managed Care Pharmacy is indexed by Index Medicus/MEDLINE and International Pharmaceutical Abstracts (IPA). 1. Standard journal article (List all authors when 6 or less; if more than 6, list only the first 3 and add et al.) Lennard EL, Feinberg PE. Overview of the New York State program for prescription drug benefits. Am J Hosp Pharm. 1994;512:944-48. 2. No author given Anonymous. Top 25 U.S. hospitals, ranked by admissions, 1992. Manag Healthcare. 1994;4(9):64. 3. Journal paginated by issue Corrigan PW, Luchins DJ, Malan RD, Harris J. User-friendly CQI for the mental health team. Med Interface. December 1994;7:89-92, 95. 4. Book or monograph by authors Tootelian DH, Gaedeke RM. Essentials of Pharmacy Management. St. Louis, MO: C.V. Mosby; 1993. 5. Book or monograph with editor, compiler, or chairman as author Chernow B, ed. Critical Care Pharmacotherapy. Baltimore, MD: Williams & Wilkins; 1995. 6. Chapter in a book Kreter B, Michael KA, DiPiro JT. Antimicrobial prophylaxis in surgery. In: DiPiro JT, Talbert RL, Hayes PE, Yee GC, Matzke GR, Posey LM, eds. Pharmacotherapy: A Pathophysiologic Approach. Norwalk, CT: Appleton & Lange; 1992:1811-12. 7. Government agency publication Akutsu T. Total Heart Replacement Device. Bethesda, MD: National Institutes of Health, National Heart and Lung Institute; April 1974. Report NIHNHLI-69-2185-84. 8. Paper (or Poster) presented at a meeting Reagan ME. Workers’ compensation, managed care, and reform. Paper (poster) presented at: 1995 AMCRA Midyear Managed Care Summit; March 13, 1995; San Diego, CA. 9. Newspaper Winslow R. Lipitor prescriptions surge in wake of big study. Wall Street Journal. March 18, 2004:D4. 10. Web site National Committee for Quality Assurance. The state of health care quality: 2004. Industry trends and analysis. Available at: http://www.ncqa.org/ communications/SOMC/SOHC2004.pdf. Accessed December 26, 2004. ■■ Manuscript Submission A paper copy of the manuscript, including originals of figures and tables and author attestation forms (see Submission Checklist), should be submitted to the JMCP Peer Review Administrator at the Academy of Managed Care Pharmacy, 100 North Pitt St., Suite 400, Alexandria, VA 22314; Tel: (800) 827-2627 or (703) 683-8416 or Fax: (703) 683-8417. The paper copy is necessary to ensure proper presentation and placement of text, figures, tables, and graphs. Please submit the manuscripts electronically at jmcp.msubmit.net. All text should be in a word processing program (preferably Microsoft Word). Tables should be prepared in a word processing program using www.amcp.org Vol. 11, No. 7 September 2005 the tab function to create columns, not using “tables” or “cells.” Figures should be saved in Photoshop or Illustrator and may be re-created by us. Figures may also be prepared in a word processing program (preferably Microsoft Word). We can accept PowerPoint graphics. Please identify the format (PC or MAC), all programs used, and all file names. P values should be expressed to no more than 3 decimal points in the format 0.xxx. Note: Please do not include author identification in the electronic manuscript document. Cover letter: the corresponding (lead) author should include a cover letter with the manuscript, which • briefly describes the importance and scope of the manuscript, • certifies that the paper has not been accepted for publication or published previously and that it is not under consideration by any other publication, and • identifies the nature and extent of any financial interest or affiliation that any author has with any company, product, or service discussed in the manuscript. All manuscripts are reviewed prior to peer review. Manuscripts may be returned to authors prior to peer review for clarification or other revisions. Peer review generally requires 4 weeks but may extend as long as 8 weeks in unusual cases. Solicited manuscripts are subject to the same peer-review standards and editorial policy as unsolicited manuscripts. ■■ Submission Checklist Before submitting the paper copy of your manuscript to the Journal of Managed Care Pharmacy, please check to see that your package includes the following: ❑ Cover letter ❑ Manuscript: prepared in 12-point type, 1.5 line spacing, including ❑ abstract: no more than 500 words ❑ keywords: follows the abstract ❑ references: cited in numerical order as they appear in the text (use superscript numbers) and prepared following modified AMA style; do not include footnotes in the manuscript ❑ tables and figures (generally no more than a total of 6). Submit referenced tables and figures on separate pages with titles (and captions, as necessary) at the end of the manuscript; match symbols in tables and figures to explanatory notes, if included. May use 10-point font. ❑ Disclosures and conflict-of-interest forms: completed and signed author attestation forms (available at www.amcp.org); clearly indicate source(s) of funding and financial support. Note: Please do not include author identification in the electronic manuscript document. For “Manuscript Submission Checklist” and “Peer Review Checklist,” see www.amcp.org. REFERENCE 1. International Committee of Medical Journal Editors. Uniform requirements for manuscripts submitted to biomedical journals. N Engl J Med. 1991;324:424-48. JMCP Journal of Managed Care Pharmacy 549 ORIGINAL RESEARCH Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs KAREN STOCKL, PharmD; LORI CYPRIEN, MS; and EUNICE Y. CHANG, PhD ABSTRACT OBJECTIVE: While cyclooxygenase-2 (COX-2) inhibitors were introduced to the U.S. market with the promise of less gastrointestinal (GI) toxicity than nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), additional research is needed to examine this outcome in the naturalistic setting. The objective of this study was to examine whether use of COX-2 inhibitors is associated with reduced risk of GI bleed in a managed care population. METHODS: Adult patients in a multistate managed care organization that were initiated on a nonselective NSAID between January 1999 and August 2002 were identified and matched using propensity scoring with patients in the same managed care organization that were initiated on a COX-2 inhibitor. Matching variables included age, gender, geographical state, comorbidity index, corticosteroid use, warfarin use, arthritis indication, and history of recent GI bleed. Patients were followed until they switched or discontinued their NSAID or COX-2 inhibitor, disenrolled from the health plan, developed a GI bleed, or reached the end of the 1-year follow-up period. A GI bleed was defined as an inpatient hospitalization for GI bleed or at least 2 medical claims with a primary diagnosis for GI bleed. The relative risk (RR) of GI bleed was calculated using proportional hazards regression. RESULTS: Overall, 35,007 pairs of COX-2 inhibitor and nonselective NSAID users were evaluated. Mean age was 63 years, and 65% were female. There were 375 cases of GI bleed among 19,201 follow-up years for COX-2 users (19.5 cases per 1,000 person-years) versus 228 cases of GI bleed among 12,680 follow-up years for NSAID users (18.0 cases per 1,000 person-years). The risk of GI bleed was not significantly different for COX-2 users compared with nonselective NSAID users (RR 1.07; 95% confidence interval [CI], 0.90-1.26). Even among high-risk patients, there was no reduction in the risk of a GI bleed among users of COX-2 inhibitors (RR 0.995; 95% CI, 0.84 -1.19). CONCLUSION: Overall, within this managed care population, COX-2 inhibitor users did not have a reduced risk of a GI bleed compared with patients with similar baseline characteristics using nonselective NSAIDs. KEYWORDS: Cyclooxygenase, Nonsteroidal anti-inflammatory drugs, Gastrointestinal bleed, Drug therapy J Manag Care Pharm. 2005;11(7):550-58 Note: An editorial on the subject of this article appears on pages 590-93 of this issue. Authors KAREN STOCKL PharmD, was a research scientist, health economics and outcomes research, Prescription Solutions, Costa Mesa, California, at the time of this study; LORI CYPRIEN, MS, is a statistician, health economics and outcomes research; EUNICE Y. CHANG, PhD, is chief statistician, health informatics and outcomes research, Prescription Solutions, Costa Mesa, California. AUTHOR CORRESPONDENCE: Karen Stockl, PharmD, c/o Eunice Y. Chang, PhD, Chief Statistician, Health Informatics and Outcomes Research, Prescription Solutions, 3515 Harbor Blvd., Mail Stop LC07-264, Costa Mesa, CA, 92626. Tel: (714) 825-3732; Fax: (714) 825-3742; E-mail: [email protected] Copyright© 2005, Academy of Managed Care Pharmacy. All rights reserved. 550 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 R ecent reports of adverse cardiovascular outcomes with two of the cyclooxygenase-2 (COX-2) inhibitors have placed the risk-benefit profile of these COX-2 inhibitors under public scrutiny; both of the COX-2 inhibitors have subsequently been withdrawn from the U.S. market).1-3 As a result, clinicians have been encouraged to weigh the potential benefits and risks of the nonsteroidal anti-inflammatory (NSAID) medications prior to prescribing them.4 Managed care has been particularly interested in examining the risks, benefits, and cost-effectiveness of the COX-2 inhibitors.5-8 The present study was designed to evaluate whether the use of COX-2 inhibitor medications was beneficial in preventing gastrointestinal (GI) bleeds compared with the use of nonselective NSAIDs in the naturalistic managed care setting. The early reported results of controlled clinical studies suggested that the COX-2 inhibitors may have less GI toxicity than nonselective NSAIDs,9-14 but the U.S. Food and Drug Administration (FDA) permitted this claim only for rofecoxib, which was later withdrawn from the U.S. market, on September 30, 2004, due to adverse cardiovascular events.1,15 In actual clinical practice, patients’ medication use is not monitored as closely as in clinical trials. As a result, patients may not take their COX-2 inhibitor medication under the same conditions as those studied (e.g., they may take higher doses than those prescribed or they may use COX-2 inhibitors concomitantly with gastrotoxic substances such as alcohol), which may lead to different outcomes than those observed in clinical trials. Furthermore, the population prescribed COX-2 inhibitors in clinical practice may have more risk factors for a GI bleed than the population selected to test the COX-2 inhibitors in clinical trials. The objective of this analysis was to examine whether a managed care population of patients who used COX-2 inhibitors in a naturalistic setting actually did have a reduced occurrence of GI bleed compared with a population of patients with similar baseline characteristics who received nonselective NSAIDs. ■■ Methods This was a retrospective analysis of electronic pharmacy and medical administrative claims from a large managed care organization and Prescription Solutions, a pharmacy benefits and medical management company. Longitudinal claims data were used from health plans (private as well as Medicare + Choice [now Medicare Advantage]) within California, Oklahoma, Oregon, Texas, and Washington, which consist of approximately 2.7 million www.amcp.org Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs lives. This database has been used in previous research studies.16-20 An estimated two thirds of members within this managed care population were subject to prior authorization for COX-2 drugs that required patients to meet certain clinical characteristics prior to receiving authorization for coverage of a COX-2 inhibitor. Prior authorization approval for a COX-2 inhibitor was more likely to be given to patients with older age, a diagnosis of rheumatoid arthritis or osteoarthritis, history of GI bleed, and/or concomitant use of oral corticosteroids or warfarin. Patient Identification and Matching Adult patients (aged 18 years or older) initiated on a COX-2 inhibitor (generic product identifier [GPI] code 661005xx, which included rofecoxib, celecoxib, and valdecoxib) or a nonselective NSAID (GPI code 661000xx, which consisted of all prescription NSAIDs other than those classified as COX-2 inhibitors) during the 44-month period from January 1, 1999, through August 31, 2002 (identification period), were identified. The index date was defined as the date of each patient’s first prescription fill of a COX-2 inhibitor or nonselective NSAID during the identification period. Patients were excluded from the analysis if they had a pharmacy claim for a COX-2 inhibitor or a nonselective NSAID during the 6-month period prior to their index date (i.e., the preperiod) or if they were not continuously enrolled in the health plan during the preperiod and at least 3 months after the index date. The first (earliest) patient that could have been identified would have had an index date on January 1, 1999, and a preperiod starting July 1, 1998. Hence, some patients’ preperiods may have started prior to the FDA approval of celecoxib. Celecoxib, the first COX-2 inhibitor to be approved, was approved on December 31, 1998.21 From these identified patients, the final study cohort was obtained by matching patients who received a nonselective NSAID on the index date with those who received a COX-2 inhibitor on the index date on a 1:1 basis using the propensity score method.22 A propensity score, which represents the likelihood of receiving a COX-2 inhibitor rather than a nonselective NSAID, was determined for each patient. Patients were matched based on their propensity score. The independent variables that were used to calculate the propensity score included demographics (age at index date, gender, geographical state of the health plan), Charlson Comorbidity Index23 (calculated during the preperiod using a method adapted for electronic claims databases),24 and the following GI bleed risk factors (measured during the preperiod): a prescription fill of a corticosteroid (GPI codes 2210xx, 2220xx, 2200xx), a prescription fill of warfarin (GPI code 83200030), a medical claim representing a recent GI bleed (Table 1), history of a GI bleed-related inpatient hospitalization (i.e., an inpatient hospitalization with at least a 1-day length of stay and a diagnosis code representing a GI bleed, Table 1), and arthritis indication (osteoarthritis [International Classification of Diseases, 9th Revision, www.amcp.org TABLE 1 Diagnosis Codes Representing Gastrointestinal Bleed ICD-9-CM Code Description 530.2 531.xx 532.xx 533.xx 534.xx 578.xx Ulcer of esophagus Gastric ulcer Duodenal ulcer Peptic ulcer site unspecified Gastrojejunal ulcer Gastrointestinal hemorrhage ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification. Clincial Modification (ICD-9-CM) code 715.xx], rheumatoid arthritis [ICD-9-CM code 714.xx], or neither). Patients diagnosed with both osteoarthritis and rheumatoid arthritis were classified into the rheumatoid arthritis group. If more than 1 patient who received a nonselective NSAID was identified as a match, 1 patient was selected at random to be included in the final study cohort. Patients who could not be matched were excluded from the analysis. Patients in the matched COX-2 inhibitor and nonselective NSAID cohorts were further stratified according to their GI risk. Patients were considered high-risk if they met any of the following criteria: (1) age greater than 65 years; (2) GI-bleed-related inpatient hospitalization during the preperiod; (3) pharmacy claim for warfarin (GPI code 83200030) during the preperiod; or (4) pharmacy claim for a corticosteroid (GPI codes 2210xx, 2220xx, 2200xx) during the preperiod. All other patients were classified as low-risk. Outcome Measures The primary outcome of interest was the risk of developing a GI bleed over the follow-up period. Patients were followed until the first occurrence of one of the following events: (1) patient discontinued (as defined below) the index COX-2 inhibitor or nonselective NSAID medication; (2) crossover of medication of interest (patient filled a prescription for a study medication in a class [COX-2 inhibitor or nonselective NSAID] other than their index class of medication); (3) patient disenrolled from the health plan; (4) patient had a GI-bleed-related inpatient hospitalization; (5) patient had 2 medical claims with a primary (first-listed) diagnosis for Gl bleed (Table 1) during the follow-up period (where the event date was defined as the date of the first of the 2 claims); or (6) the end of the 1-year follow-up period. A discontinuation was defined as a gap of at least 60 days between the run-out date of the last index COX-2 inhibitor or nonselective NSAID medication fill (fill date plus the days of supply of that last prescription) and the end of the follow-up period. Patients were considered to have a GI bleed event if they experienced a GI bleed-related inpatient hospitalization (event 4 above) or had 2 medical claims with a primary diagnosis for a GI bleed (event 5 above) within the follow-up time frame. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 551 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs FIGURE 1 Study Population Identification Patients Who Filled at Least 1 COX-2 Inhibitor or Nonselective NSAID During the ID Period (N = 1,038,437) Patients Who Did Not Have a COX-2 Inhibitor or Nonselective NSAID Filled During the 6-Month Preperiod (N = 958,075; 92.3%) Patients Who Were Continuously Enrolled in the Health Plan in the Preperiod and at Least 3 Months After the Index Date (N = 613,526; 59.1%) Patients Who Were Aged at Least 18 years or Older (N = 585,634; 56.4%) COX-2 Users (N = 36,401; 3.5%) Nonselective NSAID Users (N = 549,233; 52.9%) COX-2 Users Matched With Nonselective NSAID Users (N = 35,007; 3.4%) Nonselective NSAID Users Matched With COX-2 Users (N = 35,007; 3.4%) COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal antiinflammatory drug. To account for the different lengths of follow-up, the number of GI bleed cases per 1,000 person years was calculated by dividing the number of patients who experienced the GI bleed event during the follow-up period by the sum of each patient’s observed follow-up time (in years) and then multiplying by 1,000. Statistical Analysis Statistical analyses were performed using SAS version 9.1 (SAS Institute Inc., Cary, NC). Chi-square tests and t tests were used to compare baseline demographics and clinical characteristics between the COX-2 inhibitor and nonselective NSAID cohorts. Propensity score calculations were conducted using logistic regression. Two-way interactions between independent variables were tested and only significant interactions (P < 0.05) were included in the final propensity model. The propensity score for each patient was rounded to 0.0001 for matching. Proportional 552 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 hazards regression was used to determine relative risk of a GI bleed event for the 2 cohorts, to study the relationship between the event and specific risk factors, and to adjust for baseline differences between the cohorts. Risk factors included in the model were age, gender, state of health plan, preperiod Charlson Comorbidity Index, preperiod corticosteroid use, preperiod warfarin use, preperiod diagnosis of GI bleed in any diagnostic field, and preperiod rheumatoid or osteoarthritis indication. Separate models were created for the total study population, the low-risk population, and the high-risk population. Only the main effects were included in the models (interaction terms were not tested for inclusion). To adjust for baseline differences in the use of gastroprotective agents between the cohorts, additional proportional hazards regression models were performed after adding preperiod use of a gastroprotective agent as a variable. To assess the proportional hazards assumption, a plot of the scaled Schoenfeld residuals by transformed time for each risk factor was investigated and a test of zero slope of the plot was conducted.25-27 Nonzero slope would indicate a violation of the proportional hazards assumption. If nonproportional hazards were found for some risk factors, then stratified proportional hazards regressions were conducted to evaluate whether the GI bleed event outcome was changed by this stratification. All statistical tests were 2-sided with an alpha of 0.05. ■■ Results There were 1,038,437 patients who filled at least 1 COX-2 inhibitor or nonselective NSAID during the identification period (Figure 1). Among them, 80,362 (7.7%) were excluded because they had a pharmacy claim for a COX-2 inhibitor or nonselective NSAID during the preperiod, and 344,549 were excluded because they were not continuously enrolled in the preperiod and at least 3 months after the index date. An additional 27,892 patients were excluded because they were younger than 18 years. Overall, 585,634 eligible patients were identified; 36,401 (6.2%) used COX-2 inhibitors and 549,233 (93.8%) used nonselective NSAIDs. A total of 70,014 patients (35,007 pairs of COX-2 inhibitor users and nonselective NSAID users) were matched according to propensity score and included in the final study cohort. Demographics and clinical characteristics were similar for COX-2 inhibitor and nonselective NSAID cohorts (Table 2), with the exception of a lower percentage of COX-2 inhibitor users participating in a Medicare + Choice health plan (53.1% versus 55.7%, P < 0.001); health plan type was not one of the variables included in the propensity score match. When the populations were stratified according to low and high GI bleed risk, there were also statistical differences in the mean age for the COX-2 inhibitor and nonselective NSAID cohorts, which were apparently due to the large sample size since the mean www.amcp.org Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs TABLE 2 Baseline Demographics and Clinical Characteristics of Study Population* Total Population COX-2 Inhibitor (N = 35,007) Age, mean (SD) Nonselective NSAID (N = 35,007) Low-Risk Population COX-2 Inhibitor (N = 14,897) High-Risk Population† Nonselective NSAID (N = 14,650) COX-2 Inhibitor (N = 20,110) Nonselective NSAID (N = 20,357) 63.4 (16.3) 63.3 (16.3) 48.9 (10.5)|| 48.6 (10.6)|| 74.2 (10.3)|| 73.9 (10.5)|| Female gender 22,909 (65.4) 22,906 (65.4) 9,219 (61.9) 9,142 (62.4) 13,690 (68.1) 13,764 (67.6) State of health plan California Oklahoma Oregon Texas Washington 24,090 (68.8) 1,771 (5.1) 1,555 (4.4) 3,843 (11.0) 3,748 (10.7) 23,936 (68.4) 1,756 (5.0) 1,560 (4.5) 3,961 (11.3) 3,794 (10.8) 10,726 (72.0) 951 (6.4) 724 (4.9) 1,373 (9.2) 1,123 (7.5) 10,506 (71.7) 902 (6.2) 729 (5.0) 1,361 (9.3) 1,152 (7.9) 13,364 (66.5) 820 (4.1) 831 (4.1) 2,470 (12.3) 2,625 (13.1) 13,430 (66.0) 854 (4.2) 831 (4.1) 2,600 (12.8) 2,642 (13.0) Medicare + Choice health plan (%)‡ 18,574 (53.1)¶ 19,488 (55.7)¶ 17,587 (87.5)¶ 18,349 (90.1)¶ 0.29 (0.95) 0.30 (0.98) 0.15 (0.69) 0.15 (0.71) 0.39 (1.09) 0.40 (1.12) 1,040 (3.0) 5,788 (16.5) 28,179 (80.5) 1,015 (2.9) 5,802 (16.6) 28,190 (80.5) 276 (1.9) 1,224 (8.2) 13,397 (89.9) 283 (1.9) 1,189 (8.1) 13,178 (90.0) 764 (3.8) 4,564 (22.7) 14,782 (73.5) 732 (3.6) 4,613 (22.7) 15,012 (73.7) Warfarin use 2,264 (6.5) 2,262 (6.5) 0 0 2,264 (11.3) 2,262 (11.1) Corticosteroid use 3,385 (9.7) 3,406 (9.7) 0 0 3,385 (16.8) 3,406 (16.7) Medical claim for GI bleed (any field) 843 (2.4) 808 (2.3) 252 (1.7) 223 (1.5) 591 (2.9) 585 (2.9) Inpatient hospitalization for GI bleed 132 (0.4) 119 (0.3) 0 0 132 (0.7) 119 (0.6) Charlson Comorbidity Index, mean (SD) Arthritis indication Rheumatoid arthritis§ Osteoarthritis None 987 (6.6)¶ 1,139 (7.8)¶ * Values are number (%) unless specified otherwise. † High-risk was defined as a patient with (a) age older than 65 years, or (b) recent history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a GI bleed. ‡ Variable was not used in the propensity score match. § Includes patients with medical claims for both osteoarthritis and rheumatoid arthritis. || P = 0.004 for the low-risk population and P = 0.02 for the high-risk population for the comparison of the COX-2 inhibitor cohort and the nonselective NSAID cohort. ¶ P < 0.001 for all 3 populations (total, low-risk, and high-risk) for the comparison of the COX-2 inhibitor cohort and the nonselective NSAID cohort. COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflammatory drug. (± standard deviation) age was similar for the 2 cohorts (48.9 ± 10.5 years versus 48.6 ± 10.6 years, respectively, for low-risk users, P = 0.004; 74.2 ± 10.3 years versus 73.9 ± 10.5 years, respectively, for high-risk users, P = 0.02). Although statistically significant, the small difference in age between the groups probably is not clinically meaningful. During the preperiod, 17.5% of COX-2 inhibitor users and 7.8% of nonselective NSAID users filled at least 1 prescription for a GI protective agent (i.e., a proton pump inhibitor or misoprostol, GPI codes 49250030xx, 6610990220, 4927xx, 6610990242). Subjects in the nonselective NSAID cohort had a shorter follow-up time (average 4.3 months per patient or 12,680 personyears, Table 3) than subjects in the COX-2 inhibitor cohort (average 6.6 months per patient or 19,201 person-years). A GI bleed event (defined as a GI-bleed-related inpatient hospitalization or at least 2 medical claims with a primary diagnosis for GI bleed) was noted in 19.5 patients per 1,000 person years in the COX-2 inhibitor cohort and 18.0 patients per 1,000 person years in the nonselective NSAID cohort (Table 3). The time to GI bleed event (median and mean) was shorter for the nonselective NSAID cohort than for the COX-2 inhibitor cohort. www.amcp.org The proportions of patients in each cohort whose follow-up period ended due to reasons other than a GI bleed event are described below. Compared with COX-2 inhibitor users, a greater percentage of nonselective NSAID users discontinued their index medication (82.2% versus 63.6%), while a smaller percentage of nonselective NSAID users had a crossover of medication of interest (1.2% versus 4.6%). Nonselective NSAID users were also less likely than COX-2 inhibitor users to disenroll from the health plan (15.8% versus 30.1%) but were less likely to reach the end of the 1-year follow-up period than COX-2 inhibitor users (0.2% versus 0.6%). Patients who had a GI bleed event had a mean age of 75.1 ± 11.3 years and 61% were female. The mean Charlson Comorbidity Index was 0.84 ± 1.57. Among these patients, 15.3% used warfarin and 13.1% used corticosteroids during the preperiod. During the preperiod, 16.1% of these patients had a medical claim for a GI bleed (any field), and 3.6% had an inpatient hospitalization for a GI bleed. At the end of follow-up, use of a GI protective agent was observed in 21.1% of COX-2 inhibitor and 8.8% of nonselective NSAID users who had a GI bleed event compared with 8.1% of COX-2 inhibitor and 4.0% Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 553 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs TABLE 3 GI Bleed Events, Number of Cases per 1,000 Person Years Total Population COX-2 Inhibitor (N = 35,007) Sum of patients’ follow-up years Inpatient hospitalization or primary diagnosis for GI bleed on 2 medical claims Low-Risk Population Nonselective NSAID (N = 35,007) High-Risk Population* COX-2 Inhibitor (N = 14,897) Nonselective NSAID (N = 14,650) COX-2 Inhibitor (N = 20,110) Nonselective NSAID (N = 20,357) 7,677 19,201 12,680 6,605 5,003 12,596 Number of cases 375 228 51 19 324 209 Number of cases per 1,000 person-years 19.5 18.0 7.7 3.8 25.7 27.2 84.0 61.5 102.0 78.0 77.0 61.0 111.5 100.9 117.3 110.6 110.6 100.0 Median time to event (days)† Mean time to event (days)† * High-risk was defined as a patient with (a) age older than 65 years, or (b) recent history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a GI bleed. † Median and mean time to event were measured among those patients who had a GI bleed event during the follow-up period. COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflammatory drug. of nonselective NSAID users who did not have a GI bleed event. Proportional hazards regression was conducted to determine the relative risk of a GI bleed event for the 2 cohorts (accounting for different patient follow-up times and adjusting for baseline characteristics) and to study the relationship between the event and specific risk factors (Table 4). Among the total study population and the high-risk population, patients who received a COX-2 inhibitor did not have a lower relative risk of having a GI bleed event (defined as an inpatient hospitalization for GI bleed or at least 2 medical claims with a primary diagnosis for GI bleed) compared with patients who received a nonselective NSAID. Other factors that were significant predictors of a GI bleed event were older age, male gender, state of health plan (California versus Texas), higher Charlson Comorbidity Index, use of warfarin, and a history of GI bleed. For low-risk patients, those who received a COX-2 inhibitor had a higher relative risk of a GI bleed event compared with patients who received nonselective NSAIDs. Male gender, higher Charlson Comorbidity Index, history of GI bleed, and diagnosis of osteoarthritis were found to be additional risk factors, but age and state of health plan were not. Within the total study population and the high-risk population, the test of the proportional hazards assumption (using the plot of the scaled Shoenfeld residuals by transformed time for each risk factor) revealed significant evidence of nonproportional hazards for the risk factor of a preperiod history of GI bleed. Stratification by preperiod history of GI bleed did not significantly alter the hazard ratio for the comparison between COX-2 inhibitors and nonselective NSAIDs or for the other specific risk factors in the model. Within the low-risk population, the test of the proportional hazards assumption found significant evidence of nonproportional hazards for the risk factor of arthritis indication on a medical claim. Stratification by arthritis indication did not significantly alter the other hazard ratios in the model. Since a higher proportion of patients treated with a COX-2 inhibitor used a gastroprotective agent during the preperiod, 554 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 additional proportional hazards regression models were performed to evaluate whether the GI bleed outcome changed after adjusting for differences in preperiod use of a gastroprotective agent (Table 5). In the total study population and the high-risk population, the hazard ratios changed slightly, but the overall study findings and level of significance were not altered. However, in the low-risk population, patients who received a COX-2 inhibitor no longer had a significantly greater risk of having a GI bleed event compared with patients who received a nonselective NSAID after adjusting for preperiod use of a gastroprotective agent. In all 3 study populations (total population, low-risk, and high-risk), preperiod use of gastroprotective agents was a factor associated with a higher risk of a GI bleed during follow-up. Similar to the previous models, stratified models were performed because there was evidence of nonproportional hazards for the risk factor of history of GI bleed for the total population and the high-risk population and for the risk factor of arthritis indication for the low-risk population. The stratified models did not significantly alter the hazard ratios. ■■ Discussion In this retrospective study of a population of COX-2 inhibitor users who were matched to nonselective NSAID users with similar baseline characteristics, a lower risk of having a GI bleed was not observed among patients receiving COX-2 inhibitors. Despite the fact that the early controlled clinical trials of the COX-2 inhibitors suggested a lower risk of GI bleed, particularly for rofecoxib,9-15 similar results were not demonstrated at the population level within this managed care setting. The results from the present study expand on those from a cross-sectional time series analysis from 1994 through 2002 among patients older than 66 years in Ontario, Canada, where a 41% increase in NSAID utilization (resulting from the increased use of COX-2 inhibitors) was accompanied by a 10% increase in hospitalization rates for upper GI bleed.28 Although causation was not proven, adverse outcomes may result when www.amcp.org Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs high-risk patients who would normally not be prescribed nonselective NSAIDs are treated with COX-2 inhibitors because these medications are perceived to be safer. In the present study, the use of COX-2 inhibitors was not associated with higher rates of GI bleed for the high-risk population, and there was no measurable reduction in the risk of GI bleed for COX-2 inhibitor users compared with nonselective NSAID users. Contrary to the findings of the present study, an earlier observational study published in 2002 examining an administrative health care database in Ontario, Canada, found a higher risk of hospitalization for a GI hemorrhage with the nonselective NSAIDs than with the COX-2 inhibitors (12.6 GI hemorrhages per 1,000 person years for nonselective NSAIDs versus 3.6 to 7.3 GI hemorrhages per 1,000 person years for COX-2 inhibitors, adjusted rate ratio ranged from 1.9 to 4.4 for nonselective NSAIDs versus COX-2 inhibitors).29 While this study was similar in concept to the present study, the 2 studies used different study populations and health systems, different cohort matching techniques, different model covariates, and different definitions of a GI bleed. The Canadian study evaluated a high-risk population (patients aged 66 years and older), while the present evaluated both low- and high-risk patients. In the Canadian study, the GI bleed outcome was limited to hospitalizations for a GI bleed, but the present study examined hospitalizations as well as outpatient claims for a GI bleed because many GI bleeds are treated in the outpatient setting (hence, the absolute rate of GI bleeds was lower in the Canadian study compared with the present study). The Canadian study did not match nonselective NSAID users to COX-2 inhibitor users but, instead, adjusted for covariates using a Cox proportional hazards model. The covariates used within the Canadian 2002 study were different than those used in the present study. One noteworthy difference was that the Canadian study measured past history of a GI bleed over the prior 5-year period, while the present study limited this look-back period to 6 months since the frequent turnover of patients in U.S. managed care plans would not allow for a longer evaluation without significantly reducing the study sample size. While the differences in study design prohibit a direct comparison of the 2 studies, the contrasting results between the studies indicate a need for additional research to further understand the risk of GI bleed among users of COX-2 inhibitors and nonselective NSAIDs. Whether COX-2 inhibitors are cost effective at the population level remains a controversial issue.5-8, 30-32 Based on data from this managed care organization during the fourth quarter of 2004, the average pharmacy ingredient cost per 30 days for a COX-2 inhibitor prescription was $95.70, which, if filled regularly for a 1-year period, would have an annual drug cost of $1,148 per patient. In comparison, the average pharmacy ingredient cost per 30 days for a nonselective NSAID was $16.56, or $199 per patient per year. In other words, it would be possible to treat www.amcp.org TABLE 4 Hazard Ratio (95% Confidence Interval) of a GI Bleed Event* Total Population (N = 70,014) NSAID Selection COX-2 inhibitor vs. nonselective NSAID Other Factors Age Gender (male vs. female) 1.07 (0.90-1.26) Low-Risk Population (N = 29,547) High-Risk Population† (N = 40,467) 2.05 (1.21-3.48)‡ 0.995 (0.84-1.19) 1.61 (1.50-1.72)‡ 1.25 (0.97-1.62) 1.53 (1.39-1.68)‡ 1.34 (1.13-1.58)‡ 1.64 (1.02-2.63)‡ 1.31 (1.09-1.56)‡ 0.55 (0.17-1.77) 0.94 (0.69-1.26) 0.75 (0.30-1.87) 0.55 (0.14-2.28) 0.66 (0.48-0.91)‡ 0.81 (0.48-1.36) 0.38 (0.09-1.54) 0.77 (0.47-1.25) State of health plan Washington 0.93 (0.69-1.24) vs. California Texas vs. California 0.68 (0.50-0.92)‡ Oregon 0.79 (0.49-1.29) vs. California Oklahoma 0.69 (0.43-1.09) vs. California Charlson Comorbidity Index 1.11 (1.04-1.18)‡ 1.26 (1.08-1.49)‡ 1.10 (1.03-1.17)‡ Corticosteroid use (yes vs. no) 1.27 (0.99-1.61) – 1.20 (0.94-1.54) Warfarin use (yes vs. no) 1.75 (1.39-2.19)‡ – 1.68 (1.33-2.12)‡ History of GI bleed (yes vs. no) 5.35 (4.00-7.15)‡ 10.30 (5.19-20.45)‡ 4.82 (3.52-6.60)‡ Arthritis indication Rheumatoid arthritis vs. none Osteoarthritis vs. none 1.05 (0.69-1.61) 1.44 (0.45-4.65) 0.98 (0.62-1.55) 1.10 (0.91-1.32) 2.01 (1.10-3.67)‡ 1.03 (0.85-1.26) * A separate regression model was performed for each of the 3 populations: the total study population, low-risk population, and high-risk population. † High-risk was defined as a patient with (a) age older than 65 years, or (b) recent history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a GI bleed. ‡ Indicates a level of significance of P < 0.05. COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflammatory drug. 6 patients with nonselective NSAIDs for the same drug cost incurred for 1 patient taking a COX-2 drug. Since patients using the COX-2 inhibitors did not demonstrate a reduction in GI bleeding in the present study, the cost benefit of the COX-2 inhibitors is questionable. Previous research has shown that use of COX-2 inhibitors may not be cost effective among patients with low or average GI risk,30 and the findings from the current study suggest that COX-2 inhibitors may not be cost effective even in high-risk populations. While the use of nonselective NSAIDs for high-risk patients may not be appropriate, the use of COX-2 inhibitors within high-risk populations must also be questioned. The use of COX-2 inhibitors may be appropriate for some patients; however, Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 555 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs Hazard Ratio (95% Confidence Interval) of a cohort. Because patients were not randomized to treatment GI Bleed Event, Adjusting for Preperiod Use group, there exists the potential for selection bias between the COX-2 inhibitor and nonselective NSAID cohorts. Physicians of a Gastroprotective Agent* may have decided to prescribe COX-2 inhibitors instead of Total Low-Risk High-Risk nonselective NSAIDs for their higher-risk patients. Population Population Population† In many of the health plans within this managed care (N = 70,014) (N = 29,547) (N = 40,467) population, COX-2 inhibitors were subjected to prior authorization, NSAID Selection COX-2 inhibitor vs. 1.01 (0.85-1.19) 1.65 (0.96-2.85) 0.96 (0.80-1.14) which required patients to meet certain clinical characteristics nonselective NSAID prior to receiving an authorization for coverage of a COX-2 Other Factors inhibitor. Such criteria commonly include age, diagnosis, GI Age 1.61 (1.51-1.73)‡ 1.22 (0.94-1.57) 1.54 (1.40-1.70)‡ bleed history, and concomitant medication use. Thus, patients Gender 1.35 (1.15-1.60)‡ 1.63 (1.02-2.62)‡ 1.32 (1.10-1.57)‡ receiving COX-2 inhibitors may have had different GI risk (male vs. female) factors than those receiving nonselective NSAIDs. Due to database State of health plan limitations, it was not possible to determine the proportion of Washington 0.95 (0.71-1.28) 0.56 (0.18-1.79) 0.96 (0.71-1.30) subjects within each cohort that required prior authorization to vs. California obtain a COX-2 inhibitor. However, propensity matching was Texas vs. California 0.69 (0.51-0.92)‡ 0.74 (0.30-1.86) 0.67 (0.49-0.92)‡ Oregon 0.81 (0.50-1.32) 0.54 (0.13-2.22) 0.83 (0.50-1.40) used to minimize demographic, clinical, and geographical vs. California differences between the cohorts. Oklahoma 0.68 (0.43-1.08) 0.38 (0.09-1.58) 0.76 (0.47-1.24) The percentage of patients in the Medicare + Choice health vs. California plan was statistically different for the COX-2 inhibitor group Charlson 1.10 (1.04-1.17)‡ 1.30 (1.11-1.51)‡ 1.09 (1.02-1.17)‡ and the nonselective NSAID group (53.1% versus 55.7%). The Comorbidity Index practical significance of this 2.6-point absolute difference is not Corticosteroid use 1.23 (0.97-1.57) – 1.18 (0.92-1.52) clear, but the Medicare + Choice health plan may have had a (yes vs. no) maximum annual pharmacy benefit ranging from $500 to Warfarin use 1.78 (1.42-2.23)‡ – 1.70 (1.35-2.15)‡ (yes vs. no) $2,000 and generic-only benefits (with or without limits), depending on the benefit year and the geographical county of the History of GI bleed 4.70 (3.50-6.30)‡ 6.41 (3.17-12.95)‡ 4.40 (3.20-6.04)‡ (yes vs. no) member. Arthritis indication The 6-month preperiod may not have been long enough to Rheumatoid 1.03 (0.68-1.58) 1.38 (0.43-4.45) 0.97 (0.62-1.53) capture the patient’s entire past history. For example, if a patient arthritis vs. none had a GI bleed prior to the preperiod, his or her history of GI Osteoarthritis 1.10 (0.91-1.32) 1.98 (1.08-3.60)‡ 1.03 (0.85-1.26) bleed would not have been captured. All of the potential risk vs. none factors for a GI bleed could not be captured in this claims Preperiod 1.58 (1.30-1.91)‡ 2.85 (1.70-4.80)‡ 1.43 (1.15-1.76)‡ database, including use of alcohol, tobacco, aspirin or over-thegastroprotective agent use (yes vs. no) counter nonselective NSAIDs. Patients may have had other risk * A separate regression model was performed for each of the 3 populations: factors for GI bleed (e.g., bleeding disorders) that were not the total study population, low-risk population, and high-risk population. considered in this analysis. As a result, some high-risk patients † High-risk was defined as a patient with (a) age older than 65 years, or (b) recent could have been mismatched or incorrectly classified as history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a GI bleed. low-risk patients because of missing data. Within the low-risk ‡ Indicates a level of significance of P < 0.05. population, the higher rate of GI bleed among patients COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflam- receiving COX-2 inhibitors versus patients receiving nonselective matory drug. NSAIDs may be the result of incorrect (false assignment) classification of high-risk patients into the low-risk population. In contrast, patients in the high-risk population were more likely further research is warranted to determine the specific patient to be correctly classified since they were required to have a high-risk condition (i.e., age older than 65 years, recent history populations where the COX-2 inhibitors are cost effective. of warfarin or corticosteroid use, or a recent hospitalization for Limitations a GI bleed) to be included in this population. Since these results were obtained among health maintenance Potential study limitations should be considered when generalizing these results to other populations. Although patients were organizations and Medicare + Choice patients in 3 western states, matched by propensity score to control for possible selection Texas, and Oklahoma, similar results may not be observed among bias, the matching may not have accounted for all of the factors populations with different demographic or socioeconomic characthat could have led to adverse selection for the COX-2 inhibitor teristics (such as the Medicaid population). These results were TABLE 5 556 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs subject to the definition used to identify a case of GI bleed. A GI bleed event was defined as an inpatient hospitalization with a diagnosis representing GI bleed or at least 2 medical claims with a primary (first-listed) diagnosis representing a GI bleed. Medical records could be used to validate this definition of a GI bleed; however, examination of patient medical records was beyond the scope of this study. Further study of the accuracy of medical claims diagnoses in identifying actual GI bleeds would add to the literature. Applying a less stringent or more stringent definition for a GI bleed could have changed the study outcomes. In order to be included in the analysis, patients were only required to have 3 months of continuous enrollment in the health plan following their fill of a COX-2 inhibitor or nonselective NSAID. While some patients could have been followed for a maximum of 1 year, others were followed for a shorter period if they disenrolled from the health plan, switched or discontinued their index class of medication (COX-2 inhibitor or nonselective NSAID), or had a GI bleed. Further research is needed to understand whether these results would have been different if patients were studied over a longer follow-up period. While rates of concomitant use of proton pump inhibitors or other gastrointestinal protective agents were reported, an evaluation of the impact of gastroprotective agents on GI outcomes was beyond the scope of this study, which was designed to examine the relative risk of having a GI bleed among patients using COX-2 inhibitors or nonselective NSAIDs within this managed care population. Introducing the use of gastroprotective agents as a variable within the analysis could have confounded the results since the use of gastroprotective agents can represent either a risk factor or a study outcome. While the use of gastroprotective agents in combination with COX-2 inhibitors or nonselective NSAIDs merits further research, a study designed specifically to measure the concomitant use of gastroprotective agents such as a case-control study could help eliminate some of the confouning associated with this measure. ■■ Conclusion This study provides insight into the rates of GI bleeding among a large population of managed care patients initiated on COX-2 inhibitors or nonselective NSAIDs. Patients using a COX-2 inhibitor did not have a reduced risk of a GI bleed compared with patients with similar baseline characteristics who were using nonselective NSAIDs. With the high direct-drug cost of COX-2 inhibitors and the uncertain risk of adverse cardiovascular events, further research is needed to reevaluate the appropriate patient populations for cost-effective treatment with COX-2 inhibitors. DISCLOSURES No outside funding supported this study. Author Karen Stockl served as principal author of the study. Study concept and design were contributed primarily by Stockl and author Eunice Y. Chang. Analysis and interpretation www.amcp.org of data were contributed by Stockl, Chang, and author Lori Cyprien. Drafting of the manuscript was the work of Stockl and Chang, and its critical revision was the work of Chang. Statistical expertise was contributed by Cyprien and Chang. Stockl discloses that she owns a small amount of Merck stock; Stockl, Cyprien, and Chang disclose no potential bias or conflict of interest relating to this article. REFERENCES 1. U.S. Food and Drug Administration. FDA issues public health advisory on Vioxx as its manufacturer voluntarily withdraws the product [FDA news release. September 30, 2004]. Available at: http://www.fda.gov/bbs/topics/news/ 2004/NEW01122.html. Accessed April 29, 2005. 2. U.S. Food and Drug Administration. FDA announces important changes and additional warnings for COX-2 selective and nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) [FDA public health advisory, April 7, 2005]. Available at: http://www.fda.gov/cder/drug/advisory/COX2.htm. Accessed April 29, 2005. 3. Ortiz E. Market withdrawal of Vioxx: Is it time to rethink the use of COX-2 inhibitors? J Manag Care Pharm. 2004;10(6):551-54. 4. U.S. Food and Drug Administration. Prescription nonsteroidal antiinflammatory drugs (NSAIDs) [FDA alert for health care professionals, April 7, 2005]. Available at: http://www.fda.gov. Accessed April 29, 2005. 5. Cox ER, Motheral B, Mager D. Verification of a decision analytic model assumption using real-world practice data: implications for the cost effectiveness of cyclo-oxygenase 2 inhibitors (COX-2s). Am J Manag Care. 2003;9:785-94. 6. Stacy J, Shaw E, Arledge MD, Howell-Smith D. Pharmacoeconomic modeling of prior-authorization intervention for COX-2 specific inhibitors in a 3-tier copay plan. J Manag Care Pharm. 2003;9(4):327-34. 7. Bull S, Conell C, Campen DH. Relationship of clinical factors to the use of COX-2 selective NSAIDs within an arthritis population in a large HMO. J Manag Care Pharm. 2002;8(4):252-58. 8. Curtiss FR. Cost-effective use of COX-2 drugs and NSAIDs. J Manag Care Pharm. 2002;8(4):295-96. 9. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med. 2000;343:1520-28. 10. Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. JAMA. 2000;284:1247-55. 11. Langman MJ, Jensen DM, Watson DJ, et al. Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA. 1999;282:1929-33. 12. Hawkey C, Laine L, Simon T, et al. Comparison of the effect of rofecoxib (a cyclooxygenase 2 inhibitor), ibuprofen, and placebo on the gastroduodenal mucosa of patients with osteoarthritis: A randomized, double-blind, placebocontrolled trial. Arthritis Rheum. 2000;43:370-77. 13. Laine L, Harper S, Simon T, et al. A randomized trial comparing the effect of rofecoxib, a cyclooxygenase 2-specific inhibitor, with that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology. 1999;117:776-83. 14. Simon LE, Weaver AL, Graham DY, et al. Anti-inflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis: a randomized controlled trial. JAMA. 1999;282:1921-28. 15. Rofecoxib (Vioxx) [prescribing information]. NDA 21-042/S-018 and NDA 21-052/S-012. Available at: http://www.fda.gov/cder/foi/label/2003/ 21042se8-018,21052se8-012_vioxx_lbl.pdf. Accessed July 27, 2005. 16. Stockl K, Vanderplas A, Tafesse E, Chang E. Costs of lower-extremity ulcers among patients with diabetes. Diabetes Care. 2004;27(9):2129-34. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 557 Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs 17. White TJ, Vanderplas A, Chang E, Dezii CM, Abrams GD. The costs of non-adherence to oral antihyperglycemic medication in individuals with diabetes mellitus and concomitant diabetes mellitus and cardiovascular disease in a managed care environment. Dis Manag Health Outcomes. 2004:12(3):181-88. 25. Schoenfeld D. Partial residuals for the proportional hazards regression model. Biometrika. 1982:69:239-41. 18. Yu-Isenberg KS, Fontes CL, Wan GJ, Geissler EC, Harada AS. Acute and continuation treatment adequacy with venlafaxine extended release compared with fluoxetine. Pharmacotherapy. 2004;24(1):33-40. 27. Therneau T, Grambsch P. Modeling Survival Data: Extending the Cox Model. New York: Springer-Verlag; 2000. 19. Stockl K, Vanderplas AM, Nicklasson L. A comparison of costs for four oral antidiabetic regimens within a managed care population. Manag Care Interface. 2003;16(7):31-36. 20. Melikian C, White TJ, Vanderplas A, Dezii CM, Chang E. Adherence to oral antidiabetic therapy in a managed care organization: A comparison of monotherapy, combination therapy, and fixed-dose combination therapy. Clin Ther. 2002;24(3):460-67. 21. Celebrex NDA 20-998, approved December 31, 1998. Available at: http://www.fda.gov/cder/approval/index.htm. Accessed July 27, 2005. 22. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med. 1997;127(8S):757-63. 23. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis. 1987;40:373-83. 24. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9CM administrative databases. J Clin Epidemiol. 1992;45: 613-19. 558 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 26. Grambsch P, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994:81:515-26. 28. Mamdani M, Juurlink DN, Kopp A, Naglie G, Austin PC, Laupacis A. Gastrointestinal bleeding after the introduction of COX 2 inhibitors: ecological study. BMJ. 2004;328:1415-16. 29. Mamdani M, Rochon PA, Juurlink DN et al. Observational study of upper gastrointestinal haemorrhage in elderly patients given selective cyclo-oxygenase2 inhibitors or conventional non-steroidal anti-inflammatory drugs. BMJ. 2002;325:624-29. 30. Spiegel BMR, Targownik L, Dulai GS, Gralnek IM. The cost-effectiveness of cyclooxygenase-2 selective inhibitors in the management of chronic arthritis. Ann Intern Med. 2003;138:795-806. 31. Pellissier JM, Straus WL, Watson DJ, Kong SX, Harper SE. Economic evaluation of rofecoxib versus non-selective nonsteroidal anti-inflammatory drugs for the treatment of osteoarthritis. Clin Ther. 2001;23:1061-79. 32. Feldman M, McMahon AT. Do cyclooxygenase-2 inhibitors provide benefits similar to those of traditional nonsteroidal anti-inflammatory drugs, with less gastrointestinal toxicity? Ann Intern Med. 2000;132:134-43. www.amcp.org ORIGINAL RESEARCH Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes SHARASHCHANDRA SHETTY, PhD; KRISTINA SECNIK, RPh, MPH, PhD; and ALAN K. OGLESBY, MPH ABSTRACT OBJECTIVE: Glycosylated hemoglobin (A1c) is a well-established measure of glycemic control, and evidence suggests that maintaining an acceptable A1c level may be associated with lower treatment costs in adults with diabetes. Understanding the impact on total treatment costs of staying at the target A1c level is of great importance to managed care organizations. The goal of this study was to determine whether type 2 diabetes patients at or below the target A1c level of 7% had lower diabetes-related costs compared with patients above an A1c level of 7%. METHODS: This study was a retrospective database analysis using eligibility data, medical and pharmacy claims data, and laboratory data from a large U.S. health care organization. Patients were included in the study if they had 2 or more claims for type 2 diabetes (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes 250.x0 or 250.x2) and at least 1 A1c value (first such date defined as the index date) during the 12-month period from January 1, 2002, through December 31, 2002. Patients with 2 or more medical claims for type 1 diabetes (ICD-9-CM codes 250.x1 or 250.x3) were excluded from the study. Study patients were divided into 2 groups, those at the target A1c level (≤7%) and those at the above-target A1c level (>7%), and were followed for a period of 1 year after their index date. Demographic, clinical, and cost variables were extracted from the administrative database. Multiple linear regression analysis was used to compare treatment costs between patients at the target A1c level and patients above target level. RESULTS: A total of 3,121 patients (46.0%) were identified as being at the target A1c level, and 3,659 patients (54%) were identified as being above the target A1c level during the study period. After controlling for confounding factors, the predicted total diabetes-related cost for the above-target group during the 1-year follow-up period was $1,540 per patient, 32% higher than the total diabetesrelated cost ($1,171) for the at-target group (P <0.001). CONCLUSION: Results of this analysis suggest that managed care members with type 2 diabetes who stayed continuously at the target A1c level of 7% or less over a 1-year follow-up period incurred lower diabetes-related costs compared with managed care members with type 2 diabetes who were continuously over the target A1c level of 7%. KEYWORDS: Type 2 diabetes, A1c level, Glycosylated hemoglobin, Costs, Managed care, Obesity J Manag Care Pharm. 2005;11(7):559-64 Note: An editorial on the subject of this article appears on pages 588-89 of this issue. D iabetes is a common chronic disease that is associated with considerable morbidity and mortality.1 Approximately 18 million people in the United States are diagnosed with diabetes. The American Diabetes Association (ADA) attributed $92 billion in direct medical expenditures to diabetes in 2002. Inadequate glycemic control is thought to be a cause of diabetic complications and higher costs.2,3 Proper management of diabetes can delay complications, reduce mortality, and reduce the costs of diabetes care. Research has shown that aggressive glycemic control can reduce longterm complications in patients with type 1 or type 2 diabetes and result in considerable medical cost savings.2, 4-6 Most of the previous studies have focused on the effects of glycemic control on long-term cost savings. However, evidence of short-term cost savings is often required before studying the long-term implications. A few studies have suggested that better glycemic control may result in cost savings within a short period of time.3,7,8 Analysis of retrospective administrative claims and laboratory data by Gilmer et al. showed that inadequate glycemic control was associated with greater health care costs over a 3-year period.3 For every 1% increase in glycosylated hemoglobin (A1c), Gilmer et al. found that health care costs rose 7% over the next 3 years. Menzin et al. found, in a study using a retrospective cohort design, a reduced rate of admission for short-term complications and reduced medical charges for these complications in patients with better glycemic control.7 In a study using data from a staff-model health maintenance organization, Wagner et al. suggested that a sustained reduction in A1c level was associated with significant cost savings within 1 to 2 years of improvement.8 The studies either used change in A1c level or created categories of A1c levels while studying the effect of glycemic control on health care costs. In the present study, we examined the relationship of diabetes-related costs in patients with type 2 diabetes who stayed at the target A1c level of ≤7% compared with patients with type 2 diabetes who stayed above the target A1c level (had A1c levels >7%). Authors SHARASHCHANDRA SHETTY, PhD, is a senior researcher, i3 Magnifi, Eden Prairie, Minnesota; KRISTINA SECNIK, RPh, MPH, PhD, is a senior scientist, Global Health Outcomes; ALAN K. OGLESBY, MPH, is a senior health outcomes consultant, U.S. Health Outcomes, Eli Lilly and Company, Indianapolis, Indiana. AUTHOR CORRESPONDENCE: Sharashchandra Shetty, PhD, Senior Researcher, i3 Magnifi, 12125 Technology Dr., MN002-0258, Eden Prairie, MN 55344. Tel: (952) 833-8002; Fax: (952) 833-6045; E-mail: [email protected] Copyright© 2005, Academy of Managed Care Pharmacy. All rights reserved. www.amcp.org ■■ Methods This study is a retrospective, longitudinal database analysis using eligibility data, medical and pharmacy administrative claims data, and laboratory data from a large U.S. managed care organization (MCO). All are commercial, preferred-provider organization-model regional health plans of a national MCO. The MCO had approximately 5.4 million members during 2002. The individuals covered by this health plan are geographically diverse across the United States. The MCO provides fully Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 559 Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes insured coverage for physician, hospital, and pharmacy services. Study patients were identified during the period January 1, 2002, through December 31, 2002. The index date was the first available A1c laboratory value recorded during the subject identification period. Patients were included in this study if they met the following criteria: (1) had 2 or more claims for type 2 diabetes in either the primary or secondary position on physician or hospital claims (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes 250.x0 or 250.x2 were used to identify patients with type 2 diabetes.); (2) had at least 1 prescription for an oral hypoglycemic agent and/or insulin; (3) had at least 1 available A1c value; (4) were commercially insured with a drug benefit; and (5) had at least 6 months of continuous enrollment prior to the index date and at least 12 months of continuous enrollment following the index date. Patients with 2 or more claims for type 1 diabetes (ICD-9-CM codes 250.x1 or 250.x3) were excluded from the study. Study patients were divided into those at the target A1c level (≤7%) and those above the target A1c level (>7%). Patients with more than 1 A1c value were required to be at target level or above target level to be included in the at-target group or abovetarget group, respectively. Patients whose A1c values were not continuously at target level or above target level during the 1-year follow-up period were excluded from the study. Demographic, clinical, and cost variables were extracted from the research database for each subject. The demographic variables included age, gender, and health plan region. The clinical variables included prescribing physician specialty and the presence of select comorbid conditions. The following comorbid conditions were identified during the preindex and the postindex periods: hypertension, congestive heart failure, ischemic heart disease, atherosclerosis, dyslipidemia, retinopathy, nephropathy, neuropathy, diseases of the extremities, obesity, and albuminuria. The practice specialty of the physician prescribing the first hypoglycemic agent during the subject identification period was determined, and 4 dichotomous variables indicating the specialty of the physicians were created (general practitioner, internist, endocrinologist, and other specialty). To account for the difference in burden posed by comorbidities between the 2 groups, the presence of select comorbid conditions was examined during the study period. Comorbid conditions were identified using ICD-9-CM codes in any position on the physician and hospital claims and using the National Drug Codes for prescription drugs in the pharmacy claims. Total baseline cost during the 6-month preindex period was also estimated to serve as a proxy for general health status of each subject. The direct medical costs associated with the treatment of diabetes were estimated 1 year following the index date. The direct medical costs included all physician office visit, outpatient visit, inpatient, emergency room, and lab costs associated with a diagnosis of diabetes in the primary 560 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 position on the medical claims. Pharmacy costs were calculated for oral hypoglycemic agents and insulin during the 1-year follow-up period. Costs were defined as the cost to the health plan as well as to the health plan member; i.e., cost in this study included the total amount paid by the health plan as well as any copayment and deductible amounts paid by the health plan member. Statistical Analysis The baseline characteristics of the 2 comparison groups were analyzed descriptively (frequencies and percentages), using paired t tests for continuous variables and chi-square tests for categorical variables. Multiple linear regression analysis was used to estimate the relationship between remaining at target A1c level and diabetes treatment cost. The primary independent variable in the regression model was the target group indicator. The regression analysis controlled for demographic and clinical confounders, including age, gender, specialty of the physician prescribing the index medication, presence of comorbid conditions, and total baseline costs. The dependent variable in the regression model was the total diabetes-related costs during the 1-year follow-up period. The distribution of the cost data was skewed. Testing for heteroscedasticity indicated that error variances were not constant. Logarithmic transformation of the cost data was done prior to the analysis to make the data normal, and the regression analysis was conducted using robust standard errors. Since all patients in the study population had a diabetes-related cost, zero values for cost were not a concern for this study. The adjusted log means were transformed to a dollar scale using a smearing estimator in order to obtain an unbiased estimate of mean diabetes-related cost.9,10 A subanalysis was also conducted to identify significant demographic predictors and comorbidities that are associated with increased diabetes-related costs when stratified by A1c level. Multiple linear regressions were conducted separately in the at-target group and above-target group to identify significant predictors of costs. These regression models included the same set of independent variables described above except for the target group indicator. All analyses were conducted using the SAS software, version 8.2.11 ■■ Results Prior to application of the exclusion criteria, 170,566 patients were identified with 2 or more claims for type 2 diabetes during calendar year 2002, representing a prevalence of approximately 3.1% in the MCO population of 5.4 million members. After application of the exclusion criteria (Table 1), 8,991 patients were identified with at least 2 or more claims for type 2 diabetes, at least 1 pharmacy claim for a hypoglycemic agent, at least 1 laboratory value for A1c, and continuous enrollment during the study period. Patients with 2 or more claims for type 1 diabetes were excluded from the study (Table 1). www.amcp.org Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes TABLE 1 Sample Selection Criteria Number of Patients Remaining (%) Number of Patients Dropped (%) Patients identified with 2 or more claims for type 2 diabetes from January 1, 2002, through December 31, 2002 170,566 (100) – Patients with 2 or more claims for type 1 diabetes from January 1, 2002, through December 31, 2002 110,042 (64.5) 60,524 (35.5) At least 1 claim for oral hypoglycemic agent or insulin from January 1, 2002, through December 31, 2002 68,518 (40.2) 41,524 (24.3) At least 1 lab value for A1c from January 1, 2002, through December 31, 2002 31,807 (18.7) 36,711 (21.5) Continuous enrollment 6 months prior to index date and 1 year following the index date 8,991 (5.3) 22,816 (13.4) Patients not continuously at either ≤7% or >7% A1c level during the 1-year follow-up period 6,780 (4.0) 2,211 (1.3) A1c = glycosylated hemoglobin. A total of 2,211 patients were excluded from the study because they were not continuously either at the target or above the target A1c level during the 1-year follow-up period. Of the 6,780 patients who were included in this study, 3,121 (46%) were identified as being continuously at the target A1c level, and 3,659 patients (54%) were identified as being above the target A1c level for the 1-year period following the index A1c value. Table 2 shows the descriptive characteristics of the 2 groups. The at-target group (a) consisted of patients who were significantly older (mean 53.9 years) as compared with the above-target group (mean 52.3 years, P <0.001); (b) had a significantly lower percentage of patients in age groups 31-40 and 41-50 and a higher percentage of patients in the above-60-years age group (P <0.001); (c) had a slightly lower proportion of patients from health plans in the Midwest and a slightly higher proportion of patients from health plans in the South (P = 0.004); (d) had a higher percentage of patients who received only oral hypoglycemic agents (P <0.001), a lower percentage who received only insulin, and a lower percentage who received a combination of insulin and oral agents (P <0.001); (e) had a higher percentage of patients with dyslipidemia (P <0.001) and a lower percentage of patients with retinopathy (P = 0.007), neuropathy (P = 0.035), or diseases of the extremities (P = 0.003); and (f) had a lower percentage of patients who had an internist prescribe the index hypoglycemic medication (P < 0.001) but a higher percentage of patients who had an endocrinologist prescribe the index hypoglycemic medication (P = 0.004). www.amcp.org Costs during the 6-month baseline period were significantly (27%) higher in the at-target group ($2,419) compared with the above-target group ($1,911, P <0.001). Prior to adjustment of costs, the at-target group during the follow-up period had significantly lower medical costs, pharmacy costs, and total diabetes-related costs compared with the above-target group (all comparisons, P < 0.001, Table 2). Table 3 presents the results of multiple regression analysis, comparing the total diabetes-related costs of the at-target and the above-target groups. The results of regression analysis are interpreted in the following way: At an A1c level of 7%, the expected cost of a subject aged 65 years with hypertension was higher than the expected cost of a subject aged 30 years without hypertension. The results of regression analysis revealed that the at-target group had significantly lower total diabetes costs (P <0.001) as compared with the above-target group after adjusting for confounding factors. After appropriate log retransformation using the smearing estimator, it was found that the predicted total diabetes-related cost for the above-target group during the 1-year follow-up period was $1,540 per patient, 32% higher than the total diabetes-related cost ($1,171) for the at-target group. Patients in the age groups <31 and 31-40 years had significantly lower diabetes-related costs as compared with patients older than 60 years. Patients who had an endocrinologist prescribe the first hypoglycemic medication during the study period had higher diabetes-related costs. The higher costs associated with endocrinologists may be related to referral of more complex cases from primary care physicians to endocrinologists. Patients with a comorbid diagnosis of hypertension, dyslipidemia, retinopathy, nephropathy, neuropathy, diseases of the extremities, and obesity had significantly higher diabetesrelated costs. Since some researchers have expressed concerns with the use of logged costs related to the potential difficulties associated with retransformation,12 we also compared the costs using a gamma distribution with a log link (generalized linear model [GLM]).13 This method avoids the retransformation problems associated with log models. The magnitude of costs in the at-target group and the above-target group using the GLM was similar to the results obtained using the logged model, thus providing us confidence in our results. Table 4 displays the results of the subanalyses that identified the significant demographic predictors and comorbidities associated with increased diabetes costs when stratified by A1c level. Patients in the age groups <31 and 31-40 years in the at-target group had significantly lower total diabetes costs as compared with patients older than 60 years. However, the diabetes-related costs of the <31 and 31-40 years age groups in the above-target group were not significantly different from the costs of those older than 60 years, but patients in the age group Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 561 Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes 51-60 years had significantly higher total diabetes costs as compared with patients older than 60 years. Patients with dyslipidemia, retinopathy, nephropathy, neuropathy, and diseases of the extremities had significantly higher diabetes costs in both groups. The major difference in the 2 groups was that obesity was significant in explaining the higher total diabetes costs in the above-target group whereas it was not significant in the at-target group. ■■ Discussion A retrospective study using eligibility data, medical and pharmacy claims, and laboratory data for 6,780 patients with type 2 diabetes was conducted to determine the potential economic benefits associated with glycemic control using the A1c value of ≤7%. The results of this study found a strong association between glycemic control and total diabetes-related costs. Patients in the at-target group had significantly lower total diabetes costs as compared with patients in the above-target group after adjusting for the demographic characteristics, presence of comorbid conditions, and baseline costs. The findings of this study suggest that patients who were at the target A1c level used fewer health care resources related to diabetes as compared with patients who were above the target level, which is reflected in the lower total diabetes costs for the at-target group patients in the 1-year follow-up period. The findings were consistent with some of the earlier studies that studied the relationship between A1c levels and health care costs. Reduced health care costs among patients with better glycemic control may be related to symptomatic relief and improvements in quality of life.14 Wagner et al. suggest that better glycemic control may increase the comfort of the primary care physician and the patient, resulting in reduction in physician visits.8 The at-target-group patients had significantly higher baseline costs as compared with the above-target-group patients. This finding is consistent with the results of some of the other studies that found that patients whose A1c levels improved had increased health care use prior to the decrease in A1c levels.8,15,16 In an effort to keep the health care costs under control, patients and physicians may have increased the intensity of treatment to achieve glycemic control. The importance of the association between obesity and total diabetes costs in the above-target group is of great interest. Obesity was significant in explaining the higher diabetes costs in the above-target-group patients but was not significant in the at-target group although both groups had a similar prevalence of obesity. It may be particularly important to focus on reducing weight in patients who are above the target A1c level. Given the above findings regarding the higher diabetes costs associated with patients with A1c levels continuously above 7%, it seems important to focus on getting diabetic patients in control. Previous research has shown that intervention programs 562 Journal of Managed Care Pharmacy JMCP September 2005 TABLE 2 Characteristic Age† [mean ± SD] Descriptive Characteristics of the Study Population At-Target Group* n = 3,121 Above-Target Group n = 3,659 Number (%) Number (%) P Values 53.9 [9.1] 52.3 [9.1] <0.001 Age, years <31 36 (1.2) 51 (1.4) 0.380 31-40 227 (7.3) 351 (9.6) <0.001 41-50 739 (23.7) 1,028 (28.1) <0.001 51-60 1,331 (42.7) 1,530 (41.8) 0.489 >60 788 (25.3) 699 (19.1) <0.001 Gender Male 1,792 (57.4) 2,156 (58.9) 0.210 Female 1,329 (42.6) 1,503 (41.1) Health plan location Northeast 157 (5.0) 195 (5.3) 0.580 Midwest 727 (23.3) 961 (26.3) 0.004 South 2,043 (65.5) 2,272 (62.1) 0.004 West 194 (6.2) 231 (6.3) 0.869 Type of treatment Oral agents only 2,954 (94.7) 2,976 (81.3) <0.001 Insulin only 61 (2.0) 140 (3.8) <0.001 Oral agents + insulin 106 (3.4) 543 (14.8) <0.001 Comorbidities Hypertension 2,444 (78.3) 2,846 (77.8) 0.601 Congestive heart 79 (2.5) 114 (3.1) 0.149 failure Ischemic heart 495 (15.9) 539 (14.7) 0.197 disease Atherosclerosis 154 (4.9) 165 (4.5) 0.410 Dyslipidemia 2,442 (78.2) 2,681 (73.3) <0.001 Retinopathy 332 (10.6) 466 (12.7) 0.007 Nephropathy 96 (3.1) 109 (3.0) 0.816 Neuropathy 272 (8.7) 374 (10.2) 0.035 Diseases of the 40 (1.3) 82 (2.2) 0.003 extremities Obesity 368 (11.8) 403 (11.0) 0.315 Albuminuria 90 (2.9) 129 (3.5) 0.136 Physician specialty General practitioner 1,332 (42.7) 1,511 (41.3) 0.250 Internist 986 (31.6) 1,302 (35.6) <0.001 Endocrinologist 261 (8.4) 239 (6.5) 0.004 Other specialties 542 (17.4) 607 (16.6) 0.395 Baseline cost† $2,419 [$3,856] $1,911 [$3,236] <0.001 [unadjusted] Outcomes† $ [unadjusted] Diabetes medical $534 [$624] $682 [$1,508] <0.001 costs Diabetes pharmacy $663 [$683] $924 [$778] <0.001 costs Total diabetes costs $1,197 [$969] $1,606 [$1,747] <0.001 * Target group: patients with A1c (glycosylated hemoglobin) value ≤7%; abovetarget group: patients with A1c value >7%. † For continuous variables, mean and standard deviations are presented by [ ]. Vol. 11, No. 7 www.amcp.org Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes TABLE 3 Comparison of Total Diabetes-Related Costs of Patients in the At-Target Group and Patients in the Above-Target Group Characteristic Parameter Estimate (SE) P Values At-target group -0.3044 (0.0189) <0.001 TABLE 4 Identification of Significant Predictors of Total Diabetes-Related Costs of Patients in the At- Target Group and Patients in the Above-Target Group At-Target Group* Age, years P Values Parameter Estimate (SE) P Values <31 -0.4361 (0.1704) 0.010 -0.0418 (0.0874) 0.632 31-40 -0.1880 (0.0661) 0.004 -0.0929 (0.0539) 0.085 41-50 0.0298 (0.0396) 0.451 -0.0214 (0.0380) 0.573 51-60 <31 -0.2023 (0.0896) 0.023 Characteristic 31-40 -0.1365 (0.0416) 0.001 Age, years 41-50 -0.0031 (0.0274) 0.909 51-60 0.0434 (0.0247) 0.080 Gender: male 0.0242 (0.0191) 0.206 Health plan: Northeast 0.0974 (0.0570) 0.087 Health plan: Midwest Health plan: South 0.0425 (0.0424) -0.0209 (0.0401) 0.316 0.602 Family practitioner 0.0418 (0.0285) 0.142 Endocrinologist 0.4266 (0.0382) <0.001 Internist Hypertension Congestive heart failure Ischemic heart disease Atherosclerosis 0.0327 (0.0297) 0.270 0.0728 (0.0237) 0.002 -0.0706 (0.0655) 0.280 0.0284 (0.0278) 0.307 -0.0239 (0.0486) 0.622 Dyslipidemia 0.1928 (0.0234) <0.001 Retinopathy 0.3044 (0.0274) <0.001 Nephropathy 0.2507 (0.0541) <0.001 Neuropathy 0.2227 (0.0312) <0.001 Diseases of the lower extremities 0.3849 (0.0868) <0.001 Obesity 0.0897 (0.0312) 0.004 Albuminuria 0.0612 (0.0476) 0.198 Baseline cost/100 0.0017 (0.0002) <0.001 Adjusted R2 (coefficient of determination) = 0.1191. Reference groups include: above-target group, >60 years, female gender, and health plan in the Western region. that facilitate early diagnosis of diabetes, regular monitoring of each patient’s progress, patient education, and systematic follow-up will be helpful in attaining glycemic goals and ultimately will result in costs savings to the health care system.17,18 Proper screening, early diagnosis, and effective management of diabetes will likely ensure a higher percentage of patients at target A1c levels. The use of effective drug therapy, along with patient education and systematic follow-up, may result in maintaining patients at target levels and may help reduce costs associated with diabetes treatment. Limitations The findings of this study must be considered within the limitations of the data and study design. First, the observational study design does not permit causal inference of the results. www.amcp.org Above-Target Group Parameter Estimate (SE) 0.0146 (0.0357) 0.682 0.0690 (0.0345) 0.045 Gender: male 0.0422 (0.0285) 0.138 0.0090 (0.0259) 0.727 Health plan: Northeast 0.1539 (0.0853) 0.071 0.0594 (0.0766) 0.438 Health plan: Midwest 0.0083 (0.0646) 0.897 0.0756 (0.0561) 0.178 Health plan: South -0.0188 (0.0602) 0.754 -0.0220 (0.0539) 0.682 Family practitioner 0.0337 (0.0415) 0.417 0.0461 (0.0392) 0.240 Endocrinologist 0.4463 (0.0553) <0.001 0.4200 (0.0521) <0.001 Internist 0.0372 (0.0443) 0.401 0.0252 (0.0402) 0.530 Hypertension 0.0666 (0.0333) 0.046 0.0710 (0.0334) 0.033 Congestive heart failure -0.0676 (0.1061) 0.524 -0.0870 (0.0834) 0.297 Ischemic heart disease 0.0258 (0.0421) 0.539 0.0294 (0.0370) 0.426 Atherosclerosis -0.0447 (0.0728) 0.538 Dyslipidemia 0.1515 (0.0369) <0.001 Retinopathy 0.3077 (0.0393) Nephropathy 0.2422 (0.0835) Neuropathy Diseases of the lower extremities 0.0002 (0.0650) 0.996 0.2237 (0.0301) <0.001 <0.001 0.3030 (0.0376) <0.001 0.003 0.2576 (0.0704) <0.001 0.1557 (0.0478) 0.001 0.2709 (0.0412) <0.001 0.4228 (0.1238) <0.001 0.3563 (0.1143) 0.001 Obesity 0.0165 (0.0474) 0.727 0.1549 (0.0409) <0.001 Albuminuria 0.1209 (0.0772) 0.117 0.0117 (0.0611) 0.848 Baseline cost/100 0.0013 (0.0004) 0.001 0.0022 (0.0005) <0.001 * Target group: patients with A1c (glycosylated hemoglobin) value ≤7%; above-target group: patients with A1c value >7%. Adjusted R2 (coefficient of determination) for at-target group=0.0762; adjusted R2 for above-target group=0.1071. Reference groups include: >60 years, female gender, and health plan in the Western region. Further, it could be determined whether patients remained at the target A1c level only if they had follow-up visits with their physicians. We made the assumption that the proportion of patients scheduling a follow-up visit in the above-target group is distributed similar to the at-target group. Patients who achieved their target A1c level are probably less likely to have a follow-up visit scheduled as compared with patients who were above target level. We may, therefore, have overestimated the number of patients at target if the number of follow-up visits decreased dramatically after the target level was achieved. However, a post hoc analysis revealed that the number of Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 563 Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes follow-up visits was higher among the at-target-group patients as compared with the above-target-group patients, thus limiting the possibility of overestimation of the number of patients at target (data not presented). The present study is limited by the inability to control for the number of A1c tests, the number of diabetes-related ambulatory visits, and the type of oral hypoglycemic agent. Reporting the values for these 3 variables would have provided a more comprehensive description of the study population. The study is also limited by the lack of control for the severity of disease, but the baseline costs may be a crude proxy for disease severity. We measured only diabetes-related costs, defined as (a) the health plan and member costs for only medical claims with a primary diagnosis of diabetes and (b) pharmacy costs for diabetes drugs and insulin. We did not measure total medical costs for these 2 groups of patients with type 2 diabetes. Patients were excluded from the study if their A1c values were not continuously at target level or above target level during the 1-year follow-up period. Since this excluded group of 2,211 patients represented 25% of the potential study population, the results of the study may not be generalizable to the general population of patients with type 2 diabetes. The study population was a sample of managed care members. Therefore, the results of this study are applicable to the management of type 2 diabetes in a managed care setting. It may not be possible to generalize the results to a nonmanaged care population. In addition, administrative claims data have some limitations that are common to all analyses conducted using this data source. Claims data are collected for the purpose of payment and not for research. They may be subject to possible coding errors. Despite the limitations, administrative claims remain a powerful source of data for research. Claims data allow for examination of health care utilization and associated expenditures in real-world settings. This analysis has important policy implications. The analysis suggests that better glycemic control is associated with significantly lower costs over a 1-year period. Further research should be conducted to explore if glycemic control is associated with cost savings outside the managed care environment. ■■ Conclusion Diabetes patients who were continuously at the target A1c level of ≤7 had significantly lower diabetes-related costs over a 1-year follow-up period compared with diabetes patients who were continuously above the target A1c level. Managed care efforts to help diabetes patients attain target A1c levels may reduce total diabetes-related medical costs. DISCLOSURES Funding for this research was provided by Eli Lilly and Company and was obtained by authors Kristina Secnik and Alan K. Oglesby, who are employees and shareholders of Eli Lilly and Company. Sharashchandra Shetty is an 564 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 employee of i3 Magnifi and a shareholder of UnitedHealth Group; i3 Magnifi, a division of UnitedHealth Group, receives funding from several pharmaceutical companies for outcomes research studies. The study results in this article were presented as a poster at the International Society for Pharmacoeconomics and Outcomes Research 10th Annual International Meeting, Washington, DC, on May 17, 2005. Shetty served as principal author of the study. Study concept and design, analysis and interpretation of data, and statistical expertise were contributed by all authors. Drafting of the manuscript was primarily the work of Shetty, and its critical revision was the work of Secnik and Oglesby. Administrative, technical, and/or material support was provided by i3 Magnifi. REFERENCES 1. American Diabetes Association. Diabetes statistics. Available at: http://www.diabetes.org/diabetes-statistics/national-diabetes-fact-sheet.jsp. Accessed November 3, 2004. 2. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and the progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329:977-86. 3. Gilmer TP, O’Connor PJ, Manning WG, Rush WA. The cost to health plans of poor glycemic control. Diabetes Care. 1997;20:1847-53. 4. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet. 1998;352:837-53. 5. Eastman RC, Javitt JC, Herman WH, et al. Model of complications of NIDDM, II: analysis of health benefits and cost-effectiveness of treating NIDDM with the goal of normo-glycemia. Diabetes Care. 1997;20:735-44. 6. Brown JB, Pedula KL, Bakst AW. The progressive cost of complications in type 2 diabetes mellitus. Arch Intern Med. 1999;159:1873-80. 7. Menzin J, Langley-Hawthorne C, Friedman M, Boulanger L, Cavanaugh R. Potential short-term economic benefits of improved glycemic control: a managed care perspective. Diabetes Care. 2001;24:51-55. 8. Wagner EH, Sandhu N, Newton KH, McCulloch DK, Ramsey SD, Grothaus LC. Effect of improved glycemic control on health care costs and utilization. JAMA 2001;285(2):182-89. 9. Duan N. Smearing estimate: a nonparametric retransformation method. J Am Stat Assoc. 1983;78:605-10. 10. Rutten-van Molken MP, van Doorslaer EK, van Vilet RC. Statistical analysis of cost outcomes in a randomized controlled clinical trial. Health Econ. 1994; 3:333-45. 11. SAS statistical software, Version 8.2, 2001; SAS Institute, Inc., Cary, NC. 12. Manning WG. The logged dependent variable, heteroscedasticity, and the retransformation problem. J Health Econ. 1998;17(3):283-95. 13. Blough DK, Madden CW, Hornbrook MC. Modeling risk using generalized linear models. J Health Econ. 1999;18(2):153-71. 14. Testa MA, Simonson DC. Health economic benefits and quality of life during improved glycemic control in patients with type 2 diabetes mellitus: a randomized, controlled, double-blind trial. JAMA. 1998;280:1490-96. 15. Wagner EH, Curry SJ, Grothaus L, Saunders KW, McBride CM. The impact of smoking and quitting on health care use. Arch Intern Med. 1995; 155:1789-95. 16. O’Connor PJ, Crabtree BF, Abourizk NN. Longitudinal study of a diabetes education and care intervention: predictors of improved glycemic control. J Am Board Fam Pract. 1992;5:381-87. 17. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus: Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 1997;20:1183-97. 18. Solberg LI, Reger LA, Pearson TL, et al. Using continuous quality improvement to improve diabetes care in populations: the IDEAL model. Improving care for Diabetics through Empowerment Active collaboration and Leadership. Jt Comm J Qual Improv. 1997;23(11):581-92. www.amcp.org ORIGINAL RESEARCH Economic Burden of Anemia in an Insured Population ALLEN R. NISSENSON, MD, FACP; SALLY WADE, MPH; L. TIM GOODNOUGH, MD; KEVIN KNIGHT, MD, MPH; and ROBERT W. DUBOIS, MD, PhD ABSTRACT OBJECTIVE: Anemia is a common hematological disorder characterized by reduced hemoglobin concentrations. Despite information on prevalence and associated outcomes, little is known about the impact of anemia on health care utilization and costs. This study examines anemia prevalence and associated medical costs and utilization, using administrative claims for adults newly diagnosed with anemia, including up to 12 months of follow-up. METHODS: Patients predisposed to anemia, based on selected comorbid conditions (chronic kidney disease, human immunodeficiency virus, rheumatoid arthritis, inflammatory bowel disease, congestive heart failure, and solid-tumor cancers), were identified. Costs for anemic patients and a random sample of nonanemic patients with these conditions were compared. Associations were evaluated after adjustment for potential confounders using a regression model. Clinical care patterns were examined overall and by condition. RESULTS: Anemia was observed in 3.5% (81,423) of approximately 2.3 million health plan members in 2000; 15% of anemic patients received an identified treatment, with transfusion being the most frequent intervention. Utilization and costs were significantly higher for anemic patients (P < 0.001). Average annualized per-patient costs were $14,535 for anemic patients (55% outpatient, 33% inpatient, 13% pharmacy), 54% higher than the $9,451 average cost for nonanemic patients (45% outpatient, 36% inpatient, 19% pharmacy). After adjustment for age, other comorbidities (e.g., chronic kidney disease and cancer), sex, and insurance type (indemnity, preferred provider organization/point of service, or health maintenance organization, in the Medstat MarketScan database), anemic patients had average costs that were more than twice the adjusted costs of nonanemic patients. CONCLUSION: Medical costs for anemic patients are as much as twice those for nonanemic patients with the same comorbid conditions. KEYWORDS: Anemia, Chronic conditions, Utilization, Medical costs J Manag Care Pharm. 2005;11(7):565-74 Authors ALLEN R. NISSENSON, MD, FACP, is a professor of medicine, associate dean for special projects, and director of the dialysis program, Division of Nephrology, David Geffen School of Medicine, University of California, Los Angeles; SALLY WADE, MPH, is a consultant, Wade Outcomes Research and Consulting, Salt Lake City, Utah; KEVIN KNIGHT, MD, MPH, is a consultants, Santa Monica, California; L. TIM GOODNOUGH, MD, is a professor of pathology, Department of Pathology and Medicine, Stanford University School of Medicine, Stanford, California; ROBERT W. DUBOIS, MD, PhD, is senior vice president, Cerner Health Insights, Beverly Hills, California. AUTHOR CORRESPONDENCE: Allen R. Nissenson, MD, FACP, Director, Dialysis Program, Division of Nephrology, David Geffen School of Medicine, UCLA, 200 Medical Plaza, Suite 565-59, Los Angeles, CA 90095-6945. Tel: (310) 825-9464; Fax: (310) 206-2985; E-mail: [email protected] Copyright© 2005, Academy of Managed Care Pharmacy. All rights reserved. www.amcp.org A nemia is an important public health concern. It occurs commonly and is characterized by reduced concentrations of hemoglobin due to a variety of underlying causes.1,2 Estimates of anemia prevalence vary considerably. The National Center for Health Statistics conservatively estimates that approximately 3.4 million individuals in the United States are anemic.3 The National Health and Nutrition Examination Survey (NHANES) provides an assessment of anemia prevalence based on laboratory testing. Results from NHANES III show anemia to be most prevalent in children through age 16, women aged 17 to 49 years, and the elderly (aged 75 years and older), especially elderly men.4 Anemia prevalence is higher among individuals with certain chronic conditions, including chronic kidney disease (CKD), human immunodeficiency virus (HIV), rheumatoid arthritis (RA), inflammatory bowel disease (IBD), congestive heart failure (CHF), and cancer.2,5 In previous studies, anemia with chronic disease has been identified in 36% of patients with CKD6 and 27% of patients with RA.7 The increasing amount of information on the prevalence of anemia among individuals with such diseases supports the view that it is a condition of growing concern.7-11 Anemia has been shown to be associated with increased mortality and morbidity as well as with decreased physical functioning and quality of life.2,12 Although anemia is often associated with disease progression or increased disease severity, evidence of its independent effect on these key outcomes is still accumulating. Despite information on anemia prevalence and associated outcomes, little is known about the impact of anemia on health resource utilization and costs. Although the cost of anemia care has been examined using Medicare data for CKD,13 cancer,14,15 and heart failure,16 the literature provides no data on anemiarelated costs in other populations. Consequently, this study was undertaken to estimate the health care costs and treatment patterns of patients with anemia in a privately insured population. ■■ Methods This study is based on retrospective administrative claims data from commercially insured and Medicare plans represented in the Medstat MarketScan database, containing the combined administrative claims for more than 2 million health plan members for employers from across the country. The study population was selected only from those enrolled in plans with complete capture of facility-based and professional services, as well as outpatient prescription medications. The combined Medicare and commercial populations from which the study sample was Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 565 Economic Burden of Anemia in an Insured Population TABLE 1 Diagnosis Codes (ICD-9-CM) Used to Identify 6 Study Conditions Chronic kidney disease Malignant hypertensive renal disease (without renal failure) 403.00 Malignant hypertensive renal disease (with renal failure) 403.01 Benign hypertensive renal disease (without renal failure) 403.10 Benign hypertensive renal disease (with renal failure) 403.11 Unspecified hypertensive renal disease (without renal failure) 403.90 Unspecified hypertensive renal disease (with renal failure) 403.91 Malignant hypertensive heart and renal disease (with renal failure) 404.02 Malignant hypertensive heart and renal disease (with heart and renal failure) 404.03 Benign hypertensive heart and renal disease (with renal failure) 404.12 Benign hypertensive heart and renal disease (with heart and renal failure) 404.13 Unspecified hypertensive heart and renal disease (with renal failure 404.92 Unspecified hypertensive heart and renal disease (with heart and renal failure) 404.93 Nephrotic syndrome 581.0-581.9 Chronic glomerulonephritis 582.0-582.9 Nephritis (NOS as acute or chronic) 583.0-583.9 Chronic renal failure 585 Renal failure, unspecified 586 Renal sclerosis, unspecified 587 Chronic pyelonephritis (without lesion of renal medullary necrosis) 590.00 Chronic pyelonephritis (with lesion of renal medullary necrosis) 590.01 HIV infection HIV infection 042 Rheumatoid arthritis Rheumatoid arthritis Felty's syndrome Other rheumatoid arthritis with visceral or systematic involvement Rheumatoid lung Other 714.0 714.1 714.2 714.81 714.89 Inflammatory bowel disease Small intestine Large intestine Small intestine with large intestine Unspecified site Ulcerative (chronic) enterocolitis Ulcerative (chronic) ileocolitis Ulcerative (chronic) proctitis Ulcerative (chronic) proctosigmoiditis Left-sided ulcerative (chronic) colitis Universal ulcerative (chronic) colitis Other ulcerative colitis Ulcerative colitis, unspecified Congestive heart failure Congestive heart failure Malignant hypertensive heart disease (with congestive heart failure) Benign hypertensive heart disease (with congestive heart failure) Unspecified hypertensive heart disease (with congestive heart failure) Malignant hypertensive heart and renal disease (with congestive heart failure) Malignant hypertensive heart and renal disease (with congestive heart failure and renal failure) Benign hypertensive heart and renal disease (with congestive heart failure) Benign hypertensive heart and renal disease (with congestive heart failure and renal failure) Unspecified hypertensive heart and renal disease (with congestive heart failure) Unspecified hypertensive heart and renal disease (with congestive heart failure and renal failure) 555.0 555.1 555.2 555.9 556.0 556.1 556.2 556.3 556.5 556.6 556.8 556.9 428.0 402.01 402.11 402.91 404.01 404.03 404.11 404.13 404.91 404.93 Cancer (primary only) Cancer diagnoses in the range of 140.0-199.x, excluding 173.0-173.9. For cancer patients, we required 2 outpatient diagnoses on separate service dates within 6 months or 1 diagnosis on an inpatient admission record. ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification; NOS = not otherwise specified. drawn resided primarily in the southern, northern central, and northeastern regions of the United States, with a smaller representation from the western region. Analyses were conducted to examine the prevalence of anemia and related utilization and health plan costs in an adult population. This paper presents results for the entire study population and separately for groups with specific conditions that are often associated with an increased occurrence of anemia or in which anemia presents particular clinical challenges. For condition-specific subgroup results, patients were identified based on the presence of diagnosis codes from the International Classification of Diseases, 9th Revision, Clinical Modification, (ICD-9-CM) for 6 conditions: CKD, HIV, RA, IBD, CHF, and solid-tumor cancer (Table 1). Patients who had multiple diagnoses during the study period were included in all condition-specific groups for which they qualified. Since laboratory values such as hemoglobin levels are not 566 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 generally available in medical claims data, anemia was identified by the presence of at least 1 diagnosis code for anemia (occurring in any position on a claim), 1 procedure code (Current Procedural Terminology, 4th Edition [CPT-4], ICD-9-CM, or Health Care Financing Administration Common Procedure Coding System [HCPCS]), or 1 drug code indicative of anemia treatment (e.g., blood transfusions, injections with recombinant erythropoietin. Diagnoses used for patient selection include iron deficiency anemia, pernicious anemia, anemia of chronic disease, nutritional anemia, other specified aplastic anemias, and other unspecified anemia. Codes for acute anemias were not included, and blood transfusion was only considered in the absence of a diagnosis of acute anemia (ICD-9-CM 285.1) (Table 2). The study was divided into 2 components. The first component assessed anemia prevalence in the year 2000. The denominator included all adult health plan members who had continuous www.amcp.org Economic Burden of Anemia in an Insured Population Diagnosis Codes (ICD-9-CM) and Procedure Codes (CPT-4, ICD-9-CM, HCPCS) Used to Identify Anemia Iron deficiency anemia 280.x: Iron deficiency anemia Anemia in chronic illness 285.2x: Anemia in chronic illness Pernicious anemia 281.0 Pernicious anemia Other anemia The “other anemia” category includes patients with evidence of anemia diagnoses other than those listed above, plus patients with evidence of anemia treatment but no corresponding diagnosis. 281.1-281.9: Other nutritional anemias 285.9: Anemia, unspecified 284.8: Other specified aplastic anemias Transfusion V58.2, 99.0x (ICD-9-CM); 36430, 36440, 96400-96549 (CPT-4); P9010, P9011, P9012, P9013, P9016, P9017, P9018, P9019, P9020, P9021, P9022, P9023 (HCPCS) Note: Only patients with the above transfusion codes but no acute anemia (DX = 285.1) in both the 12-month period preceding the transfusion and during the follow-up period were included in the anemic population. Epoetin alfa injection HCPCS: Q9920, Q9921, Q9922, Q9923, Q9924, Q9925, Q9926, Q9927, Q9928, Q9929, Q9930, Q9931, Q9932, Q9933, Q9934, Q9935, Q9936, Q9937, Q9938, Q9939, Q9940, Q0136 NDC: 55513014401, 55513047810, 55513047801, 55513028301, 55513026701, 55513014810, 55513082301, 55513014410, 55513028310, 55513012610, 55513012601, 05551326710, 05551314810, 05551314410, 05551312610, 55513014801, 55513026710, 55513082310, 59676031200, 00062740103, 00062740201, 00062740501, 00403489718, 59676740104, 59676740000, 59676034001, 59676032001, 54868252300, 59676031201, 59676031002, 59676031001, 59676030402, 59676030401, 59676030302, 00062740003, 59676030301, 59676740503, 59676030202, 59676030201, 54868252301 CPT-4 = Current Procedural Terminology, 4th Edition; DX = Diagnosis; HCPCS = Health Care Financing Administration Common Procedure Coding System; ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification; NDC = National Drug Code. medical and drug benefits coverage during 2000, and the numerator included all members with evidence of a specified anemia diagnosis or treatment. The second component used administrative claims for dates of service between January 1, 1998, and June 30, 2001, to compare the health care cost and utilization patterns between anemic and nonanemic patients within the 6 study conditions. We captured the sequence of anemia-related services in the year following the initial diagnosis. For the second part of the study, we selected only those patients who were “newly diagnosed” with anemia, defined as those with at least 1 year of continuous medical and drug benefits coverage prior to their anemia index date (date of first anemia diagnosis or procedure in the study period [Table 2]) and no evidence of anemia diagnoses or treatment during this 1-year “history” period. For comparison, we selected patients who met the same health plan enrollment requirements as the anemic patients but had no evidence of anemia. We identified a comparison group for the overall anemic population and also constructed 6 condition-specific comparison groups using the www.amcp.org FIGURE 1 Year 2000 Anemia Prevalence Overall and by Condition 40 35 33.5% 30 % of Patients TABLE 2 25 21.4% 20 17.8% 13.2% 15 11.9% 10 10.2% 3.5% 5 0 CKD Cancer CHF IBD RA HIV Overall Condition Overall n=81,423 (out of overall Medstat MarketScan research database population= 2,296,832); CKD=4,834; cancer=14,023; CHF=7,234; IBD=1,349; RA=1,885; HIV=151. Percentages for each condition are percentages of the anemic population. N values for anemia prevalence by condition and age: CKD: age 18-49 = 378; age 50-64 = 771; age 65+ = 1,160 Cancer: age 18-49 = 1,201; age 50-64 = 2,784; age 65+ = 2,768 CHF: age 18-49 = 216; age 50-64 = 804; age 65+ = 2,726 IBD: age 18-49 = 296; age 50-64 = 283; age 65+ = 213 RA: age 18-49 = 326; age 50-64 = 550; age 65+ = 572 HIV: age 18-49 = 26; age 50-64 = 7; age 65+ = 2 CKD = chronic kidney disease; CHF = congestive heart failure; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis. same diagnostic criteria that we used for the anemic patients. Since the objective of this study component was to characterize anemia care in the first year after diagnosis, follow-up data were examined for a maximum of 12 months after the anemia index date. Individual follow-up periods were determined by the amount of time that each patient had continuous benefits coverage following the anemia index date. To avoid skewing the study toward a sicker population, we did not require that patients remain in the health plan for the full 12 months of potential follow-up; this variable follow-up period was taken into consideration in the analyses. For the cost model, we developed a variable to adjust for disease severity for each of the 6 conditions. Severity adjustment was based on specific ICD-9 codes, HCPCS codes for durable medical equipment, or pharmacy codes. Patients were separated into mild, moderate, and severe categories based on specific ICD-9 codes, HCPCS codes for durable medical equipment, or pharmacy codes. Cancer patients who were actively receiving chemotherapy were categorized as part of the moderate severity category, whereas those with evidence of metastasis were categorized as severe. CKD patients with 1 CKD hospitalization during follow-up were categorized as moderate while those with either more than 1 CKD hospitalization or with a kidney transplant were categorized as severe. Use of biologic therapies Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 567 Economic Burden of Anemia in an Insured Population FIGURE 2 Year 2000 Anemia Prevalence in Females by Condition and Age 40 ■ Age 18-49 ■ Age 50-64 ■ Age 65+ 35 % of Patients 30 25 20 15 10 5 0 CKD Cancer CHF IBD Condition RA HIV N values for anemia prevalence by condition and age: CKD: age 18-49 = 378; age 50-64 = 771; age 65+ = 1,160 Cancer: age 18-49 = 1,201; age 50-64 = 2,784; age 65+ = 2,768 CHF: age 18-49 = 216; age 50-64 = 804; age 65+ = 2,726 IBD: age 18-49 = 296; age 50-64 = 283; age 65+ = 213 RA: age 18-49 = 326; age 50-64 = 550; age 65+ = 572 HIV: age 18-49 = 26; age 50-64 = 7; age 65+ = 2 CHF = congestive heart failure; CKD = chronic kidney disease; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis. FIGURE 3 Year 2000 Anemia Prevalence in Males by Condition and Age 40 ■ Age 18-49 ■ Age 50-64 ■ Age 65+ 35 % of Patients 30 25 20 15 10 5 0 CKD Cancer CHF IBD Condition RA HIV N values for anemia prevalence by condition and age: CKD: age 18-49 = 280; age 50-64 = 763; age 65+ = 1,482 Cancer: age 18-49 = 298; age 50-64 = 1,774; age 65+ = 5,198 CHF: age 18-49 = 105; age 50-64 = 708; age 65+ = 2,675 IBD: age 18-49 = 152; age 50-64 = 211; age 65+ = 194 RA: age 18-49 = 55; age 50-64 = 167; age 65+ = 215 HIV: age 18-49 = 64; age 50-64 = 43; age 65+ = 9 CHF = congestive heart failure; CKD = chronic kidney disease; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis. 568 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 and selected nonbiologic therapies indicated moderate RA while joint surgery indicated severe RA. For IBD, more advanced medications, surgery, and multiple hospitalizations were indicators of severe disease. For CHF, the severity increased as the number of concomitant CHF medications and hospitalizations increased. The Charlson Comorbidity Index (CCI) was used to quantify burden of illness in the study population based on claims incurred during the 6 months before each patient’s index date.17 In order to assess treatment patterns, we identified specific procedures and medications that are commonly used in the management of anemia. These included blood transfusions, erythropoietin injections, B12 injections, iron injections, and use of testosterone, nandrolone, folate, or folic acid. It is important to note that the study database captured information only on outpatient prescription medications and did not include utilization of any inpatient or over-the-counter medications. Anemia treatment regimens were assessed for up to 1 year following patients’ anemia index dates; follow-up periods ended either at the conclusion of the study period or when patients left the health plan. All relevant therapies provided on or after the anemia index date were counted. The costs evaluated in this study were the payments made by the health plan, after subtraction of member cost-share, as reported on the final adjudicated version of each claim. No adjustments were made to standardize costs across the study period. Unfortunately, the structure and level of detail of administrative claims precluded us from simply summing up payments in order to determine the cost of anemia. As is common in cost studies such as this one, we created an algorithm to estimate the direct health plan payments attributable to anemia.14-16 For inpatient and outpatient (nonpharmacy) services, claims with either a primary or secondary diagnosis resulted in the attribution of a portion of the costs to anemia. In general, if anemia was listed as the primary diagnosis, 50% of the costs on the claim were attributed to anemia. If anemia was listed only as a secondary diagnosis, 25% of the costs were attributed to anemia. Allocations for individual claims ranged from 0% to 100%, depending on whether the anemia diagnosis was primary or secondary, how many additional diagnoses were on the claims, and whether anemia-specific services (e.g., erythropoietin injection) appeared on the claim. In addition, costs for anemia-specific procedure codes were attributed to anemia even when anemia was not listed explicitly as a diagnosis (e.g., transfusion). For outpatient pharmacy claims, all erythropoietin costs were attributed to anemia. Pairwise comparisons were performed for each variable according to the nature of the data involved: continuous variables were compared using t tests or nonparametric equivalents, and categorical variables were compared using chi-square tests. A multivariate analysis was also conducted to estimate cost www.amcp.org Economic Burden of Anemia in an Insured Population TABLE 3 Number of Patients and Average Follow-Up Months by Condition and Anemia Status Based on Claims Incurred January 1, 1998, to June 30, 2001 Newly Diagnosed Anemic Patients Overall population N (Average Follow-up Months) Average Age 118,332 (8.9) 56.9 7,545 (9.0) 65.8 Nonanemic Comparison Patients % Female N (Average Follow-up Months) Average Age % Female 66.3 35,948 (9.3) 61.6 53.0 49.3 5,814 (9.3) 62.7 44.4 Condition-Specific Populations Chronic kidney disease Human immunodeficiency virus 354 (8.7) 45.9 31.9 232 (9.1) 44.5 29.7 3,852 (9.1) 61.2 76.1 3,303 (9.2) 59.8 71.6 Inflammatory bowel disease 2,538 (9.3) 55.9 61.5 2,139 (9.3) 53.8 55.1 Congestive heart failure 14,985 (8,7) 72.5 53.5 11,886 (9.2) 71.3 49.5 Cancer 22,030 (8.7) 65.7 51.6 17,542 (9.3) 65.3 53.0 Rheumatoid arthritis differences between anemic and nonanemic patients, adjusting for factors likely to influence health care utilization and expenditures. An exponential model was fit using a generalized linear modeling technique, with patient age and gender, coverage type (e.g., preferred provider organization, indemnity), predisposing condition (i.e., the 6 study conditions), and disease severity for each predisposing condition as covariates and a binary indicator variable for presence or absence of anemia. Due to the skewed nature of distributions of payment data, a gamma variance function was chosen using the Park test, and bootstrap standard errors were estimated. All analyses were conducted using SAS software version 8.02 (Cary, NC) and STATA version 7.0 (College Station, TX). ■■ Results Based on data for the 2,296,832 adult health plan members with continuous benefits coverage during 2000, the overall anemia prevalence was 3.5% (81,423) (Figure 1). Although statistical comparisons cannot be made because the condition groups were not mutually exclusive, it is clear that the prevalence of anemia varied significantly by condition, with CKD defining the upper end. Within each of the 6 study conditions, the relationship between anemia and age among females was not consistent. However, among males, anemia prevalence increased with age (Figures 2 and 3). Overall, 118,332 anemic patients and a random sample of 35,948 nonanemic patients were identified for inclusion in the cost and utilization component of the study (Table 3). (Case-matching was determined to be unnecessary in order to provide reasonably precise adjusted measures of association since the size of the study population was large). The number of patients in the condition-specific subgroups ranged from 354 anemic patients and 232 nonanemic patients in the HIV subgroup to 22,030 anemic patients and 17,542 nonanemic www.amcp.org patients in the cancer subgroup (Table 3). Females made up the majority of both the overall anemic and control populations (66% and 53%, respectively), and the proportion of females in the anemic population was statistically higher (P<0.001). With the exception of the HIV population, females were more common in the anemic populations for all study conditions (P < 0.01). Overall, nonanemic patients were nearly 5 years older on average compared with the anemic patients (61.6 years vs. 56.9 years, P < 0.001). The opposite was true in the condition-specific populations: anemic patients were older, on average, than nonanemic patients in all of the condition-specific populations except for HIV (P < 0.002). With respect to the CCI, the overall population of anemic patients did not differ statistically from nonanemic comparison subjects (P = 0.22). However, in each of the 6 condition-specific populations, anemic patients had a statistically higher burden of illness as evidenced by higher CCI scores (P < 0.001). The difference in CCI scores was greatest in patients with CKD (1.6 for anemic vs. 0.98 for nonanemic) and those with cancer (1.6 for anemic and 0.56 for nonanemic). Anemia Management The average follow-up period for patients in this study was approximately 9 months. Overall Population In the overall population, the majority of anemic patients (86.5%) did not receive any of the therapies evaluated in this study (Figure 4). Among therapies evaluated, transfusion was the most commonly used, with nearly 1 out of every 10 anemic patients (9.3%) receiving at least 1 blood transfusion during the follow-up period. These patients averaged 1.1 transfusions per month. For most transfused patients, transfusion was the only therapy used; approximately 1 in 5 also received erythropoietin, which was nearly always given after the transfusion. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 569 Economic Burden of Anemia in an Insured Population FIGURE 4 lowest average number of transfusions per patient per month (0.8). Therapy involving both transfusion and erythropoietin was most common among anemic HIV patients (27.0% of transfused patients) and least common among anemic RA patients (15.5% of transfused patients). Nearly 1 in 5 transfused CKD patients (19.1%) and cancer patients (19.2%) also received erythropoietin. Among condition-specific anemic populations, erythropoietin was used most commonly by HIV patients (4.5%) and least commonly by RA patients (2.8%). The majority of erythropoietin use in these populations was in conjunction with (usually following) blood transfusion. Anemia Management Strategies by Condition 16 14 % of Patients 12 10 8 6 4 2 0 All CKD Treatment Cancer All CHF Condition IBD CKD Cancer CHF RA HIV IBD RA HIV ■■ Any 118,322 7,545 22,030 14,985 2,539 3,852 354 ■■ Transfusion 11,017 738 2,130 1,528 257 349 37 ■■ EPO 3,743 271 739 527 86 109 16 ■■ B12 injections 3,146 213 599 438 61 94 6 ■■ Iron injections 238 11 40 22 1 6 2 CHF = congestive heart failure; CKD = chronic kidney disease; EPO = erythropoietin; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis. TABLE 4 Average Annualized Utilization of Key Services per Patient Based on Claims Incurred January 1, 1998, to June 30, 2001 Total Population Service Anemic N = 118,332 (Average Follow-up Months = 8.9) Nonanemic N = 35,948 (Average Follow-up Months = 9.3) P Value (2-Tailed Student’s t Test) Outpatient visits 6.1 5.3 <0.001 Emergency room visits 0.2 0.1 <0.001 Inpatient admissions 0.4 0.3 <0.001 Hospital days 3.1 2.0 <0.001 Laboratory tests 5.3 2.6 <0.001 Condition-Specific Populations As in the overall population, the majority of patients (85.2%86.9%) in each of the 6 condition-specific populations had no documented anemia treatment. In these populations, the pattern of anemia treatment was similar to that in the overall population. Anemic HIV patients had the highest transfusion use (10.5% with at least 1 transfusion), but they also had the 570 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 Utilization and Costs Utilization of selected key services was significantly higher for anemic patients (P < 0.001 in all cases), as presented in Table 4. Similarly, per-patient payments for health care services (Tables 5 and 6) were higher for anemic patients than for nonanemic patients, both overall and within each of the 6 study conditions (P < 0.001). With the exception of outpatient pharmacy-based prescription drugs, anemic patients exhibited higher costs than did nonanemic patients for all types of care, including inpatient, outpatient, emergency room, and outpatient laboratory (P < 0.001). With the exception of outpatient facility care for HIV patients, this pattern of higher costs (including, in this case, higher outpatient prescription costs) among anemic patients persisted in the condition-specific populations (P < 0.03). Among anemic patients, average total annualized costs were $14,535 per patient. Outpatient care, including physician office visits, accounted for more than half ($7,927, 54.5%) of the average total costs. Inpatient care accounted for nearly one third ($4,775, 33%) of the average total costs. Payments for pharmacybased outpatient prescriptions averaged $1,833. In the nonanemic population, average total annualized costs were $9,450 per patient. Outpatient care accounted for 45% ($4,262) of the total average costs, while inpatient care accounted for 36% ($3,375). Outpatient pharmacy payments averaged $1,813 (19%). Cost differences between anemic and nonanemic patients were statistically significant (P <0.001) for all types of care except outpatient pharmacy (P = 0.24). Table 7 presents adjusted and unadjusted differences in average annualized per-patient costs. Cost differences persisted after adjusting for differences in patient gender and age, coverage type (e.g., preferred provider organization, indemnity), predisposing condition (i.e., the 6 study conditions), and disease severity for each predisposing condition. The average annualized total cost per anemic patient was more than twice the average for nonanemic patients. Both outpatient and inpatient costs were more than twice as high for anemic patients as for nonanemic patients. Services that we could attribute to anemia, based on our algorithm (i.e., claims for anemia treatment or claims containing a www.amcp.org Economic Burden of Anemia in an Insured Population diagnosis of anemia), accounted for only 5% to 11% of the cost differential between anemic and nonanemic patients. Most of the difference was accounted for by services without an anemia diagnosis code or another unambiguous relationship to anemia. Average annualized anemia-attributed payments, using our algorithm, were $563 per patient. Outpatient care accounted for the largest share of anemia-attributed expenditures (which included a proportion of costs for services with either an anemia diagnosis or anemia-specific procedure such as erythropoietin administration), averaging $412 per patient annually (73% of total anemia-attributed costs). The costs of inpatient care attributable to anemia, based on our algorithm, were slightly lower than those for professionally administered outpatient medications, $90 and $99 per patient annually. Anemia-attributed outpatient prescription costs (i.e., costs for anemia-related medications) averaged $61 per patient annually. ■■ Discussion For this study, we examined anemia prevalence, current treatment patterns, and associated costs of care in a privately insured population in order to determine the impact of anemia on the use of health care resources. In general, anemia occurred with noticeable frequency even in a relatively healthy privately insured population and resulted in higher health care utilization and costs. Overall, 3.5% of the study population was anemic at some point during the study year. It is somewhat challenging to compare this estimated anemia prevalence with estimates from previous studies, given that no standard definition of anemia is currently used and that reported and actual prevalence vary widely depending on the nature of the population studied. It is likely that the prevalence estimates of anemia in this study underestimate the true prevalence since, in order to be considered anemic, a patient was required to have a diagnosis of anemia recorded on a claim during the study period or to have received one of the specified anemia therapies. These results underscore the fact that anemia is common enough to merit attention, even outside the context of those conditions with which it has historically been associated. Nearly 4% of the overall study population had anemia that was serious enough to be recorded as a diagnosis on a medical claim, to receive an anemia-related prescription medication, or to require an anemia-related procedure. In general, these results highlight the importance of understanding the demographic and clinical risk factors that increase the likelihood that a particular individual will be anemic. From both public health and provider perspectives, such profiles of “at-risk” populations are critical for improving anemia screening, detection, and treatment. This study suggests that anemia merits particular attention in routine clinical care for women and the elderly. Anemic patients used significantly more health care services www.amcp.org TABLE 5 Average Annualized Health Care Payments by Type of Care for Anemic and Nonanemic Patients Based on Claims Incurred January 1, 1998, to June 30, 2001 Type of Care Payments ($) Payments ($) per per Nonanemic Anemic Patient Patient N = 118,332 N = 35,948 P Value (Average (Average (2-Tailed Follow-up Follow-up Student’s Months = 8.9) Months = 9.3) t Test) Total 14,535 9,451 <0.001 4,775 3,375 <0.001 137 $101 <0.001 Outpatient care 7,927 4,262 <0.001 Pharmacy-based outpatient medications 1,833 1,813 0.238 Professionally administered outpatient medications 157 0.13 <0.001 Inpatient care Emergency room and had higher costs ($14,535 vs. $9,451, P <0.001), even compared with patients with the same underlying condition who were not anemic. Since it is often assumed that anemia may simply be a marker for the severity of a key underlying disease (e.g., RA) and that disease severity would therefore be the primary driver for any observed cost differences, our multivariate analysis was designed to adjust for the presence of key conditions and the severity of those conditions as well as for other factors that could influence costs (i.e., patient age, gender, and coverage type). Our results indicate that costs of anemic patients were more than twice those of nonanemic patients even after adjusting for these other potential confounders. The majority of the anemic patients (85%) did not receive any of the therapies assessed. This is quite striking since these patients were primarily identified through the presence of explicit anemia diagnoses on the medical claims and, therefore, those with mild anemia may be underrepresented. One possible explanation is the use of oral iron, which was not captured in the study database because it is an over-the-counter medication. Nonetheless, the finding that only 15% of patients received any apparent treatment raises serious concern that anemia is inadequately managed. It is possible that physicians in general do not attribute much clinical importance to anemia, especially in patients who do not have medical conditions that may be exacerbated by anemia, and/or for whom anemia management is part of the standard of care. If that is the case, then efforts to increase awareness of anemia’s risk factors and consequences are needed, along with practical clinical treatment guidelines. It may not be surprising that anemia in the overall health plan population appears to be either undertreated or minimally treated. It is surprising, however, that despite explicit guidelines Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 571 Economic Burden of Anemia in an Insured Population TABLE 6 Average Annualized Health Care Payments for Anemic and Nonanemic Patients Anemic Patients Nonanemic Patients N Average Payment ($) per Patient 118,332 14,535 35,948 9,451 5,084* 563 (11%) 7,545 41,292 5,814 12,535 28,757* 2,324 (8%) 354 37,424 232 13,579 23,845* 1,300 (5%) Rheumatoid arthritis 3,852 17,186 3,303 7,777 9,409* 502 (5%) Inflammatory bowel disease 2,538 19,113 2,139 7,678 11,435* 676 (6%) Congestive heart failure 14,985 29,703 11,886 12,459 17,244* 1,141 (7%) Cancer 22,030 34,009 17,542 9,034 24,975* 1,480 (6%) Condition All Chronic kidney disease HIV Average Payment ($) per Patient N Payment ($) Difference AnemiaAttributed Payment ($) per Patient (% Difference) * P < 0.001 based on 2-tailed Student’s t test. HIV = human immunodeficiency virus. TABLE 7 Unadjusted and Adjusted Differences* in Average Annualized Health Care Payments for Anemic and Nonanemic Patients Average Total Payments ($) per Patient† Average Inpatient Payments ($) per Patient Average Outpatient Payments ($) per Patient Unadjusted costs for anemic patients (N = 118,332) 14,535 4,775 7,927 Unadjusted costs for nonanemic patients (N = 35,948) 9,451 3,375 4,262 Adjusted costs for nonanemic patients 7,106 1,996 3,297 Unadjusted cost differences 5,084 1,400 3,665 Adjusted cost differences‡ 7,429 2,779 4,630 * Costs for the nonanemic patients were standardized using the covariate levels in the anemic population. Covariates included age, sex, coverage type (preferred provider organization), presence and severity of 6 study conditions, Charlson Comorbidity Index. †Total payments include inpatient, outpatient, and outpatient pharmacy costs. ‡ P values for differences were all <0.001 based on 2-tailed Student’s t test. emphasizing aggressive anemia management in CKD18 and cancer,19 such patients are not receiving adequate or appropriate care.19,20 These results also suggest that when anemia treatments are employed, their usage is remarkably similar across the 6 study conditions. The one exception is HIV, where the use of erythropoietin is more commonly observed than in the other conditions. This pattern suggests that the presence of underlying conditions may not play a significant role in clinical judgments about which anemia treatment is most appropriate 572 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 or even about how aggressively to manage anemia. Some of the observed treatment patterns also highlight specific quality-of-care considerations. Despite growing concern about the risks associated with transfusions and a wide array of initiatives to promote blood conservation, transfusions represented the predominant treatment among patients with newly diagnosed anemia in the study database. There is a clear gap between current practice relative to blood conservation and recommendations for transfusion alternatives. For example, current recommendations for use of blood products list iron, folate, B12, and erythropoietin therapy as specific therapies that should be administered instead of blood transfusions if the patient’s condition permits time for these agents to be utilized.21 Since the current study focused only on use of injectable drugs, further analysis is necessary to understand the full extent to which these recommended first-line pharmaceutical therapies are used. Limitations Administrative claims data are one of the richest sources of information on health care utilization and cost and have historically served as the foundation for many areas of health services research. Like any data source, claims data present limitations: the most important is that the level of detail available is limited to that required for claims adjudication and internal and external health plan reporting. This limitation may lead to underidentification of anemia in the study population since the anemia diagnosis codes and anemia-related procedures and pharmaceutical therapies used for patient selection are only proxy indicators of anemia. The lack of actual hemoglobin levels for the patients in the study population also precludes the assessment of anemia severity. While claims data do present a reasonable amount of clinical information, no simple standard methodology was available to stratify patients by severity for www.amcp.org Economic Burden of Anemia in an Insured Population the 6 study conditions. Finally, although the costs presented in this paper may differ from those experienced by another health plan due to differing fee schedules, the more critical finding is the statistically higher health care costs associated with anemia. Although the data collected are not sufficient to explain why health care costs are higher among anemic patients, we would like to suggest a few possible explanations. First, despite our best efforts to control for disease severity and comorbidity burden in this analysis, it is possible that anemic patients are simply sicker than the controls and therefore use more health care services. It is also possible that the presence of anemia may contribute to a higher rate of detection of comorbidities in these patients or that anemic patients with comorbidities may be more likely to have their anemia detected, resulting in an association that is not causal. These adjusted cost comparisons should also be considered in light of our examination of anemia-attributed costs (e.g., costs due to specific anemia visits, tests, and therapies). Although these anemia-attributed costs certainly contributed to overall health care costs in the study population, they represented only a small percentage of the total costs per patient. These results suggest that anemia may be responsible for excess costs in areas that cannot be captured by our algorithm or, alternatively, that the association is a marker for disease severity associated with increased costs (i.e., that the association is not causal). ■■ Conclusions The results of this study demonstrate that the elevated clinical burden that anemia imposes at the patient level in turn increases the resource burden at the health plan level. As anemia gains greater recognition as both an important clinical and public health issue, careful consideration should be given to determining the most cost-effective approaches to anemia screening in highrisk populations and efforts to improve anemia diagnosis and treatment. Given the challenges inherent in isolating the true costs of anemia, future research should examine the economic impact of more aggressive anemia treatment to determine if expected short-term cost increases incurred by earlier treatment would be offset by savings from fewer admissions, shorter lengths of stay, and less use of other expensive services in the treatment of anemia-related outcomes. Target research should also examine the extent to which patient care is consistent with current guidelines. DISCLOSURES Funding for this research was provided by the National Anemia Action Council, Inc. (NAAC); additional funding was also provided by the Richard Rosenthal Dialysis Fund. Corporate sponsors of NAAC include Amgen, Watson Pharmaceuticals, American Regent Laboratory, and Fibrogen. Author Allen R. Nissenson discloses that he has received research support from Amgen, OrthoBiotech, Roche, and Watson pharmaceutical companies and American Regent Laboratory and has been a speaker for ARL. L. Tim Goodnough is on the speaker’s bureau of Amgen, Novo Nordisk, Watson, American Regent Laboratory, and OrthoBiotech; Sally Wade and Kevin Knight provide consulting services to the health care industry; Robert W. Dubois is an employee Cerner Health Insights, which provides consulting services to the health care industry. The authors disclose no potential bias or conflict of interest relating to this article. Nissenson served as principal author of the study. Study concept and design were contributed by all authors. Analysis and interpretation of data were contributed by Wade and Knight. Drafting of the manuscript was primarily the work of Wade and Dubois, and its critical revision was the work of all authors. Statistical expertise was contributed by Knight. REFERENCES 1. Dallman PR, Yip R, Johnson C. Prevalence and causes of anemia in the United States, 1976 to 1980. Am J Clin Nutr. 1984;39:437-45. 2. Goodnough LT, Dubois RW, Nissenson AR. Anemia: not just an innocent bystander? Arch Intern Med. 2003;163:1400-04. 3. Adams PF, Hendershot GE, Marano MA. Current estimates from the National Health Interview Survey, 1996. Vital Health Stat 10. 1999:1-203. 4. Guralnik JM, Eisenstaedt RS, Ferrucci L, Klein HG, Woodman RC. Prevalence of anemia in persons 65 years and older in the United States: evidence for a high rate of unexplained anemia. Blood. 2004;104:2263-68. 5. Goodnough LT, Dubois RW, Nissenson AR. Anemia: not just an innocent bystander? [correction]. Arch Intern Med. 2003;163:1820. 6. Kausz AT, Steinberg EP, Nissenson AR, Pereira BJG. Prevalence and management of anemia among patients with chronic kidney disease in a health maintenance organization. Dis Manag Health Outcomes. 2002;10:505-13. 7. Baer AN, Dessypris EN, Krantz SB. The pathogenesis of anemia in rheumatoid arthritis: a clinical and laboratory analysis. Semin Arthritis Rheum. 1990; 19:209-23. 8. Knight K, Wade S, Balducci L. Prevalence and outcomes of anemia in cancer: a systematic review of the literature. Am J Med. 2004;116(suppl 7A):11S26S. 9. Belperio PS, Rhew DC. Prevalence and outcomes of anemia in individuals with human immunodeficiency virus: a systematic review of the literature. Am J Med. 2004;116(suppl 7A):27S-43S. 10. Wilson A, Reyes E, Ofman J. Prevalence and outcomes of anemia in inflammatory bowel disease: a systematic review of the literature. Am J Med. 2004;116(suppl 7A):44S-49S. 11. Wilson A, Yu HT, Goodnough LT, Nissenson AR. Prevalence and outcomes of anemia in rheumatoid arthritis: a systematic review of the literature. Am J Med. 2004;116(suppl 7A):50S-57S. 12. Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J Med. 2005; 352:1011-23. 13. Collins AJ, Li S, Ebben J, Ma JZ, Manning W. Hematocrit levels and associated Medicare expenditures. Am J Kidney Dis. 2000;36:282-93. ACKNOWLEDGMENTS The authors wish to thank Lisa Kaspin, PhD, Cerner Health Insights, Beverly Hills, CA, for her assistance in preparing this article and Onur Baser, PhD, and Xue Song, PhD, of Medstat, Ann Arbor, MI, for providing data and technical programming for this study. www.amcp.org 14. Berndt E, Crown W, Kallich J, et al. The impact of anaemia and its treatment on employee disability and medical costs. Pharmacoeconomics. 2005; 23:183-92. 15. Lyman GH, Berndt ER, Kallich JD, et al. The economic burden of anemia in cancer patients receiving chemotherapy. Value Health. 2005;8:149-56. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 573 Economic Burden of Anemia in an Insured Population 16. Nordyke RJ, Kim JJ, Goldberg GA, et al. Impact of anemia on hospitalization time, charges, and mortality in patients with heart failure. Value Health. 2004;7:464-71. 20. Nissenson AR, Collins AJ, Hurley J, et al. Opportunities for improving the care of patients with chronic renal insufficiency: current practice patterns. J Am Soc Nephrol. 2001;12:1713-20. 17. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613-19. 21. American Association of Blood Banks, America’s Blood Centers, and the American Red Cross. Circular of information for the use of human blood and blood components. Available at: http://www.aabb.org/All_About_Blood/ COI/coi0702.pdf. Accessed June 15, 2005. 18. NKF-DOQI clinical practice guidelines for the treatment of anemia of chronic renal failure. National Kidney Foundation-Dialysis Outcomes Quality Initiative. Am J Kidney Dis. 1997;30:S192-S240. 19. Groopman JE, Itri LM. Chemotherapy-induced anemia in adults: incidence and treatment. J Natl Cancer Inst. 1999;91:1616-34. 574 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org C O N T E M P O R A RY S U B J E C T Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative TROY K. TRYGSTAD, PharmD, MBA; DALE CHRISTENSEN, RPh, PhD; JENNIFER GARMISE, PharmD; ROBERT SULLIVAN, MD, MPH; and STEVEN E. WEGNER, MD, JD ABSTRACT OBJECTIVE: In response to burgeoning drug costs, North Carolina (NC) Medicaid encouraged pharmacists and prescribers to develop collaborative programs to reduce drug expenditures. One of these programs, the North Carolina Polypharmacy Initiative, was a focused drug therapy management intervention aimed at reducing polypharmacy in nursing homes. This intervention targeted patients with more than 18 prescription fills in 90 days, beginning in November 2002. These patients were believed to have a high likelihood of experiencing potential drug therapy problems (PDTPs). Consultant pharmacists were asked to utilize profiles displaying alerts generated from pharmacy claims to guide interventions in addition to usual-care drug regimen reviews. The pharmacists documented their reviews, recommendations, and resulting changes in drug therapy. Our objectives were to determine (1) the persistence of PDTP alerts following interventions by consultant pharmacists and (2) the impact of these interventions on patient drug costs from a payer perspective. METHODS: A before-after study with comparison group design was used. Medicaid prescription claims data were compared for the 90-day periods prior to the intervention (June-August 2002) and following the intervention (March-June 2003). The 90-day postintervention period allowed for 2 to 3 follow-up prescriptions and reduced the drop-out rate. The 5 categories of potential problem alerts included potentially inappropriate medications (Beers criteria), substitution opportunity for a lower-cost drug, 16 drugs or drug classes with specific quality improvement opportunities (Clinical Initiatives list), therapeutic duplication, and length of drug therapy evaluation. RESULTS: A total of 253 nursing homes, involving 110 consultant pharmacists and 6,344 patients, were in the intervention arm, with 5,160 patients (81.3%) remaining at the end of the follow-up period. At baseline, study-group patients used an average of 9.7 prescriptions per month, costing the NC Medicaid program $517 per patient per month (PPPM). There were 6,360 recommendations offered for 3,400 patients, or an average of 1.87 recommendations per patient. Physicians concurred with 59.8% (3,801 of 6,360) of all recommendations to change drug therapy, about half involving a switch to a lower-cost drug. Two of 5 alert categories had significant (P <0.01) reductions in alert persistence: -10.8% for the study group versus -0.7% for the comparison group for the Clinical Initiatives list and -29.7% for the study group versus -14.1% in the comparison group for the drug substitution opportunity. Median drug costs per patient in the study group decreased by $12.14 (-0.92%), from $1,329.46 to $1,317.32, and increased in the comparison group by $44.98 (3.35%), from $1,341.25 to $1,386.23, creating a relative cost reduction of $57.12 per patient in the 3-month follow-up period, or $19.04 PPPM. CONCLUSION: A supplemental program of medication reviews for nursing home patients targeted by high drug utilization resulted in a reduction in the persistence of PDTP alerts and was cost beneficial based solely on drug cost savings. This intervention may be a model for future medication therapy management services provided by prescription drug plans under Medicare Part D for patients in long-term-care settings and possibly ambulatory patients. KEYWORDS: Nursing homes, Pharmaceutical care, Medication therapy management, Drug use review, Polypharmacy, Drug regimen review J Manag Care Pharm. 2005;11(7):575-83 Note: An editorial on the subject of this article appears on pages 586-87 of this issue. www.amcp.org T he passage of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) and the ensuing rollout of the outpatient drug benefit in January 2006 have focused attention on ensuring elderly patient access and cost-effective prescribing and use of drugs. Those responsible for Part D program administration within the Centers for Medicare and Medicaid Services (CMS) and prescription drug program sponsors share the formidable task of managing both the cost and quality of drug regimens for more than 40 million Medicare beneficiaries. Medicare will become the largest single payer of drug benefits in the United States, with a projected $70 billion in expenditures in 2006.1 The elderly have more chronic illnesses and use more prescription drugs than any other age segment, increasing the likelihood of adverse drug events, many of which are avoidable.2-4 In an attempt to ameliorate the cost burden and ensure rationale and optimal drug use, Congress took the novel approach of requiring prescription drug plans (PDPs) and Medicare Advantage PDPs to offer a Medication Therapy Management Program (MTMP) as part of their drug benefit. Despite considerable variations in strategy and implementation, prior MTMP-like programs have demonstrated significant cost savings and reductions in drug therapy problems for other targeted patient populations.5-7 Authors TROY K. TRYGSTAD, PharmD, MBA, is a contract data analyst, AccessCare, Inc., Morrisville, North Carolina, and a PhD candidate (American Foundation for Pharmaceutical Education fellow), Division of Pharmaceutical Policy and Evaluative Sciences, School of Pharmacy, University of North Carolina, Chapel Hill; DALE CHRISTENSEN, RPh, PhD, is a professor, Division of Pharmaceutical Policy and Evaluative Sciences, School of Pharmacy, University of North Carolina, Chapel Hill; JENNIFER GARMISE, PharmD, is a pharmacy projects manager, AccessCare, Inc., Morrisville, North Carolina; ROBERT SULLIVAN, MD, MPH, is medical director, Community Care of North Carolina, and associate professor, Department of Community and Family Medicine, Department of Medicine, Center for the Study of Aging and Human Development, Duke University Medical Center, Durham, North Carolina; STEVEN E. WEGNER, MD, JD, is president and medical director, AccessCare, Inc., Morrisville, North Carolina, and an adjunct assistant professor, Department of Pediatrics, University of North Carolina, Chapel Hill. AUTHOR CORRESPONDENCE: Troy K. Trygstad, PharmD, MBA, Division of Pharmaceutical Policy and Evaluative Sciences, School of Pharmacy, C.B. 7360, University of North Carolina, Chapel Hill, NC 27599-7360. Tel: (919) 843-3775; Fax: (919) 966-8466; E-mail: [email protected] Copyright© 2005, Academy of Managed Care Pharmacy. All rights reserved. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 575 Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative FIGURE 1 Inclusion Criteria and Cohort Development 131 Nonparticipating Nursing Homes 253 Participating Nursing Homes >12,000 North Carolina Nursing Home Residents With at Least 18 Prescription Fills in 90 Days BeforePeriod Screen 1,204 Patients Discharged or Deceased 9,208 Patient Drug Profiles Sent Intervention Period 1,660 Profiles Not Returned 6,344 Profile Reviews Completed AfterPeriod Screen 2,202 Patients Remain in Comparison Group 5,160 Patients Remain Following Intervention Period 3,400 Patients With Recommendations 2,305 Patients With Accepted Recommendations Defining the nature and scope of MTMP services within Medicare Part D continues to be a dynamic and ongoing endeavor. A consortium of pharmacy trade and professional associations published a working definition in July 2004.8 This definition was expanded by the American Pharmacists Association and the National Association of Chain Drug Stores in April 2005.9 However, CMS’s final rules pertaining to MTMP services remain broadly defined, leaving the operational details to PDP sponsors.10 MMA was not the first federal legislation to require pharmacist involvement in the drug-use process. Beginning in the 1970s, federal regulations imposed a requirement that monthly drug regimen reviews (DRRs) be conducted in long-term-care facilities by consultant pharmacists.11 Subsequent Omnibus Reconciliation Act legislation (OBRA ’87) required that this review be accomplished in collaboration with the attending physician. These regulations contained explicit requirements for reviewing therapy for targeted drugs and drug classes deemed to be over- 576 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 used in long-term-care settings. While such reviews have resulted in improved care since first mandated,12 there is room for improvement, and a more holistic approach based upon the optimization of both the type and use of all drugs taken by Medicare Part D recipients seems prudent.13 Medicaid recipients are also subject to drug reviews through OBRA ’90 regulations that require ongoing statewide drug utilization review (DUR) activities. These programs typically focus on drug use by ambulatory Medicaid recipients. The legislation compelled states to establish committees and systems to review patterns of drug use believed to be problematic but did not go as far as MMA to allow for explicit compensation of pharmacists as providers of care. MMA legislation effectively shifts the burden of drug costs incurred by elderly Medicaid recipients from the state-federal program to the federal government. Prior to the passage of MMA, states were burdened with Medicaid drug expenditures that were ballooning at unsustainable rates despite the federal sharing of Medicaid costs. North Carolina (NC) Medicaid spent more than $1.2 billion on drugs in 2003, with the elderly accounting for 11% of recipients but 32% of all prescription drug costs.14,15 In response to these trends, NC Medicaid introduced a program that combined the state-level, top-down administration characteristic of DUR activities with patient-level, pharmacist-driven activities typical of DRRs. This program was titled the North Carolina Polypharmacy (NCPP) Initiative. Following a successful pilot study, the NCPP Initiative was launched in 253 nursing homes in North Carolina with emphasis on elderly Medicaid recipients. In addition to mandated DRRs, the initiative provided a targeted drug therapy management consultation provided by a pharmacist with the treating physician. In these targeted drug therapy management consultations, pharmacists were to (1) review a drug profile generated from Medicaid pharmacy claims with potential drug therapy problem (PDTP) alerts and medical records of Medicaid patients in nursing homes, (2) determine if a drug therapy problem existed, (3) recommend a change if needed, and (4) perform a follow-up to determine if the change was implemented. The NCPP Initiative was organized as a collaborative activity that incorporated a physician primary care practice network (AccessCare of North Carolina), a pharmacy consultant coalition, and a network of nursing home medical directors. The nature of the NCPP Initiative and its organization was described in an earlier paper that reported the type and frequency of pharmacist interventions and estimated the cost impact of drug therapy changes by type of PDTP.16 Intervention documents submitted by pharmacists were used as a single data source. For the 6,344 patients with reviews, pharmacists responded to approximately 20,000 drugs with alerts by making 6,520 recommendations, resulting in changes in drug therapy 58% of the time.16 These changes were projected to save NC Medicaid $30.33 per patient per month (PPPM).16 www.amcp.org Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative FIGURE 2 Patient Profile Generated From Medicaid Pharmacy Claims— With Potential Drug Therapy Problem Alerts In the present article, we reconcile the projected drug cost impact of pharmacist intervention activities with actual Medicaid claims data spanning a 6-month period. We describe the nature of PDTP alerts, drugs involved, recommendations, and actions taken after physician consultation. We also assess changes in drug therapy from a qualitative and economic perspective using a before-after study design with a comparison group. Our working hypothesis was that a systematic program of pharmacist-directed DUR that supplements requisite OBRA ’87 DRRs in nursing homes would produce drug therapy changes that maintain or improve the quality of care while decreasing drug costs. The specific objectives of the current study were to determine (1) the persistence of PDTP alerts following interventions by consultant pharmacists and (2) the impact of these interventions on patient drug costs from a payer perspective. This study received approval from the Institutional Review Board at the University of North Carolina at Chapel Hill. www.amcp.org ■■ Methods Setting and Participants Phase 1 of the NCPP Initiative was conducted by 110 pharmacists in 253 nursing homes, representing approximately 70% of all nursing homes in North Carolina (Figure 1). Participation in the intervention was solicited through the North Carolina Long Term Care Pharmacy Alliance, a representative group of pharmacists serving nursing homes throughout the state. Exempted were 13 homes that contracted with a single pharmacy provider and were involved in a separate, ongoing intervention project. All Medicaid residents of the participating facilities who had 18 or more prescription fills in the 90-day period prior to the start of the study were eligible for an on-site profile review by a consultant pharmacist. This time horizon was chosen to capture, on average, 3 monthly supplies of medications while limiting the dropout rate as much as possible. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 577 Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative Pharmacist Responsibilities Participating pharmacists were introduced to the project, toolkit, and documentation form during two 1-hour group meetings and one 1-hour conference call. Other professional interactions took place throughout the course of the project, including informational meetings with geriatric associations, nursing home medical directors, and network physicians, as well as the use of telephone follow-ups. The toolkit contained instructions for documenting interventions and explained the screening criteria used to select (flag) drugs for attention. Each consultant pharmacist was provided with drug profiles computer-generated from Medicaid pharmacy claims that displayed flags for patients and suggestions for modifications of drugs and classes of drugs. Pharmacists were asked to record both the result of the review (i.e., the intervention) and the result of the consultation with the prescribing physician (i.e., the outcome) on a documentation form (Figure 2). Recording the result of the intervention required awaiting the prescriber’s response to the recommendation. Pharmacists were required to conduct these assessments during their regularly scheduled visits to each home. Consultant pharmacists employed their usual methods of communicating with physicians (fax, phone, or written notation in the medical record) to make recommendations and to learn the outcome of the change recommendation. We categorized the drug therapy flags as (1) unnecessary drug therapy, (2) more cost-effective drug available, (3) wrong dose/delivery, (4) potential for adverse drug reaction, (5) needs additional therapy, and (6) other problem. We coded intervention results as (1) dose/delivery changed, (2) drug added, (3) drug changed (from one to another), (4) drug discontinued, (5) no change, and (6) other intervention. If an intervention resulted in a drug therapy change of any type, the new drug, dose, and quantity were noted. Drug, dose, and quantity were also reported for each new drug added for previously untreated indications. Pharmacists were compensated $12.50 for each comprehensive profile review for which results were clearly documented on the forms provided (i.e., the patient profile). This compensation amount was based on our estimate of the additional time required for these focused reviews above and beyond normal review activities and a customary rate of pay of $50 per hour. Pharmacists were compensated regardless of problem determination and/or the offering of a recommendation. Drug Profiles and PDTP Alert Criteria Patient drug profiles were generated from Medicaid claims data and contained, for each listed drug, a space for all alert categories, marked with the appropriate flag/alert if a PDTP was determined by matching claims data with drug lists generated from alert categories. The profiles were a compilation of all drugs for which a claim was paid in the 90 days prior to generation, regardless of the presence of an alert. The first alert criterion 578 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 was receipt of a drug widely considered to be inappropriate for use in the elderly (Beers drug list).17 In order to engender participation and maximize the quality of the PDTP alerts, program administrators also elicited input from local physicians and consultant pharmacists. Thus, the second criterion was receipt of a drug on the Community Care of North Carolina Prescription Advantage List (PAL), which encourages substitution of less expensive drugs within a therapeutic class. This voluntary preferred drug list was conceived and is maintained by a committee of practicing physicians in North Carolina specifically for NC Medicaid. There are 3 categories of PAL drug alerts. PAL-3 drugs are considered to incur “significant cost” to the Medicaid program (e.g., Nexium, Prilosec, Zestril, Prinivil, as of November 2002), while PAL-2 drugs offered “no clear cost advantage” (e.g., Prevacid, Aciphex, Accupril, Monopril, Lotensin, Altace, as of November 2002), and PAL-1 drugs offer “significant cost savings” to the Medicaid program (e.g., Protonix, lisinopril, enalapril, captopril, as of November 2002). The third criterion was the appearance of a drug on a “Clinical Initiatives” list. The Clinical Initiatives list was developed by consultant pharmacists participating in the NCPP Initiative and included 16 drugs and/or drug classes (e.g., COX-2 inhibitors, statin drugs, sleep aids, low-sedating antihistamines) that had the potential for quality improvement and cost savings. Program administrators offered 2 additional alerts: therapeutic duplication and a “consider length of therapy” alert that was derived from classes of drugs considered appropriate only for short-term use (e.g., antibiotics, injectable enoxaparin). Research Design We first evaluated pharmacist action and reporting by reconciling the response to alerts with downstream prescribing activity using the Medicaid dispensed prescription claims database. Using a before-after, study-comparison-group design, we compared prescription use during the 3 months before intervention (JuneAugust 2002) with a period of equal length at the end of Phase 1 (March-June 2003). Second, we assessed whether or not PDTP alerts were reduced during the follow-up period compared with usual-care controls (nonrandomized comparison group). Third, we describe the economic consequences of pharmacist activities in terms of changes in drug cost using pharmacy paid claims data. Study-group patients were Medicaid recipients residing in participating nursing homes who received a completed profile review by a consultant pharmacist. The comparison group consisted of patients in nursing homes not responding to the invitation for inclusion in Phase 1 of the intervention. Inclusion of patients in comparison-group homes was determined by criteria identical to study-group patients (i.e., more than 18 prescription fills in 90 days, Figure 1). Several of the nursing homes in the comparison group became participants in later phases of the project, but only after the 6-month study window in this www.amcp.org Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative TABLE 1 Baseline Characteristics of Continuously Enrolled Patients by Treatment Group Characteristic Study Group (n = 5,160) Study Group (With Recommendation*) (n = 3,400) Study Group† (With Acceptance) (n = 2,305) Comparison Group (n = 2,202) Sex, no. (%) Male Female 1,289 (24.98) 3,871 (75.02) 820 (24.12) 2,580 (75.88) 533 (23.99) 1,752 (76.01) 484 (21.98) 1,718 (78.02) Race, no. (%) White Other 3,533‡ (68.47) 1,627 (31.53) 2,325‡ (68.38) 1,075 (31.62) 1,588‡ (68.89) 717 (31.11) 1,667 (75.70) 535 (24.30) 77.57 ± 12.72 (80.0) 77.63 ± 12.42 (80.0) 77.67 ± 12.44 (80.0) 78.65 ± 12.46 (81.0) 29.04 ± 9.92 (27.0) 29.86 ± 10.27 (28.0) 30.19 ± 10.53 (28.0) 28.18 ± 10.74 (26.0) $1,549.89 ± 1,652.49 ($1,329.46) $586.91 ± 919.17 ($1,392.14) $1,610.02‡ ± 926.77 ($1,427.13) $1,543.67 ± 921.98 ($1,341.25) Age, years, mean ± SD (median) No. of prescription fills, 3 month period, mean ± SD (median) Amount of paid claim ($), 3-month period, mean ± SD (median) Note: Difference of proportions tests were used to determine differences in sex and race. T-testing was used to determine differences in age, number of prescription fills, and amount of paid claims. * Study group (with recommendation) = those patients having a recommendation resulting from pharmacist consultation, regardless of outcome. † Study group (with acceptance) = those patients having a recommendation and a change in therapy as a result of a recommendation provided by a pharmacist. ‡ Denotes significantly different from comparison group at P <0.01. analysis. For study-group patients, we linked prescription drug use elicited through claims data to pharmacist-reported interventions (or lack thereof) on patient profiles. We examined 2 study subgroups: (1) patients whose drug use received pharmacist recommendations and (2) patients for whom recommendations were accepted. Studies in the long-term-care arena are often burdened by a high attrition rate. Using a combination of claims data and pharmacist report, we estimated an annual nursing home resident attrition rate of 36% due to death or discharge in North Carolina. Since we were not able to verify dropout from prescription claims with certainty, only residents having claims in the last 35 days of the 90-day follow-up period were included in both the study and comparison groups. Statistical testing was performed using SAS statistical software, version 8.2 (1999-2001, SAS Institute Inc., Cary, NC). We used nonparametric statistical testing to account for possible skewness in the data. ■■ Results Prescription profiles were generated and sent to consultant pharmacists for 9,208 patients. Pharmacists returned 7,548 (82%) of all profiles sent to them (Figure 1). After excluding 1,204 patients (13%) who were discharged or deceased, 6,344 patients were subjected to profile reviews. This number diminished to 5,160 patients who remained in the Medicaid population throughout the follow-up period, constituting an 18% dropout rate over 6 months due to death or discharge. This is consistent with historical dropout rates for Medicaid recipients. Remaining patients had received an average of 9.7 prescription fills (median www.amcp.org 9) per month during the 3-month period prior to profile generation. Exclusive of manufacturer rebates, the average PPPM drug cost to NC Medicaid was $517, with a median of $443. The comparison group consisted of 2,202 patients selected in the same manner as study-group patients (having 18 or more prescription fills in a 90-day period). We compared study and comparison groups based on age, gender, race, baseline prescription use, and dropout rates (Table 1). The groups differed with respect to race, with a lower proportion of whites in study nursing homes versus comparison-group homes (69% vs. 76%, respectively, P <0.01). At baseline, drug usage and costs were similar for study and comparison-group patients with one exception: the study subgroup with changes resulting from recommendations had higher baseline prescription costs. Dropout rates from the original cohorts were also similar across the groups (at 18% to 19%). Among study group patients, the most common PDTP alert was for a drug with a potential therapeutic duplication with an average of 5.11 alerts (Table 2). Therapeutic duplication alerts were common because a single potential duplication triggered at least 2 alerts. Clinical Initiative alerts averaged 2.77 alerts per patient. This was followed by PAL-2 or PAL-3 drugs (1.58 per patient) and Beers list drugs (0.78 per patient). A total of 6,360 interventions were offered for 3,400 patients in the study group, an average of 1.87 per patient with intervention. Based on pharmacist reporting, physicians concurred with 59.8% (3,801 of 6,360) of all interventions to change drug therapy (Table 3). Pharmacist suggestion for a more cost-effective drug was the most popular recommendation (3,327) with the greatest frequency of success (2,088, 62.8%). A recommendation for a Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 579 Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative TABLE 2 Comparison of Potential Drug Problem Alert Rates Before and After a Single Retrospective Intervention No. of Alerts Per Patient Before After (3 months) (3 months) Alert Type Difference (%) Beers List§ Study 0.78 0.70 -0.08 (-10.8) Study (w/recommendation*) 0.82 0.72 -0.10 (-12.2) Study (w/acceptance†) 0.83 0.71 -0.12 (-14.5) Comparison 0.83 0.74 -0.09 (-10.8) PAL List (2 or 3)|| Study 1.58 1.11 -0.47‡ (-29.7) Study (w/recommendation*) 1.76 1.16 -0.60‡ (-34.1) Study (w/acceptance†) 1.82 1.10 -0.72‡ (-39.6) Comparison 1.63 1.40 -0.23 (-14.1) Clinical Initiatives List¶ Study 2.77 2.47 -0.30‡ (-10.8) Study (w/recommendation*) 3.00 2.67 -0.33‡ (-11.0) Study (w/acceptance†) 3.09 2.68 -0.41‡ (-13.3) Comparison 2.73 2.71 -0.02 (-0.7) Consider Duration Flag# Study 0.16 0.15 -0.01 (-6.3) Study (w/recommendation*) 0.15 0.15 0.00 (0.0) Study (w/acceptance†) 0.14 0.14 0.00 (0.0) Comparison 0.18 0.15 -0.03 (16.7) Therapeutic Duplication** Study 5.11 4.63 -0.48 (-9.4) Study (w/recommendation*) 5.15 4.78 -0.37 (-7.2) Study (w/acceptance†) 5.22 4.75 -0.47 (-9.0) Comparison 5.00 4.56 -0.44 (-8.8) Note: The Wilcoxon 2-sample test was used to assess differences in alert rates between the comparison group and study. Sample sizes: Study group: n = 5,160 Study group with recommendations: n = 3,400 Study group with accepted recommendations: n = 2,305 Comparison group: n = 2,202 * Study group (with recommendation) = those patients having a recommendation resulting from pharmacist consultation, regardless of outcome. † Study group (with acceptance) = those patients having a recommendation and a change in therapy as a result of a recommendation provided by a pharmacist. ‡ Denotes significantly different from comparison group at P <0.01. § The Beers List is a list of drugs generally considered to be inappropriate in the elderly.17 || PAL = Prescription Advantage List, a categorization of drug alerts proposed by practicing physicians in North Carolina. PAL 3 drugs are considered to “incur significant cost.” PAL 2 drugs are considered to offer “no clear cost advantage.” PAL 1 drugs are considered to offer “significant cost savings.” The rates of PAL 2 and 3 drug alerts are shown in this table. ¶ Clinical Initiatives List refers to potential drug therapy problem alerts proposed by consultant pharmacists in North Carolina. # Consider Duration alerts were derived from classes of drugs considered appropriate only for short-term use. ** Therapeutic Duplication alerts were generated based upon duplications within hierarchical drug class codings. TABLE 3 Frequency of Recommendation by Type With Resultant Success* (n = 3,400) Recommendation Type Frequency Success, No. (%) Wrong dose or strength 545 444 (81.5) More cost-effective drug available 3,327 2,088 (62.8) Drug has potential for ADRs 632 328 (51.9) Needs additional therapy 167 69 (41.3) Other (not specified) 432 146 (33.8) Total 6,360 3,801 (59.8) * Recommendations were considered successful when a change in therapy occurred subsequent to a recommendation by the clinical pharmacist. ADRs = adverse drug reactions. 580 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 different dose garnered the highest rate of success (444 of 545, 81.5%). For Clinical Initiatives and PAL-2 or PAL-3 flags, pharmacists made interventions for change in 46.2% (2,271 of 4,916) of patients; physicians endorsed 60.2% (1,939 of 3,222) of the recommendations. Beers drugs and “consider length” (of drug therapy) categories garnered considerably fewer recommendations. We next examined persistence of computer-generated alerts in the drugs received by patients before and after intervention. Our working hypothesis was that, if the intervention program was successful, there should be a decrease in the number of PDTP alerts on subsequent patient drug profiles using the same computer-screening process employed in the before-intervention period. We found statistically significant declines in the number of alerts per patient for both PAL and Clinical Initiatives flags (P < 0.01) for all study groups (-29.7% and -10.8%, respectively) compared with the comparison group (-14.1% and -0.7%, respectively) using the Wilcoxon 2-sample test (Table 2). As expected, even greater declines in alert rates were observed in the study subgroup that received intervention (-34.1% and -11.0%) and in the subgroup that had drug therapy changes as a result of dispensing pharmacist recommendations (-39.6% and -13.3%). When compared with baseline drug use, all flag categories in all study groups had statistically significant reductions (P <0.01; Wilcoxon signed rank test), with the exception of the “consider length” (of drug therapy) flag. Finally, we examined before-after changes in the amount paid for prescriptions (Table 4). Median drug costs per patient in the intervention group decreased by $12.14 (-0.92%) from $1,329.46 to $1,317.32 and increased in the comparison group by $44.98 (3.35%) from $1,341.25 to $1,386.23, creating a relative cost reduction of $57.12 per patient in the 3-month follow-up period, or $19.04 PPPM. Even larger reductions in drug costs were observed in the study subgroups with (1) documented profile reviews and with recommendations for change, where a median decline of $25.83 per patient was observed and (2) in the subgroup for which drug therapy changes occurred as a result of the recommendations, where a decline of $61.68 per patient was observed. ■■ Discussion The results indicate that the addition of PDTP alerts to usualcare DRRs was associated with more changes in drug therapy and a reduction in computer-generated drug therapy alerts during the follow-up period. Among drug problem alert categories, we found statistically significant differences between the study group and the comparison group in alert persistence for Clinical Initiatives and PAL drugs. These 2 categories were constructed by physician and pharmacist leaders, suggesting that practitioner involvement with a centralized DUR process aids in program response. Beers list and therapeutic duplication alerts decreased in all study groups and in the comparison www.amcp.org Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative group, but persistence was not statistically different between study and comparison groups. This finding is consistent with the role of DRRs as outlined in OBRA ’87.11 These types of drugs and drug problems are explicitly mentioned as part of the guidelines for conducting customary mandated DRRs. Residents in comparison homes were not subject to drug profile reviews with PDTP alerts generated from pharmacy claims as part of the NCPP Initiative. However, residents in both study and comparison homes were subject to requirements based on OBRA ’87 and screening guidelines for the overuse of particular prescription drugs. This may explain the reduction in both groups. A JMCP article published in April 2005 demonstrated significant reductions in the use of Beers list drugs associated with an intervention involving letters to prescribers, pharmacist phone consultations, and written literature disseminated in a predominantly ambulatory population of Medicare + Choice (now Medicare Advantage) recipients.18 It would seem prudent, given previous success, to attempt to replicate the NCCP Initiative in an ambulatory Medicare setting. Notably, few recommendations were made pursuant to the “consider length” (of therapy) flag category in all study and comparison groups. This length of therapy category generated only 205 alerts in total and did not contain drugs such as benzodiazepines or psychotropic medications customarily scrutinized for length of therapy during regularly scheduled DRRs. This analysis of prescription claims data supports previous findings with regard to drug cost savings resulting from the NCPP Initiative as well as its pilot project. The NCPP Pilot Project was found to have generated an approximate 4% reduction in drug costs.19 A previously published article utilizing primary data from pharmacist reports found that the NCPP Initiative produced savings of $30.33 PPPM savings in the month immediately following the intervention.16 The resulting cost minimization ratio was determined to be 12:1.16 In the present study, we utilized Medicaid claims data to reconcile documented pharmacist interventions and to determine the downstream effects of those interventions. We also added a comparison group to further strengthen its internal validity. Using the results from Medicaid claims data in conjunction with comparison group findings, we observed a savings of $19.04 (P=0.06) PPPM for all patients receiving profile reviews, $23.60 for patients receiving interventions (P <0.01), and $35.55 (P <0.01) for patients having at least 1 accepted intervention. The 3-month PPPM difference between the study group and comparison group of $57.12 remains substantial and justifies the implementation of the Polypharmacy Initiative on the basis of drug cost savings alone. Previous projections based upon the first month immediately following the interventions did not allow us to consider the persistence of the intervention effect. An intervention may not have been carried out for reasons unknown to the consultant pharmacist. The intervention decision may have been reversed www.amcp.org TABLE 4 Total Amount Paid for Prescriptions in the Before and After Periods Before After Period Period (3 Months) (3 Months) (Median) (Median) Difference (%) P Value Study group (n = 5,160) $1,329.46 $1,317.32 -$12.14 (-0.92) 0.06 Study group (n = 3,400) (w/recommendation*) $1,392.14 $1,366.31 -$25.83 (-1.86) <0.01 Study group (n = 2,305) (w/acceptance†) $1,427.13 $1,365.45 -$61.68 (-4.32) <0.01 Comparison group (n = 2,202) $1,341.25 $1,386.23 $44.98 (3.35) n/a Note: the Wilcoxon 2-sample test was used to assess differences in total amount of paid claims between the comparison and study groups. * Study group (with recommendation) = those patients having a recommendation resulting from pharmacist consultation, regardless of outcome. † Study group (with acceptance) = those patients having a recommendation and a change in therapy as a result of a recommendation provided by a pharmacist. n/a = not applicable. by the physician after the pharmacist documented acceptance in the report. Pharmacists may also have underreported new drugs found on the nursing home medical record but not appearing on the drug profile generated from Medicaid pharmacy claims due to lag time from profile receipt to regularly scheduled DRR activities. We noted an average difference of $15 per month between claims analysis and pharmacist-reported drug cost data ($516.63 in claims analysis versus $502.96 in pharmacist-reported data) in baseline costs between these studies. This difference illustrates the importance of reconciling pharmacist intervention reporting with administrative claims. Using both data sources, as we did in the present study, is advantageous since we can tie observed medication-level interventions derived from pharmacist reporting with actual costs incurred from claims data to validate pharmacist action. The NCPP Initiative combines population-level, drug-specific surveillance of DUR programs with patient-level, comprehensive reviews characteristic of DRR activities. Alerts were generated by the payer, in this case NC Medicaid, and were provided to prescribing physicians to encourage change in targeted drugs and drug classes. In line with usual care in long-term-care settings, pharmacists were free to review and recommend therapy changes for any drug in a patient’s profile for any problem they discovered. Beginning in 2006, Medicare PDP sponsors will take on a DUR role with differing approaches to MTMP under the MMA. Standard DUR approaches have offered little evidence, to date, of effectively improving patient outcomes for state Medicaid recipients despite the large budget outlays to these programs.20-23 However, targeted, populationspecific interventions such as the NCPP Initiative have shown some success.24-26 Focused reviews based upon patient-specific Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 581 Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative profiles generated from administrative pharmacy claims, in combination with collaborative activities that individualize care,27 such as DRRs, may be a better strategy for PDPs to adopt through the MTMP service requirement. This strategy is not limited to the long-term-care setting and may in fact be more effective in an ambulatory setting where less frequent review of drug use profiles takes place. The strategy is generally applicable to any group of beneficiaries that use online adjudication for processing pharmacy claims. Limitations It was not possible to draw a true random sample of patients, nursing homes, or pharmacist consultants due to the intermingling of providers. Our comparison group was not, by design, a randomized sample of patients. Due to clustering effects, it is difficult to construct a truly randomized patient-level sample within a nursing home because physicians often provide care to patients in more than one nursing home. Additionally, groups of pharmacists are often clustered through consulting organizations serving multiple nursing homes, and multiple nursing homes often operate under a common ownership structure. Fortunately, baseline demographic characteristics and prescription drug costs between the study group and the comparison group were remarkably similar ($516.63 in the study group versus $514.56 in the comparison group). The study group, its subgroups, and the comparison group did not differ statistically with respect to gender, age, or number of prescriptions filled at baseline. There was a statistically significant difference with respect to race, with the study group and its subgroups having a greater proportion of nonwhite participants. We do not suspect that this difference confounded the results. Whatever unmeasured population differences existed, we believe our use of before-after comparisons within groups and the relatively large sample sizes enhance the validity of study results. We assume that contamination effects arising from sharing of pharmacist consultant firms between study and comparison facilities was trivial. While some pharmacist consulting firms served several different nursing homes, no individual pharmacist provided consulting services to both study and comparison group homes. Pharmacist turnover was not a problem since the time period was relatively short. To the extent that contamination effects were present, they would serve to diminish observed between-group differences. We do not know the effect of repeated interventions, the effects of continually evolving PDTP alerts criteria,28 or intervention persistence beyond 6 months. We cannot confidently project the long-term impact of these interventions. Our 3-month follow-up period reflected a balance in our approach. On the one hand, we wanted at least two 1-month follow-up periods to ensure that drug therapy changes were reflected in claims data and persisted. On the other hand, a longer follow-up period of 6 to 12 months would 582 Journal of Managed Care Pharmacy JMCP September 2005 have incurred problems of patient attrition within the nursing homes, given the statewide average attrition rate of 36% per year. Yet another factor was the strong desire by the sponsor to finish the analysis of Phase 1 as soon as possible for public policy planning and budgeting purposes. Using administrative claims data to measure differences in drug costs is not without limitations. Drugs may have been filled without submission of a claim, or nursing homes may have paid for products such as over-the-counter medications out of a separate budget. However, this study takes a payer perspective, and paid claims are the most meaningful measurement from this perspective. Administrative claims are also poor standalone proxies for measuring changes in quality, particularly in such areas as adverse effects or health status. On the other hand, the very large sample sizes involved in our study suggest that our findings are real and replicable. As with any nonrandomized observational study, regression toward the mean must be considered. We chose our comparison group in the same way we chose study group patients; hence, both should have equally incurred this regression effect, and it is, in essence, neutralized for purposes of differential analysis. Using a payer perspective, we assessed the impact of all drug claims not just those drugs flagged in profiles from preintervention screening. It is likely that our broader focus diluted our findings toward the null. Yet we found important drug cost differences on a PPPM basis. ■■ Conclusions A program of supplemental pharmacist review targeting patients with high drug use and the potential for multiple drug therapy problems was successful in generating changes in drug therapy. We believe that involving pharmacists and physicians in the creation of PDTP alerts was crucial to widespread adoption. The changes in drug therapy that resulted from a single (compensated) pharmacist retrospective review significantly reduced the number of PDTP alerts at follow-up. Currently, regulations governing DRRs do not explicitly focus on costeffectiveness or cost reductions of pharmaceuticals received by patients, nor do they explicitly compensate reviewers for such services. Results from this study suggest that a program to encourage and compensate pharmacists for conducting focused reviews of drug therapy regimens for targeted high-risk patients as a supplement to usual mandated review activities can lower drug therapy costs and maintain or enhance the quality of drug therapy. Interventions by pharmacists were economically beneficial when labor costs and savings in drug costs are considered. Elements of this program can be applied to both ambulatory and long-term-care settings. ACKNOWLEDGMENTS The authors thank the North Carolina Long Term Care Pharmacy Alliance; North Carolina Office of Research, Demonstrations and Rural Health Development; North Carolina Medical Directors Association; North Carolina Vol. 11, No. 7 www.amcp.org Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting: Results From the North Carolina Polypharmacy Initiative Health Care Facilities Association; and the University of North Carolina School of Pharmacy at Chapel Hill for their endorsement of this project. The authors also thank CoraLynn B. Trewet, MS, PharmD, University of Iowa, Des Moines, and Brandy Newman, PharmD, King Soopers Pharmacy, Aurora, CO, for their assistance with data compilation and manuscript guidance. DISCLOSURES Financial and technical support of this project was provided by AccessCare, Inc., a provider network within Community Care of North Carolina, and the American Foundation for Pharmaceutical Education and was obtained by authors Dale Christensen and Steven E. Wegner (AccessCare) and Troy K. Trygstad (AFPE). Trygstad, Wegner, and author Jennifer Garmise are employed by AccessCare. The authors disclose no potential bias or conflict of interest relating to this article. Author Trygstad served as principal author of the study. Study concept and design were contributed by Trygstad, Christensen, Wegner, and author Robert Sullivan. Analysis and interpretation of data were contributed by Trygstad, Christensen, and Sullivan. Drafting of the manuscript was primarily the work of Trygstad, and its critical revision was the work of all authors. Statistical expertise was contributed by Trygstad and Christensen. REFERENCES 1. Heffler S, Smith S, Keehan S, Clemens MK, Won G, Zezza M. Health spending projections for 2002-2012 [Web exclusive, February 7, 2003]. Health Aff. Available at: http://www.healthaffairs.org. Accessed Feburary 24, 2005. 2. Hanlon JT, Artz MB. Drug-related problems and pharmaceutical care: what are they, do they matter, and what’s next? Med Care. 2001;39:109-12. 3. Hanlon JT, Schmader K, Gray S. Adverse drug reactions. In: Delafuente JC, Stewart RB, eds. Therapeutics in the Elderly. 3rd ed. Cincinnati, OH: Harvey Whitney Books; 2000:289-314. 4. Hanlon JT, Schmader KE, Ruby CM, Weinberger M. Suboptimal prescribing in elderly inpatients and outpatients. J Am Geriatr Soc. 2001;49(2):200-09. 5. McPeak ME. State-by-state look at Medicaid retail pharmacy delivery: trials and triumphs. J Manag Care Pharm. 1998;4(6):546-47, 551-52, 554. 6. Carter BL, Chrischilles EA, Scholz D, Hayase N, Bell N. Extent of services provided by pharmacists in the Iowa Medicaid Pharmaceutical Case Management program. J Am Pharm Assoc. 2003;43(1):24-33. 7. Christensen DB, Neil N, Fassett WE, Smith DH, Holmes G, Stergachis A. Frequency and characteristics of cognitive services performed by pharmacists in response to a financial incentive. J Am Pharm Assoc. 2000;41:609-17. 8. Medication Therapy Management Services Definition and Program Criteria. Academy of Managed Care Pharmacy, American Association of Colleges of Pharmacy, American College of Apothecaries, American College of Clinical Pharmacy, American Society of Consultant Pharmacists, American Pharmacists Association, American Society of Health Systems Pharmacists, National Association of Boards of Pharmacy, National Association of Chain Drug Stores, National Community Pharmacists Association, National Council of State Pharmacy Association Executives. Approved July 27, 2004. Available at: http://www.aphanet.org/AM/Template.cfm?Section=APhA_Resources_Medicare &Template=/CM/ContentDisplay.cfm&ContentID=1681. Accessed June 15, 2005. Also available at: www.aacp.org/Docs/MainNavigation/Resources/ 6308_MTMServicesDefinitionandProgramCriteria27-Jul-04.pdf. Accessed June 15, 2005. 9. American Pharmacists Association, National Association of Chain Drug Stores. Medication therapy management in community pharmacy practice: core elements of an MTM service. April 29, 2005. Available at: http://www.aphanet.org/AM/Template.cfm?Template=/CM/ContentDisplay.cfm &ContentID=3303. Accessed June 15, 2005. www.amcp.org 10. The Centers for Medicare and Medicaid Services. Medication therapy management: excerpt from the Medicare Prescription Drug Benefit Final Rule 42 CFR Parts 400, 403, 411, 417, 423. Available at: http://www.cms.hhs.gov/ medicarereform/pharmacy/mtm.pdf. Accessed June 15, 2005. 11. Clark TR, ed. Nursing Home Survey Procedures and Interpretive Guidelines. Alexandria, VA.: American Society of Consultant Pharmacists; 1999. 12. Bootman JL, Harrison DL, Cox E. The health care cost of drug-related morbidity and mortality in nursing facilities. Arch Intern Med. 1997;157:208996. 13. Grymonpre RE, Mitenko PA, Sitar DS, et al. Drug-associated hospital admissions in older medical patients. J Am Geriatr Soc. 1988;36:1092-98. 14. North Carolina Medicaid, Table 10, state fiscal year 2003, Medicaid service expenditures by recipient group. Available at: http://www.dhhs. state.nc.us/dma/ncms.htm. Accessed August 13, 2005. 15. North Carolina Medicaid, Table 11, state fiscal year 2003, Medicaid service expenditures by recipient group. Available at: http://www.dhhs. state.nc.us/dma/ncms.htm. Accessed August 13, 2005. 16. Christensen DB, Trygstad TK, Sullivan R, et al. A pharmacy management intervention for optimizing drug therapy for nursing home patients. Am J Geriatr Pharmacother. 2004;2(4):248-56. 17. Beers MH. Explicit criteria for determining potentially inappropriate medications use by the elderly. Arch Intern Med. 1997;157:1531-36. 18. Kaufman MB, Brodin KA, Sarafian A. Effect of prescriber education on the use of medications contraindicated in older adults in a managed Medicare population. J Manag Care Pharm. 2005;11(3):211-19. 19. Christensen DB, Trygstad TK. Assessment of the Polypharmacy Initiative in Nursing Homes: a preliminary analysis. Report to Carmen Hooker Odom, Secretary, North Carolina Department of Health and Human Services. July 2002. 20. A Soumerai SB, Lipton HL. Computer-based drug-utilization review— risk, benefit, or boondoggle? N Engl J Med. 1995;332(24):1641-45. 21. Hennessy S, Bilker WB, Zhou L, et al. Retrospective drug utilization review, prescribing errors, and clinical outcomes. JAMA. 2003;290(11):1494-99. 22. Stuart B, Fahlman C. Outcomes of prospective drug-use review of betaagonist inhaler use in an elderly Medicaid population. Clin Ther. 1999;21(12): 2094-112. 23. Kidder D, Bae J. Evaluation results from prospective drug utilization review: Medicaid demonstrations. Health Care Financ Rev. 1999;20(3):107-18. 24. Smith DH, Christensen DB, Stergachis A, Holmes G. A randomized controlled trial of a drug use review intervention for sedative hypnotic medications. Med Care. 1998;36(7):1013-21. 25. Walker S, Willey CW. Impact on drug costs and utilization of a clinical pharmacist in a multisite primary care medical group. J Manag Care Pharm. 2004;10(4):345-54. 26. Zuckerman IH, Weiss SR, McNally D, Layne B, Mullins CD, Wang J. Impact of an educational intervention for secondary prevention of myocardial infarction on Medicaid drug use and cost. Am J Manag Care. 2004;10(7 pt 2): 493-500. 27. Crownover BK, Unwin BK. Implementation of the Beers criteria: sticks and stones—or throw me a bone. J Manag Care Pharm. 2005;11(5):416-17. 28. Curtiss FR. Evidence-based medicine: which drugs are truly contraindicated for use in older adults? J Manag Care Pharm. 2005;11(3):259-60. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 583 EDITORIAL ■■ Fish Oil for Heart Disease—Happy to Be in Second Place From 1970-1977, the Minnesota Vikings played in the National Football League’s championship game—the Super Bowl—four times. They lost every time. The Denver Broncos repeated the feat until finally winning in their fifth attempt. Both teams were mocked and pitied for their futility; however, there were worse fates than second place. The rest of their conference teams had spent millions of dollars and countless hours of work for the same opportunity and fell short. Obviously, first place was the goal, but coming in second was still better than 26 other competitors. In the highly competitive field of lipid treatment, many groups strive to stand out from their peers. Niacin, fibrates, and resins were the early favorites but were overtaken by statins. Statins are the usual first-line recommended medical treatment by multiple expert panels for high-risk primary and secondary prevention of atherosclerotic cardiovascular disease (ASCVD).1-4 The large volume of morbidity and mortality data makes them the undisputed champions for treating dyslipidemia.5-8 Ezetimibe has been touted as add-on therapy for additional low-density lipoprotein cholesterol (LDL-C) lowering; however, we are still waiting for data to show that the LDL-C effect of this agent results in any benefit for myocardial infarction, stroke, cardiovascular mortality, or overall mortality.9 Hormone replacement therapy also showed value for LDL-C lowering in observational studies; however, we now know that stroke and myocardial infarction risk actually increase despite the improvement in lipid values.10 Obviously, other known factors are at work, such as antiinflammatory and antiplatelet effects to explain why some agents that lower LDL-C help morbidity and mortality, while others are neutral or actually harmful. Into the fray, a new challenger is now gaining notoriety— omega fatty acids, also known as fish oil supplements. (Terminology is complex regarding marine versus plant sources and subcomponents of omega fatty acids. An extensive description is beyond the scope of this editorial, and excellent reviews are available.11) Many clinicians have rightfully been leery of unproven claims from the herbal sector. As food supplements, they carry the classic label: “(T)his product is not intended to diagnose, treat, cure, or prevent any disease.” A few notable exceptions exist, such as the Rotta preparation of glucosamine for joint pain, but the majority of agents lack substantive support of effectiveness.12 Evidence has been building lately for omega fatty acids, especially for secondary ASCVD prevention. In a randomized controlled trial of patients with recent myocardial infarction, 2,836 individuals were given omega fatty acids (dose 850-882 mg) while 2,828 were given placebo. After 3.5 years, the treatment group had a lower rate for the combined end point of overall mortality, nonfatal stroke, and recurrent infarction relative to placebo (12.3% vs. 14.6%, number needed to treat = 44, P = 0.048).13 A meta-analysis in 2002 pooled 11 trials involving almost 16,000 patients for primary and secondary prevention, performed over the previous 30 years.14 With an average followup of 20 months and dosing ranging from 0.3-6.0 g/day, results showed omega fatty acids significantly decreased the risk of fatal myocardial infarction (risk ratio = 0.7; 95% confidence interval [CI], 0.6-0.80), sudden death (risk ratio=0.6; 95% CI, 0.6-0.9), and overall mortality (risk ratio = 0.8; 95% CI, 0.7-0.9). Pulling the evidence into tighter focus was a meta-analysis in 2005 that compared true end points, such as overall mortality, among the various classes of lipid-lowering agents.15 In more than 100,000 patients enrolled in 35 statin trials, the risk ratio for overall mortality was 0.87 for statins compared with placebo (95% CI, 0.81-0.94), with a 20% average reduction in total cholesterol. Surprisingly, for overall mortality in 15 studies involving 20,000+ patients with preexisting heart disease treated with omega fatty acids, the risk ratio was 0.77 (95% CI, 0.630.94), despite a negligible 2% average reduction in total cholesterol. Statins lowered cholesterol 10-fold more than omega fatty acids yet did not outperform on mortality measures in patients with ASCVD. Fibrates, niacin, resins, and diet therapy all failed to show significance in lowering overall mortality. For primary prevention with statins, it is necessary to treat 228 persons (95% CI, 123-2,958) for 3.3 years. In patients with known heart disease, 50 patients (95% CI, 38-78) would have to be treated with a statin to prevent one additional death, and 44 patients (95% CI, 31-84) would need to be treated with fish oil to prevent one additional death, each over an average 4.4 years of therapy (excluding one low-quality study).15 Not all reviews of omega fatty acids have been positive. An earlier Cochrane meta-analysis that was last reviewed in 2004 failed to find a reduction in overall mortality for omega fatty acids. No safety concerns were found, but the authors recommended additional research.16 An evidence report from the Agency for Healthcare Research and Quality evaluated 123 trials for the effect of omega fatty acids on intermediate markers for ASCVD. Consistent with prior studies, they found a reduction of triglycerides ranging from 10% to 33%, but negligible effects on LDL-C or high-density lipoprotein cholesterol.17 The report did not assess the effect on overall mortality. Despite the differences among reviews, the U.S. Food and Drug Administration determined in November 2004 that sufficient data existed to approve the first prescription omega fatty acid agent in the United States; the agent was first launched in Austria in July 2002.18,19 Containing 1 g per capsule, the agent is indicated for adjunct therapy to diet at 4 g daily to reduce very high (≥ 500 mg/dL) triglyceride levels in adult patients. Of note, the prescription agent will vary markedly from over-the-counter preparations. For example, a 1 g capsule of the prescription agent will have roughly triple the amount of fish oil as a popular 1 g capsule available commercially at present (eicosapentaenoic acid 465 mg vs. 180 mg; docosahexaenoic acid 375 mg vs. 120 mg).20 It is expected that the 584 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Editorial manufacturer will push for approval of 1 g daily dosing in postmyocardial infarction patients for secondary prevention of ASCVD. Although more studies are needed, omega fatty acids are now standing out from the crowd. They will likely never replace statins as first-line therapy for dyslipidemia, but they have more data to support a benefit for reducing mortality in patients with preexisting heart disease than do the other antilipid agents. Central to their success in population management of coronary heart disease and mortality risk will be low discontinuation rates. As discussed previously in JMCP, 18-month discontinuation rates greater than 50% are common for all antilipid drug classes.21 Despite various hurdles, omega fatty acids are now ready to take over second place in the lineup for treatment of hypertriglyceridemia and secondary prevention of ASCVD. Brian K. Crownover, MD, FAAFP, Lt. Col., MC, USAF Associate Editor [email protected] 7. Ross DS, Allen IE, Connelly JE, et al. Clinical outcomes in statin treatment trials. A meta-analysis. Arch Intern Med. 1999;159:1793-1802. 8. MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5,963 people with diabetes: a randomised placebo-controlled trial. Lancet. 2003;361:2005-16. 9. Ballantyne CM, Houri J, Notarbartolo A, et al. Effect of ezetimibe coadministered with atorvastatin in 628 patients with primary hypercholesterolemia. Circulation. 2003;107:2409-15. 10. Writing Group for the Women's Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women's Health Initiative randomized controlled trial. JAMA. 2002;288:321-33. 11. Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the food chain in the United States. Am J Clin Nutr. 2000;71(1)179S-188S. 12. Towheed TE, Maxwell L, Anastassiades TP, et al. Glucosamine therapy for treating osteoarthritis. Cochrane Database Syst Rev. 2005; issue 2; art. no.: CD002946.pub2. DOI: 10.1002/14651858.CD002946.pub2. 13. GISSI-Prevenzione investigators. Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Lancet. 1999;354:447-55 14. Bucher HC, Hengstler P, Schindler C, Meier G. N-3 polyunsaturated fatty acids in coronary heart disease: a meta-analysis of randomized controlled trials. Am J Med. 2002;112(4):298-304. 15. Studer M, Briel M, Leimenstoll B, Glass TR, Bucher HC. Effect of different antilipidemic agents and diets on mortality. A systematic review. Arch Intern Med. 2005;165:725-30. DISCLOSURE The author is a board-certified family physician assigned to Eglin AFB Florida, where he serves as residency program director. The opinions and assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of any organization, including the U.S. Air Force medical department or the U.S. Air Force. He discloses no potential bias or conflict of interest relating to this editorial. REFERENCES 1. Veterans Health Administration, Department of Defense. VHA/DoD clinical practice guideline for the management of dyslipidemia in primary care. Washington, DC: Veterans Health Administration, Department of Defense; December 2001. 16. Hooper L, Thompson RL, Harrison RA, et al. Omega 3 fatty acids for prevention and treatment of cardiovascular disease. Cochrane Database Syst Rev. 2004, issue 4; art. no.: CD003177.pub2. DOI: 10.1002/14651858. CD003177.pub2. 17. Balk E, Chung M, Lichtenstein A, et al. Effects of omega-3 fatty acids on cardiovascular risk factors and intermediate markers of cardiovascular disease. Summary, Evidence Report/Technology Assessment: Number 93. Rockville, MD: Agency for Healthcare Research and Quality; March 2004. AHRQ Publication no. 04-E010-1. Available at: http://www.ahrq.gov/clinic/ epcsums/o3cardrisksum.htm. Accessed July 2, 2005. 18. Omacor [product label]. Washington, DC: U.S. Food and Drug Administration; 2005. Available at: http://www.fda.gov/cder/foi/label/2004/ 21654lbl.pdf Accessed July 2, 2005. 2. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110(2):227-39. 19. Omacor [product information]. Marietta, GA: Solvay Pharmaceuticals; 2005. Available at: http://www.solvaypharmaceuticals.com/html/products/ Cardiology/omacor.html#Worldwide%20avail. Accessed July 2, 2005. 3. Haffner SM. Dyslipidemia management in adults with diabetes. Diabetes Care. 2004;27(suppl 1):S68-S71. 20. GNC Fish Body Oils 1000 [product information]. Pittsburgh, PA: General Nutrition Centers, Inc.; 2005. Available at: http://www.gnc.com/ productDetails.aspx?id=887911&lang=en. Accessed July 2, 2005. 4. Institute for Clinical Systems Improvement. Lipid management in adults. Bloomington, MN: Institute for Clinical Systems Improvement; July 2004. 5. Scandinavian Simvastatin Survival Study Group. Randomized trial of cholesterol lowering in 4,444 patients with coronary heart disease: the Scandinavian simvastatin survival study (4S). Lancet. 1994;344:1383-89. 21. Abugosh SM, Kogut SJ, Andrade SE, Larrat EP, Gurwitz JH. Persistence with lipid-lowering therapy: influence of the type of lipid-lowering agent and drug benefit plan option in elderly patients. J Manag Care Pharm. 2004;10(5): 404-11. 6. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med. 1998;339:1349-57. www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 585 EDITORIAL ■■ Medication Therapy Management Services for Long-term Care Patients: No Road Maps for Those Trying to Find Their Way As required by the Medicare Modernization Act,1 organizations offering Medicare Part D prescription drug plans (PDPs) are obligated to provide medication therapy management programs (MTMPs) as part of the benefit. While medication therapy management (MTM) has been performed by pharmacists in a variety of settings for years, a consensus definition of MTM was not developed until last year.2 As defined in the Act, MTMPs are “furnished by a pharmacist [and] designed to assure . . . that covered part D drugs are appropriately used to optimize therapeutic outcomes through improved medication use, and to reduce the risk of adverse events, including adverse drug interactions.”1 While the Act defines the beneficiaries who are targeted for MTM (those with multiple chronic conditions, taking multiple medications, and likely to have high drug expenses) and basic elements of the program (promotion of enhanced enrollee understanding of the appropriate use of medications, adverse event risk reduction, increased medication adherence, and detection of adverse drug events and patterns of prescription drug overuse or underuse), the government has taken a decidedly hands-off approach to regulating MTMPs. Citing insufficient standards and performance measures to evaluate MTMPs and the inability to determine what requirements would enhance MTMPs and improve patient outcomes, the Centers for Medicare and Medicaid Services (CMS) has given what it considers maximum flexibility to plans so they may develop MTMPs that can achieve the goal of improving therapeutic outcomes. Organizations that bid on the Part D drug plans have remained silent about their MTMP plans, but some guidance is available as to what can be expected for MTM services at certain pharmacy settings. In April 2005, the American Pharmacists Association and the National Association of Chain Drug Stores Foundation released a model framework for implementing MTM services in a community pharmacy setting.3 The framework (a) provides 9 factors that community pharmacists can consider when targeting patients for MTM services, (b) identifies 5 core components of MTM programs in community pharmacy (medication therapy review, a personal medication record, a medication action plan, intervention and referral, and documentation and follow-up) and describes pharmacist responsibilities for each, and (c) provides sample personal medication records and medication action plans for community pharmacists to use in their programs. Regarding senior care, the American Society of Consultant Pharmacists (ASCP) released an issue paper on MTM services for ambulatory Medicare beneficiaries in April 2004, reviewing the goals of MTM services in the elderly and the types of MTM services that a pharmacist could provide to ambulatory seniors.4 An MTMP for low-income ambulatory seniors was described by Stebbins et al. in a recent issue of JMCP.5 However, there is little guidance for developing MTM services for those residing in long-term care (LTC) facilities. At this time, the only point that is clear is that MTM services need to be distinct from the monthly drug regimen review (DRR) process mandated for all LTC residents. While the DRR process may save as much as $3.6 billion per year in medication-related problems, these problems still generate an estimated $4 billion in annual costs in the LTC setting.6 Inappropriate prescribing remains a problem for the nation’s 1.8 million LTC and assistedliving residents and is a significant factor contributing to the risks for morbidity and mortality.7-9 While the deadline has passed for submitting bids for Part D PDPs for 2006, organizations preparing bids for PDPs for 2007 and beyond should look for evidence of success in MTM services in the LTC arena when formulating their bids. In this issue, Trygstad et al. describe results of the North Carolina Polypharmacy Initiative, a targeted drug therapy management consultation where a pharmacist reviews medical records and Medicaid pharmacy claim drug profiles to determine if a potential drug therapy problem alert requires action and, if so, to make interventions and determine if a therapy change was enacted.10 As measured by the study objectives, the program was successful because it was associated with a reduction in potential drug therapy problem alerts and median per-patient-per-month prescription drug costs. The program further demonstrated the utility of consultant pharmacists in the LTC setting since physicians implemented nearly 60% of pharmacist recommendations. However, the study by Trygstad et al. did not look at health outcomes and therefore did not determine if the drug therapy changes would maintain or improve the quality of care that residents received. It is unlikely that a consultant pharmacist would make a recommendation that would result in lower quality of care for an LTC resident, but the study by Trygstad et al. measured only intermediate (process of care) and drug cost outcomes. Research is still necessary to demonstrate that a pharmacy intervention program designed for multiple disease states in LTC leads to improvements in resident health outcomes. Such research would provide powerful evidence of an intervention’s utility. It is easy to sympathize with the researchers; by intervening in many different disease states, it is very difficult to quantify the impact of the program on LTC resident health outcomes. Limiting the study to retrospective pharmacy claims data makes even more remote the possibility of determining health outcomes. In the MTMP described in the study, the pharmacist had the ability to recommend that drug therapy be withdrawn, modified, or added, as each case required. Changes to drug therapy impact more than prescription drug expenses; these changes can have an effect on immediate and longer-term medical expenses as well. However, the study only presented prescription drug costs and did not report the impact of the intervention on medical costs or overall health care costs. The true cost for a 586 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Editorial health system of an intervention or series of interventions is best assessed by examining overall costs. How Part D providers choose to examine the costs of MTMPs may very well depend upon the type of program they are offering. Medicare Advantage prescription drug (MA-PD) plans are at risk for overall health costs, so their level of interest in the impact of a MTMP on overall costs is likely to be greater than a PDP, which is not at risk for overall health costs. One must wonder if the Part D program will suffer from the silo approach to health care finance that managed care plans have struggled to overcome. CMS’s proposal11 for linking deidentified Part D prescription claim data to hospital, physician, and other medical utilization data in a public database may allow health policy analysts to compare prescription drug expenses and overall health care expenses for PDPs and MA-PDs. More questions than answers presently remain pertaining to MTM services in LTC facilities. While the forthcoming report from the Department of Health and Human Services on standards of practice for pharmacy services (including clinical services) provided to patients in LTC settings will be a useful tool for PDP developers to consider when designing or refining MTMPs, the experiences that the LTC industry encounter starting on January 1, 2006 (“Part D-Day”), will be even more valuable. The sharing of MTM best practices by PDPs and LTC facilities can reduce the trial-and-error approach to MTM services and can improve the quality of services delivered to LTC residents. Joshua J. Spooner, PharmD, MS Director, Clinical and Outcomes Services Advanced Concepts Institute University of the Sciences in Philadelphia Philadelphia, Pennsylvania [email protected] DISCLOSURE The author discloses no potential bias or conflict of interest relating to this editorial. REFERENCES 1. Public Law 108-173. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003. December 8, 2003. Available at: http://www.cms. hhs.gov/medicarereform/. Accessed August 12, 2005. 2. Medication therapy management services definition and program criteria. July 27, 2004. Available at: http://www.amcp.org/data/jmcp/Letters-179186.pdf and http://www.aacp.org/Docs/MainNavigation/Resources/ 6308_MTMServicesDefinitionandProgramCriteria27-Jul-04.pdf. Accessed August 12, 2005. 3. The American Pharmacists Association and the National Association of Chain Drug Stores Foundation. Medication therapy management in community pharmacy practice: core elements of an MTM service. Version 1.0. April 29, 2005. Available at: http://www.aphanet.org/AM/Template.cfm?Template=/ CM/ContentDisplay.cfm&ContentID=3303. Accessed August 12, 2005. 4. The American Society of Consultant Pharmacists. Medication therapy management services for ambulatory Medicare beneficiaries. April 28, 2004. Available at: http://www.ascp.com/medicarerx/docs/ASCPMTMS.pdf. Accessed August 12, 2005. 5. Stebbins MR, Kaufman DJ, Levens Lipton H. The PRICE clinic for lowincome elderly: a managed care model for implementing pharmacist-directed services. J Manag Care Pharm. 2005;11(4):333-41. 6. Bootman JL, Harrison DL, Cox E. The health care cost of drug-related morbidity and mortality in nursing facilities. Arch Intern Med. 1997;157:2089-96. 7. Gurwitz JH, Field TS, Avorn J, et al. Incidence and preventability of adverse drug events in the nursing home setting. Am J Med. 2000;109:87-94. 8. Sloane PD, Gruber-Baldini AL, Zimmerman S, et al. Medication undertreatment in assisted living settings. Arch Intern Med. 2004;164:1621-25. 9. Gerety MB, Cornell JE, Plichta DT, Eimer M. Adverse events related to drugs and drug withdrawal in nursing home residents. J Am Geriatr Soc. 1993;41:1326-32. 10. Trygstad TK, Christensen D, Garmise J, Sullivan R, Wegner SE. Pharmacist response to alerts generated from Medicaid pharmacy claims in a long-term care setting: results from the North Carolina Polypharmacy Initiative. J Manag Care Pharm. 2005;11(7):575-83. 11. Centers for Medicare and Medicaid Services. Medicare prescription drug data strategy: improving evidence for patient care through the Medicare prescription drug benefit. Available at: http://www.cms.hhs.gov/medicarereform/ CMSPaper-DataStrategyforMedicareDrugBenefitOverview.pdf. Accessed August 8, 2005. www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 587 EDITORIAL ■■ Evaluating the Relationship Between Diabetes Treatment and Health Care Costs—Measuring the Atom With a Yardstick? In this issue of JMCP, Shetty et al. evaluated the relationship between poor glycemic control, defined as a glycosylated hemoglobin (HbA1c, now commonly abbreviated as simply A1c) >7%, and diabetes-related medical and pharmacy costs.1 They used 18-months of data from medical and pharmacy administrative claims as well as laboratory data. Other studies have been conducted with similar data sources that also characterized the relationship of poor glycemic control to health care costs, hospitalizations, and resources used.2-4 A principal limitation of each of these analyses was the inability of the investigators to determine disease duration. In their analysis, Shetty et al. attempted to adjust for a number of potential confounders, including age, gender, physician specialty, comorbid conditions, and total baseline costs. However, they looked at patients for only a 6-month preindex and a 12-month follow-up period, so they could not estimate whether patients were newly diagnosed diabetics or whether they had disease of long-standing duration. The average diabetic many years after diagnosis is very different from a patient with new-onset diabetes5; they are more dependent on insulin, more difficult to control, and have more insulin resistance. Thus, if investigators had been able to control for the length of time each patient had the disease, the relationship between A1c level and costs would have been better elucidated. Of course, the duration of disease is difficult to determine in these types of analyses. To measure resource utilization directly, investigators require health plan data for patients who are continuously enrolled over the entire study period in order to ensure that all resources used by the patient are captured. Because patients frequently change health plans over the course of their lifetimes, the follow-up period required to determine disease duration would be prohibitive for this type of analysis. The ideal database for this type of analysis would include patient data for an entire lifetime, but such an administrative claims database does not exist anywhere in the United States. Alternatively, small-scale analyses could be conducted using medical and pharmacy data from small, rural communities where populations are relatively stable and receive health care from a closed set of hospitals and clinics. However, even then, a large proportion of patients would likely have to be excluded because of moving into or out of the geographic area. In addition, patients would not be representative of the larger population as a whole, and sample sizes might be too small to ensure adequate power. Another limitation to these types of analyses is that International Classification of Diseases, Ninth Revision (ICD-9) coding is replete with errors. Shetty et al. used a range of ICD-9 codes indicative of type 2 diabetes and excluded patients with two or more claims for type 1, presumably because investigators realized that disease misclassification was likely present in the database. Interestingly, the investigators reported that 35.5% of the patients with two or more ICD-9 codes for type 2 diabetes also had two or more ICD-9 codes for type 1. As a result, more than one third of the source population had to be excluded because of an inability to determine which disease the patient actually had. This illustrates the error inherent in using a database such as this, and one has to ask whether excluding the patients with conflicting diagnoses really eliminated all of the misclassification bias. Any residual non-differential misclassification of disease in this database would bias the results of the analysis to the null hypothesis.6 Thus, we might expect an even stronger association between a lack of glycemic control and increased medical and pharmacy costs. The accuracy of administrative databases has been questioned previously. Peabody et al. attempted to characterize the extent of error in administrative data sources by sending scripted actors to simulate the symptoms of 4 diseases, including diabetes.7 Although a study designed to measure how accurately patients who are feigning illness are diagnosed with that illness has its own inherent problems, investigators reported that administrative data sets contained the correct primary diagnosis only about 57% of the time. They broke down the places in the chain of events following the patient encounters where errors were made and found that the clinicians made incorrect diagnoses 13% of the time, encounter forms were missing or incomplete 8% of the time, and the data were incorrectly entered from encounter forms 22% of the time. Even if the clinicians in this study made incorrect diagnoses because the patients did not really have the disease, the administrative database was still wrong 30% of the time. Interestingly, Peabody et al. found that the accuracy of the database for comorbid or secondary diagnoses among diabetes patients, such as hypertension and hypercholesterolemia, was even lower. They reported finding accurate secondary diagnoses only 36% of the time, and most of the error (42%) occurred when the data were incorrectly or incompletely entered from the encounter forms. The investigators attributed the poor quality of recording of comorbid condition to specific variations in administrative systems at each site; not all of the sites required the entry clerks to document secondary diagnoses. In light of the questionable quality of these data sources with respect to comorbidities, one has to wonder about the true impact of these conditions on the outcomes reported by Shetty et al. The investigators adjusted for 11 comorbidities in their analysis, but only 6 of them were significantly correlated with cost (hypertension, dyslipidemia, retinopathy, nephropathy, neuropathy, and diseases of the lower extremities). If comorbidity tracking was as inaccurate in their dataset as it was in that evaluated by Peabody et al., the impact on the study outcomes could have been substantial. An additional problem with these data sources is the degree of resolution. Encounter forms used in the clinic setting usually have a list of the top 20 to 30 potential diagnoses clinicians are 588 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Editorial likely to see. These diagnoses tend to correspond to the most general classification of a disease, including the NEC (not elsewhere classified) or NOS (not otherwise specified) categories. A report by a Canadian investigator into the accuracy of ICD-9 reporting of complications in gall-bladder surgery found that when non-specific codes were interpreted as indicating clinically important complications, the rate of these complications appeared much higher.8 The interpretation of nonspecific diagnosis codes must be tempered with the understanding that these categories may be used excessively, despite the existence of more-specific codes that are more accurate. This results in poor resolution for investigators using these data to answer outcomes research questions, and yet, they are often asked pointed, specific questions that require telescopic resolution in order to answer adequately. Outcomes research is often an attempt to measure minute details with somewhat inaccurate instruments— analogous to measuring atomic particles with a yardstick. Frequently, investigators attempt to answer these questions anyway, and it is left up to the reader to determine the magnitude of the limitations of the reported results. Data sources for outcomes researchers are very crude tools. Understanding the inaccuracies and limitations of administrative datasets is critical. To discount all outcomes research because of these limitations is to be overcautious and is unwarranted. These databases are valuable, particularly because they allow us to readily and inexpensively evaluate large numbers of patients across geographically diverse areas and because they are often closed systems, capturing all patient encounters over the enrollment period. Few other data sources have such characteristics. However, readers should bear in mind the limitations when interpreting and evaluating these studies and when basing patient care decisions on them. Perhaps a utopian future exists where national health tracking databases include patient data for a lifetime and administrative systems document patient data with high resolution. Until then, our only choice is to keep sharpening our instruments and learning what we can from the available sources. DISCLOSURE The author is program coordinator, Drug Regimen Review Center, and research assistant professor, Department of Pharmacotherapy, University of Utah College of Pharmacy, Salt Lake City, Utah. She discloses no potential bias or conflict of interest relating to this editorial. REFERENCES 1. Shetty S, Secnik K, Oglesby AK. Relationship of glycemic control to total diabetes-related costs for managed care health plan members with type 2 diabetes. J Manag Care Pharm. 2005;11(7):559-64. 2. Gilmer TP, O’Conor PJ, Manning WG, Rush WA. The cost to health plans of poor glycemic control. Diabetes Care. 1997;20:1847-53. 3. Menzin J, Langley-Hawthorne C, Friedman M, Boulanger L, Cavanaugh R. Potential short-term economic benefits of improved glycemic control: a managed care perspective. Diabetes Care. 2001;24(1):51-55. 4. Wagner EH, Sandhu N, Newton KM, McCulloch DK, Ramsey SD, Grothaus LC. Effect of improved glycemic control on health care costs and utilization. JAMA. 2001;285:182-89. 5. Fox CS, Sullivan L, D’Agostino RB Sr, Wilson PW. Framingham Heart Study. The significant effect of diabetes duration on coronary heart disease mortality: the Framingham Heart Study. Diabetes Care. 2004;27(3):704-08. 6. Rothman KJ, Greenland S. Precision and validity in epidemiologic studies. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins; 1998:115-34. 7. Peabody JW, Luck J, Jain S, Bertenthal D, Glassman P. Assessing the accuracy of administrative data in health information systems. Med Care. 2004;42: 1066-72. 8. Taylor B. Common bile duct injury during laparoscopic cholecystectomy in Ontario: does ICD-9 coding indicate true incidence? Can Med Assn J. 1998; 158:481-85. Joanne LaFleur, PharmD Associate Editor [email protected] www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 589 EDITORIAL SUBJECTS—IN THIS ISSUE AND IN PREVIOUS ISSUES ■■ COX-2 Inhibitors: Little or No GI Protection, Increased Risk of Cardiovascular Events, High Cost, and Other Class-less Effects On September 30, 2004, the U.S. Food and Drug Administration (FDA) requested the market withdrawal of rofecoxib (Vioxx) due to safety concerns associated with an apparent increased risk of cardiovascular events, particularly heart attack and stroke.1 On February 1, 2005, about 9,000 physicians associated with Kaiser Permanente in northern and southern California agreed to stop writing prescriptions for valdecoxib (Bextra).2 The Kaiser notice regarding the moratorium cited “significant concerns raised about the cardiovascular safety of Bextra…” and “compelling” evidence about cardiovascular safety risks and “modest benefit” for valdecoxib. On April 7, 2005, the FDA asked the manufacturer of valdecoxib to voluntarily withdraw the drug from the market.3 Steve Galson, acting director of the FDA Center for Drug Evaluation & Research, said that there was “no added advantage, and a special risk,” in the higher rate of adverse skin reactions with Bextra, and “the cardiovascular risks of these drugs are what we consider a class effect.”4 On August 1, 2005, the FDA approved celecoxib (Celebrex), the sole remaining cyclooxygenase (COX)-2 inhibitor in the U.S. market, for an additional (sixth) indication for the relief of signs and symptoms associated with ankylosing spondylitis, but with the label warning of increased risk of serious cardiovascular thrombotic events, myocardial infarction, and stroke.5 The class effect of COX-2 inhibitors on increased cardiovascular risk remains controversial. Using administrative claims data from Kaiser Permanente in California, Graham et al. evaluated 2,302,029 person-years of follow-up for users of nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) and celecoxib and rofecoxib.6 There were 8,143 cases of serious coronary heart disease, of which 2,210 (27.1%) were fatal. The multivariate adjusted odds ratio for the incidence of serious coronary heart disease for rofecoxib (all doses) was 1.59 (95% confidence interval [CI], 1.10-2.32; P = 0.015) compared with celecoxib. The authors concluded that rofecoxib use increases the risk of serious coronary heart disease compared with celecoxib use, and naproxen use does not protect against serious coronary heart disease. The evidence produced by the scrutiny of administrative claims data and results from clinical trials suggests that the COX-2 inhibitors do not have equivalent (class) effects in either protection from adverse gastrointestinal (GI) effects or in their adverse cardiovascular effects. It is important to remember that the FDA (a) allowed the manufacturer to add to the label for rofecoxib (Vioxx) the results of the VIGOR (Vioxx Gastrointestinal Outcomes Research) trial that suggested a GI protective effect compared with naproxen, (b) never permitted a claim of less GI harm for any COX-2 inhibitor except rofecoxib, and (c) rejected the request from the manufacturer to 590 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 drop the warning of possible GI adverse effects from the label for celecoxib.7 Comparison of the adverse event data across all NSAIDs and COX-2 inhibitors shows an incidence of myocardial infarction of <1% for all NSAIDs except tolmetin (3% to 9%), rofecoxib (3.5% to 10%), valdecoxib (1.6% to 2.1%), celecoxib (<2%), and meloxicam (Mobic, <2%).8 The results of the CLASS (Celecoxib Long-term Arthritis Safety Study) trial were published in September 2000 by Silverstein et al., including the finding that the annualized incidence rates of upper GI ulcer complications were 0.76% for celecoxib versus 1.45% for ibuprofen or diclofenac (P = 0.09).9 Based largely upon the results of the CLASS trial, the Celebrex label was revised in June 2002 to state, “Differences in the incidence of complicated ulcers between Celebrex and the combined group of ibuprofen and diclofenac were not statistically significant.” The FDA written statement elaborated that Celebrex “did not show a safety advantage in upper GI events for Celebrex compared to either ibuprofen or diclofenac.”10 Under the Precautions, the Celebrex label states, “Because of its lack of platelet effects, Celebrex is not a substitute for aspirin for cardiovascular prophylaxis.” A wealth of information regarding the advisories, the request for market withdrawal of valdecoxib on April 7, 2005, and the history of label changes for the COX-2 inhibitors can be found on the FDA Web site.11 On April 7, 2005, the FDA asked manufacturers of all marketed prescription NSAIDs, including celecoxib, to revise the labeling (package insert) for their products to include a boxed warning, highlighting (a) the potential for increased risk of cardiovascular events and (b) the serious, potential life-threatening GI bleeding associated with their use.11 The revised celecoxib label also included the warning about longterm use.12 The additional warning for celecoxib was based, in part, on results of the Adenoma Prevention with Celecoxib (APC) trial in which, over a range of 2.8 to 3.1 years of follow-up, the composite end point of death from cardiovascular causes (myocardial infarction, stroke, or heart failure) was reached in 7 of 679 patients in the placebo group (1.0%), compared with 16 of 685 patients who received 200 mg celecoxib twice daily (2.3%; hazard ratio [HR], 2.3; 95% CI, 0.9-5.5), and 23 of 671 patients who received 400 mg celecoxib twice daily (3.4%; HR 3.4; 95% CI, 1.4-7.8).13 The FDA Alert on April 7, 2005, for celecoxib specifically states, “Celebrex has been associated with an increased risk of serious adverse cardiovascular (CV) events in a long-term placebo controlled trial. Based on the currently available data, FDA has concluded that an increased risk of serious adverse CV events appears to be a class effect of NSAIDs (excluding aspirin).”14 Aside from the safety risks associated with the use of COX-2 inhibitors, they are no more efficacious than nonselective NSAIDs. Simon et al. found that celecoxib and naproxen are equally effective in treating the symptoms of rheumatoid arthritis (RA).15 Evidence-based evaluations such as that produced by www.amcp.org Editorial Subjects—In This Issue and in Previous Issues the National Institute for Clinical Excellence conclude that COX-2 selective inhibitors “are more effective than placebo and of equivalent efficacy to other NSAIDs in their ability to reduce pain (resting and active) and to improve physical and global function in both OA [osteoarthritis] and RA [rheumatoid arthritis] patients.”16 So, the COX-2 inhibitors appear to share an increased cardiovascular risk with the NSAIDs (except aspirin), which the FDA defines as a “class effect,” and they are no more efficacious than nonselective NSAIDs. The COX-2 inhibitors would suffer a third blow if they proved to be no more protective for GI bleed events compared with the nonselective NSAIDs. In real-world use, Stockl, Cyprien, and Chang, in this issue of JMCP, found no difference in the risk of GI bleed for the COX-2 inhibitors compared with NSAIDs in their study of 50 months of administrative claims data from a large managed care organization (MCO).17 There were 375 cases of GI bleed among 19,201 follow-up years for COX-2 inhibitor users (19.5 cases per 1,000 person-years) versus 228 cases of GI bleed among 12,680 follow-up years for NSAID users (18.0 cases per 1,000 personyears). The risk of GI bleed was not significantly different for COX-2 inhibitor users compared with nonselective NSAID users (relative risk [RR] 1.07; 95% CI, 0.90-1.26). Even among high-risk patients, a significant reduction in the risk of a GI bleed was not found among COX-2 inhibitor users (RR 0.995; 95% CI, 0.84-1.19). Examination of the study methods and the characteristics of the MCO that was the subject of the research by Stockl et al. is warranted. First, this MCO operated a prior authorization (PA) requirement for COX-2 inhibitors that affected an estimated two thirds of the enrolled population. Therefore, one would expect the COX-2 inhibitor users in this MCO to have a higher risk of GI bleed compared with a population of health plan beneficiaries not subject to PA requirements for COX-2 inhibitors. However, this is not evident in the data presented by Stockl et al. in their Table 3, in which 57.4% of the COX-2 inhibitor users were categorized at high-risk of GI bleed compared with 58.2% of the NSAID users. The influence of this PA program on the study results is not obvious, but it seems plausible that the population of COX-2 inhibitor users in this MCO would be more likely to experience a GI bleed than the general population of COX-2 inhibitor users. Therefore, the finding of no difference in the rate of GI bleed for COX-2 inhibitor users versus NSAID users seems stronger as a result of the coincident operation of a PA program for COX-2 inhibitors in this MCO that may have led to selection bias in the group of COX-2 inhibitor users. Second, the prevalence of COX-2 inhibitor and NSAID use is of interest. In this MCO, 1,038,437 members received one or more prescriptions for either an NSAID or COX-2 inhibitor over the 44-month identification period of the study. Assuming 5% turnover among MCO members per year, the prevalence of www.amcp.org use of NSAID or COX-2 inhibitors was 31.7% (1,308,437 users divided by 3.28 million MCO members) during this 44-month study period. The proportion of COX-2 inhibitor users to total NSAID/COX-2 inhibitor users was 6.2% after exclusion for discontinuous enrollment, age younger than 18 years, and prior use of either a COX-2 inhibitor or NSAID in the 6-month preperiod. Third, Stockl et al. did not use the type of health plan (Medicare + Choice [now Medicare Advantage] or commercial) in their propensity matching, with the result that a smaller proportion (53.1%) of COX-2 inhibitor users were enrolled in Medicare + Choice compared with NSAID users (55.7%). While this 2.6% absolute difference (5% relative difference) is statistically significant, its practical significance is uncertain, particularly because there was no difference in the mean age for COX-2 inhibitor users (63.4 years ± 16.3) versus NSAID users (63.3 years ± 16.3). However, most of the Medicare + Choice plans had annual maximum benefits of $1,000 or less; e.g., in the Texas health plans of this MCO, the annual maximum benefit was in the range of $500 to $600, and in 2002, brand drugs such as the COX-2 inhibitors were excluded from coverage entirely.18 The potential value of the use of COX-2 inhibitors withers further under economic evaluation. For administrative claims data for dates of service in 2003-2004, Gleason et al. found that denial of coverage of COX-2 inhibitors by a PA program was associated with savings in pharmacy costs with no increase in total medical costs.19 The prevalence of use of COX-2 inhibitors was 4.2% in the pre-PA period and 1% in the post-PA period. The direct COX-2 inhibitor cost savings were $1.35 per member per month (PMPM) or $1.00 PMPM after subtraction of administrative costs and lost drug manufacturer rebates. In addition to the direct cost savings, there was no evidence of GI problems after coverage of COX-2 inhibitors ended in 737 patients who were using COX-2 inhibitors before institution of a PA program. In the 3 months after PA program implementation (for January 1, 2003, through March 31, 2003), 620 (84.1%) of 737 members had no claims for a COX-2 inhibitor, and during this period, their pharmacy costs initially declined by 40% (P <0.001) and their medical costs declined by 18.7% (P<0.001) and remained significantly lower. Among a subgroup of 156 members (21.2%) who tried to fill a COX-2 inhibitor prescription but were denied coverage, pharmacy costs initially declined by 48.1% (P <0.001) and medical costs declined by 10.3% (P <0.001). These findings are not surprising in light of the fact that 68% of patients who attempted to obtain a COX-2 inhibitor were denied coverage because they did not meet the PA criteria for patients at elevated risk for a GI bleed, nearly identical to the 65% of COX-2 inhibitor users determined not to be high risk in the work reported in 2003 (for 2000-2001 data) by Cox, Motheral, Frisse, et al.20 Spiegel et al. performed a cost-utility analysis and found that Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 591 Editorial Subjects—In This Issue and in Previous Issues use of a COX-2 inhibitor instead of a nonselective NSAID in average-risk patients had an incremental cost of $275,809 to gain 1 additional quality-adjusted life-year (QALY).21 Adding the potential for adverse cardiovascular events to the pharmacoeconomic analysis pushed the incremental cost to nearly $400,000 per QALY. The pharmacoeconomic considerations around the use of COX-2 inhibitors versus other NSAIDs for the relief of joint pain, OA, RA, and other pain management needs have contributed to the amount of attention that the COX-2 inhibitors have received in prior issues of JMCP. As pointed out by Ortiz, no drug class has received more attention in articles published in JMCP.22 In calendar year 2001, celecoxib and rofecoxib were the number 5 and number 8 drugs by total expenditure dispensed in community pharmacies,23 propelled, in part, by the large spending on direct-to-consumer advertising (DTCA). In the first quarter of 2001, celecoxib was the second most heavily advertised drug in the United States, and rofecoxib was the third most heavily advertised drug.24 For all of 2001, the midpoint of the study period in Stockl et al., roefecoxib and celecoxib were the top 2 drugs promoted by DTCA, accounting for $135 million and $130.4 million, respectively.25 Celecoxib was the second fastest selling drug in U.S. history. In its first 13 weeks on the market in early 1999, celecoxib had 2.5 million new prescriptions.26 Bull, Conell, and Campen found, in their analysis of data from 1999, that physician education and use of a clinical practice guideline in a large health maintenance organization were associated with a 5-fold difference in the utilization of COX-2 inhibitors among patients determined to be in the highest-risk decile versus patients in the lowest-risk decile, 8.3% versus 1.5%, respectively, and the authors speculated that DTCA probably influenced patient requests for COX-2 inhibitors.27 So, the available evidence indicates that COX-2 inhibitors (a) are expensive, (b) have no pain relief benefit over NSAIDs, (c) offer little if any GI protection in the general population of NSAID users, and (d) may cause increased risk of cardiovascular events. Just how much value does the sole COX-2 inhibitor remaining on the U.S. market have in patients with a risk of a GI bleed? Chan et al. found, in a randomized controlled trial, that 7 of 144 patients (4.9%) who received 200 mg of celecoxib twice daily plus placebo experienced recurrent GI bleeding versus 9 of 143 patients (6.3%) who received 75 mg diclofenac twice daily plus 20 mg omeprazole daily for 6 months. All patients had arthritis treated by NSAIDs, had suffered a GI bleed, and were randomly assigned after ulcer healing and testing negative for Helicobacter pylori infection. This absolute difference of 1.5% was not statistically significant.28 It is also important to note that this study used an endoscopic end point rather than clinical ulcer disease. The minimum 3 mm size of the endoscopic ulcer would appear to be practically significant, but there is no convincing evidence that endoscopic ulcers are a reasonable surrogate for 592 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 clinical ulcer disease or for ulcer complications.29 So, the study by Chan et al. suggests that omeprazole 20 mg per day plus diclofenac 75 mg twice daily is comparable to celecoxib 200 mg twice daily in the incidence of recurrent GI bleed in patients at high risk for GI bleed, and this study attempted to account for the 2 potentially confounding variables in studies of COX-2 inhibitors: (a) aspirin use, which carries a risk of complications of 1.5% to 2.5% per year,30 and (b) status of infection with Helicobacter pylori.31 Another way to view the data is that, for the many patients currently on omeprazole, celecoxib offers no additional GI bleed protection compared with diclofenac. If GI bleeding is a significant risk in a given patient, and we suspend our concerns for cardiovascular death from COX-2 inhibitor use, how do the costs compare? At current discounted drug prices, the cost of celecoxib 200 mg twice daily is $5.29 per day of therapy versus $1.30 per day for diclofenac 75 mg twice daily plus omeprazole 20 mg (over-the-counter) once daily.32 Therefore, the diclofenac + omeprazole combination has a drug cost 75% lower than the COX-2 inhibitor alternative therapy in patients at high risk of a GI bleed. Even non-math majors can appreciate that, for patients at high risk of a GI bleed, it is possible to treat 4 patients with diclofenac + omeprazole for the same drug cost to treat 1 patient with celecoxib. Frederic R. Curtiss, PhD, RPh, CEBS Editor-in-Chief [email protected] REFERENCES 1. U.S. Food and Drug Administration. FDA issues public health advisory on Vioxx as its manufacturer voluntarily withdraws the product. P04-95, September 30, 2004. Available at: http://www.fda.gov/bbs/topics/news/2004/ NEW01122.html. Accessed August 1, 2005. 2. Bextra suspended by Kaiser; Celebrex Alzheimer’s data under scrutiny. Green Sheet. February 14, 2005:1, 2. 3. U.S. Food and Drug Administration. Alert for healthcare professionals— valdecoxib (marketed as Bextra)—April 7, 2005. Available at: http://www.fda.gov/cder/drug/InfoSheets/HCP/valdecoxibHCP.pdf. Accessed July 27, 2005. 4. Mathews AW, Hensley S. FDA stiffens painkiller warnings, pushing Pfizer to suspend Bextra. Wall Street Journal. April 8, 2005:A1, A4. 5. FDA approves new use for Pfizer’s Celebrex; finalizes prescribing information, including expected warnings, for all six approved uses. Available at: http://www.pfizer.com/pfizer/are/news_releases/2005pr/mn_2005_0801.jsp Accessed August 2, 2005. 6. Graham DJ, Capen D, Hui R, et al. Risk of acute myocardial infarction and sudden cardiac death in patients treated with cyclo-oxygenase 2 selective and non-selective non-steroidal anti-inflammatory drugs: nested case-control study. Lancet. 2005;365(9458):475-81. 7. Harris G, Carroll J. Panel approves Merck’s claims for arthritis drug. Wall Street Journal. February 9, 2001:B10. 8. Non-steroidal anti-inflammatory agents, NSAIDs adverse reactions (%). Facts and Comparisons, CliniSphere Version (ISBN 1-57439-036-8). St. Louis, MO: Wolters Kluwer Health, Inc.; May 2005. 9. Silverstein FE, Faich G, Goldstein JL, et al. GI toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. JAMA. 2000;284:1247-55. www.amcp.org Editorial Subjects—In This Issue and in Previous Issues 10. Harris G. Study finds Celebrex as likely to cause ulcers as older drugs. Wall Street Journal. June 10, 2002:B3. 22. Ortiz E. Market withdrawal of Vioxx: is it time to rethink the use of COX-2 inhibitors? J Manag Care Pharm. 2004;10(6):551-54. 11. U.S. Food and Drug Administration. COX-2 selective (includes Bextra, Celebrex, and Vioxx) and non-selective non-steroidal anti-inflammatory drugs (NSAIDs). Available at: http://www.fda.gov/cder/drug/infopage/cox2/ default.htm. Accessed July 16, 2005. 23. Curtiss FR. Cost-effective use of COX-2 drugs and NSAIDs. J Manag Care Pharm. 2002;8(4):295-96. 12. U.S. Food and Drug Administration. FDA statement on halting of a clinical trial of the COX-2 inhibitor Celebrex. Available at: http://www.fda.gov/bbs/ topics/news/2004/NEW01144.html. Accessed August 1, 2005. 13. Solomon SD, McMurray JJV, Pfeffer MA, et al. for the Adenoma Prevention with Celecoxib (APC) study investigators. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med. 2005;352:1071-80. 14. U.S. Food and Drug Administration. Alert for healthcare professionals— celecoxib (marketed as Celebrex)—April 7, 2005. Available at: http://www.fda.gov/cder/drug/infopage/celebrex/celebrex-hcp.pdf. Accessed July 27, 2005. 15. Simon LS, Weaver AL, Graham DY, et al. Anti-inflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis. A randomized controlled trial. JAMA. 1999;282:1921-28. 16. National Institute for Clinical Excellence. Guidance on the use of cyclooxygenase (Cox) II selective inhibitors, celecoxib, rofecoxib, meloxicam and etodolac for osteoarthritis and rheumatoid arthritis. Technology Appraisal Guidance–no. 27. Available at: http://www.nice.org.uk/pdf/ coxiifullguidance.pdf. Accessed August 2, 2005. 17. Stockl K, Cyprien L, Chang E. Gastrointestinal bleeding rates among managed care patients newly started on COX-2 inhibitors or nonselective NSAIDs. J Manag Care Pharm. 2005;11(7):550-58. 24. Burton TM, Harris G. Note of caution—study raises specter of cardiovascular risk for hot arthritis pills. Vioxx and Celebrex marketers dispute the research, sought to downplay it. Wall Street Journal. August 22, 2001:A1. 25. Burton TM. Backlash is brewing among companies who believe flashy ads drive up costs. Wall Street Journal. March 13, 2001:B1. 26. Sharpe R. Several deaths show a link to Celebrex. Wall Street Journal. April 20, 1999:B6. 27. Bull SA, Conell C, Dampen DH. Relationship of clinical factors to the use of COX-2 selective NSAIDs within an arthritis population in a large HMO. J Manag Care Pharm. 2002;8(4):252-58. 28. Chan FK, Hung LC, Suen BY, et al. Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis. N Engl J Med. 2002;347:2104-10. 29. Sung JY, Lau JY, Chan FK, Graham DY, How often are endoscopic ulcers in NSAID trials diagnosed as actual ulcers by experienced endoscopists? [abstract]. Gastroenterology 2001;120(suppl)1A-597. 30. Derry S, Loke YK. Risk of gastrointestinal haemorrhage with long-term use of aspirin: meta-analysis. BMJ. 2000;321:1183-87. 31. Graham DY, NSAIDs, Helicobacter pylori, and Pandora’s box. N Engl J Med. 2002;347:2162-63. 32. Prices obtained from www.drugstore.com; e.g., Celebrex price available at: http://www.drugstore.com/pharmacy/prices/drugprice.asp?ndc=00025152531 &trx=1Z5006. Accessed July 31, 2005. 18. Personal communication with the pharmacy manager for the Texas health plans of PacifiCare on August 2, 2005. 19. Gleason PP, Williams C, Hrdy S, et al. Medical and pharmacy expenditures after implementation of a cyclooxygenase-2 inhibitor prior authorization program. Pharmacotherapy. 2005;25(7):924-34. 20. Cox ER, Motherall B, Frisse M, et al. Prescribing COX-2s for patients new to cyclo-oxygenase inhibition therapy. Am J Manag Care. 2003;9:735-42. 21. Spiegel BM, Targownik L, Dulai G, Gralnek IM. The cost-effectiveness of cyclooxygenase-2 selective inhibitors in the management of chronic arthritis. Ann Intern Med. 2003;138(10):795-806. www.amcp.org Letters to the Editor JMCP welcomes letters that serve to clarify subjects published in previous issues of the Journal or regarding subject matter of interest to managed care pharmacists. Letters in JMCP are not peer reviewed but are subjected to editorial review. Letters should be prepared in a word processing program, preferably Microsoft Word, and submitted electronically at jmcp.msubmit.net. See www.amcp.org for details. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 593 LETTER ■■ Standardization Is Necessary in the Methods to Assess the Value of Electronic Prescribing Systems In the May 2005 issue of JMCP, we reported that an electronic prescribing system with integrated clinical decision support was effective at improving prescribing behavior and lowering prescription costs.1 In contrast, Ross et al. reported in the June 2005 issue of JMCP that a different electronic prescribing system (without clinical decision support capabilities) had no impact on generic drug utilization or formulary compliance.2 When designing our study, we also considered assessing changes in generic prescribing rates as a surrogate marker for cost savings but, instead, we decided to focus on the actual differences in prescription costs. Since publishing our results, we have been asked many times about the impact of our system on generic drug utilization and have been involved in subsequent study design discussions in which customers wanted to focus primarily on this outcome measure. Given the ongoing interest in evaluating the impact of electronic prescribing systems, we decided to assess the changes in generic drug utilization observed in our study by conducting a post hoc analysis of the new prescriptions dataset. These results (1) highlight a potential limitation of prior studies assessing electronic prescribing systems, (2) confirm that an electronic prescribing system with integrated clinical decision support can increase generic prescribing rates, and (3) demonstrate how focusing only on savings related to increasing generic drug utilization may substantially underestimate the total savings associated with electronic prescribing. Some studies attempting to quantify the impact of electronic prescribing systems have compared outcomes for physicians using the system with a group of traditional prescribers over a single time frame.2,3 For example, Ross et al. retrospectively compared physicians using an electronic prescribing system with a group of closely matched controls and found no difference in generic drug utilization during a 12-month evaluation period (37.3% versus 36.9%). While these studies can provide useful information, the results are somewhat difficult to interpret without also knowing if the groups had similar prescribing practices at baseline (i.e., prior to implementation of the electronic prescribing system). Potential pitfalls related to comparing groups from a single time period can be demonstrated by examining the generic drug utilization data from our study. During our 12-month follow-up period, from June 2002 through May 2003, the generic dispensing rates were 55.2% in the electronic prescribing group and 48.7% in the control group. If we had looked only at this time period, we would have incorrectly assumed that the electronic prescribing system resulted in a 6.5% difference in generic drug utilization. When baseline generic drug utilization is considered, the results show that the generic dispensing rate increased from 50.8% to 55.2% (8.7% relative increase) in the intervention TABLE 1 Savings on Brand-Name and Generic Medications During the 12-Month Follow-up Period (June 2002-May 2003)1 New Prescriptions Data Script Volume Average Cost ($) Expenditures ($) Generic medications Intervention group 14,712 (55.2%) 16.79 247,014 Control (expected) values* 13,444 (51.3%) 16.35 219,809 Difference (1,268) (0.44) (27,205) 845,594 Brand-name medications Intervention group 11,962 (44.8%) 70.69 Control (expected) values* 12,786 (48.7%) 76.55 978,768 Difference 824 5.86 133,174 Script Count Savings/(Loss) Associated With Each Script Total Savings/ (Loss) Decreased utilization of brand-name medications associated with switching to a generic alternative 824 76.55 63,077 Increased utilization of generic medications associated with switching from a brand-name medication (824) (16.79) (13,835) Increased utilization of generic medications beyond those related to brand-to-generic conversions (444) (16.79) (7,455) Decreased cost of brandname medications 11,962 5.86 70,097 Increased cost of generic medications 13,444 (0.44) (5,915) Contribution to Savings Utilization† Cost Total savings 105,969 * Control values in Table 1 use the changes observed in the control group to represent what would have been expected in the intervention group. Because the intervention group (n = 19) and the control group (n = 19) were closely matched prior to implementation of the electronic prescribing system, the impact of the system can be determined by comparing the intervention group’s expected values with their actual values. † Based on changes observed in the control group, the intervention group had 824 fewer brand-name prescriptions than expected and 1,268 more generic drug prescriptions than expected. The savings calculations in the table assume that 824 of the excess generic prescriptions were related to brand-to-generic conversions; however, it is possible that some brand-name medications were discontinued rather than switched to a generic alternative. This assumption does not affect the total savings amount but, if incorrect, would further decrease the proportion of savings associated with brand-to-generic switches. group and from 48.2% to 48.7% (1.0% relative increase) in the control group. To correct for the difference in baseline generic dispensing rates, the data can be adjusted using the methodology 594 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Letter proposed by the Centers for Medicare and Medicaid Services physician group demonstration project.4 This methodology assumes that the change observed in the control group represents normal variation (i.e., what would have been expected in the intervention group if they had not started using the electronic prescribing system). In this case, the 1% relative change observed in the control group is applied to the intervention group’s baseline value of 50.8% to achieve an expected generic dispensing rate of 51.3% (i.e., 101% of the 50.8% baseline generic dispensing rate). The difference between the intervention group’s expected generic dispensing rate (51.3%) and their actual generic dispensing rate (55.2%) represents a net absolute increase of 3.9 points. Although these results confirm that the system was effective at increasing generic prescribing, as confirmed by the generic dispensing ratio, it is important to recognize that the savings related to generic drug utilization account for less than half of the total savings. As demonstrated in Table 1, the 3.9% higher generic dispensing rate in the intervention group translates into 824 fewer brand-name prescriptions being dispensed during the study period. While each avoided brand-name prescription saved an average of $76.55 (total savings: $63,077), these savings were partially offset by an increased utilization of generic medications (total cost: $21,290); therefore, the final savings associated with increased generic drug utilization was $41,787. To estimate the final savings, however, differences in the average cost of brand-name and generic medications should also be examined. Our data suggest that the system was not only effective at decreasing utilization of brand-name prescriptions, but it also had a significant impact on the cost of each remaining brand-name prescription. The average cost of brand-name prescriptions was $5.86 (7.7%) lower in the intervention group; therefore, interventions that focused on improving prescribing behavior within these drug categories saved an additional $70,097. These savings were most likely related to electronic messages that focused on improving brand-to-brand conversions or helping clinicians choose the most cost-effective dosing regimens. These cost-related savings on brand-name medications were partially offset by a $0.44 higher cost for each generic prescription (total cost: $5,915); therefore, the total savings attributed to price differences for both brand and generic drug claims was $64,182. Combining the savings estimates from interventions affecting utilization ($41,787) and price ($64,182) results in total savings of $105,969 (8.8%), or an average of $465 per prescriber per month. Because all calculations in the post hoc analysis were based on data adjusted for differences in per-member-permonth (PMPM) utilization, these savings are almost identical to the savings that we originally estimated based on the PMPM costs for new prescriptions.1 We hope that these results will encourage future researchers to assess actual savings related to differences in the cost and www.amcp.org utilization of brand and generic medications rather than simply to focus on changes in generic utilization. We also encourage future researchers to identify closely matched controls based on data gathered prior to the implementation of the electronic prescribing system because we believe that comparing 2 groups of clinicians during an isolated time frame may result in spurious conclusions. S. Troy McMullin, PharmD Clinical Decision Support Purkinje 10426 Baur Blvd. St. Louis, MO 63110 [email protected] DISCLOSURE The author is an employee of Purkinje, distributor of the electronic prescribing system that was the subject of this study. REFERENCES 1. McMullin ST, Lonergan TP, Rynearson CS. 12-month drug cost savings related to use of an electronic prescribing system with integrated decision support in primary care. J Manag Care Pharm. 2005;11(4):422-32. 2. Ross SM, Papshev D, Murphy EL, Sternberg DJ, Taylor J, Barg R. Effects of electronic prescribing on formulary compliance and generic drug utilization in the ambulatory care setting: a retrospective analysis of administrative claims data. J Manag Care Pharm. 2005;11(5):410-15. 3. Cap, Gemini, Ernst & Young. Allscripts: Touchscript medication management system financial analysis on pharmacy risk pools. Available at http://www.allscripts.com. Accessed August 31, 2004. 4. Medicare program: solicitation for proposals for the physician group demonstration project. Fed Regist. 2002;67:61116-29. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 595 ABSTRACTS Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference T he following poster presentations have been prepared for the Academy of Managed Care Pharmacy’s 2005 Educational Conference, October 5-8, 2005, in Nashville, Tennessee. Poster presentations are selected by the Program Planning Committee from proposals that are submitted to AMCP. Authors of posters are responsible for the accuracy and completeness of data presented in the posters and in the abstracts published here. For more information about the studies described below, please contact the corresponding authors, indicated by an asterisk (*), whose addresses are listed in full. The names of individuals who are scheduled to present at the meeting are underlined. ■■ ANALYSIS OF THE CLINICAL AND ECONOMIC IMPACT OF A STEP-THERAPY SEQUENCE ON NONSEDATING ANTIHISTAMINE FAILURE RATES Seifert RD*, Sturm LL. University of Minnesota, College of Pharmacy, 366 Kirby Plaza, 1208 Kirby Dr., Duluth, MN 55812-3095 INTRODUCTION: The purpose of this study is to determine baseline failure rates for secondary nonsedating antihistamines (NSAs) and the clinical and economic outcomes of a step-therapy sequence after patients have received prior treatment with overthe-counter (OTC) loratadine. METHODS: To establish a baseline, a retrospective study of patients who failed OTC loratadine was conducted, reviewing physician prescribing habits for subsequent therapy with secondary NSAs or other allergy therapies. Subsequently, a steptherapy sequence was implemented in which patients who failed on OTC loratadine were required to try and fail intranasal corticosteroids and intranasal antihistamines before receiving secondary NSA therapy. These prescribing methods were compared for clinical outcomes and economic impact in a managed care setting. RESULTS: Prior retrospective analysis demonstrated that 1,713 members received at least one prescription for OTC loratadine from January 1, 2004, through March 31, 2005. The goal of the step therapy was to reduce the number of prior authorizations (PAs) by requiring failure with an alternative therapeutic class of agents. Of those members, 349 required PA to obtain approval for use of an alternative NSA. The estimated potential savings, with a 50% reduction in failure rates, would amount to approximately $30,000 per year in this population. CONCLUSIONS: Several published clinical studies have demonstrated that there is a high secondary failure rate when using 596 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 one NSA after another NSA has already been tried. Implementing step therapy (failure of OTC loratadine and additional failure of a nasal antihistamine [Astelin] and a formulary corticosteroid) before applying secondary NSA therapy would provide potential cost savings in a managed care population. ■■ ANEMIA TREATMENT COSTS IN CANCER PATIENTS TREATED WITH EPOETIN ALFA OR DARBEPOETIN ALFA IN A MANAGED CARE POPULATION Harris HM, DelAguila M, Lopez JM*, McKenzie RS, Piech CT. Ortho Biotech Clinical Affairs Outcomes, LLC, 430 Route 22 East, Bridgewater, NJ 08807 INTRODUCTION: Anemia management with epoetin alfa (EPO) or darbepoetin alfa (DARB) is a common occurrence in cancer patients in managed care populations. This study compares anemia-related costs in cancer patients receiving EPO and DARB. METHODS: Administrative medical, lab, and pharmacy claims from more than 30 diverse managed care plans were examined. Cancer patients aged 18 years and older with at least 1 medical claim for EPO or DARB from January 2003 through February 2004 were identified. Patients continuously eligible for benefits for at least 3 months prior to and 4 months after the date of the initial claim were included. Only patients new to therapy (no EPO or DARB claims during the 3 months prior to initiating therapy) and those with at least 1 additional medical claim for the index agent were included in the study. Costs were evaluated on a per-patient-month (PPM) basis, for patients with at least 28 days of therapy. RESULTS: 6,584 cancer patients (EPO 4,535; DARB 2,049) were identified. The EPO group was older, had a higher proportion of men, and had higher baseline rates of cardiac, respiratory, digestive, and neurologic comorbidities. A higher proportion of EPO patients was initiated in the hospital outpatient setting than DARB patients (EPO 21%, DARB 6%). Treatment duration was similar between the groups (EPO 77 days, DARB 78 days), with a higher number of injections of erythropoietic agent in the EPO group (EPO 8.4, DARB 6.5). Transfusion rates were similar between groups (EPO 6.6%, DARB 5.9%). Compared with DARB patients, EPO patients had lower erythropoietic agent drug cost (EPO $2,057 vs. DARB $2,462, P < .0001). CONCLUSIONS: Despite lower rates of baseline comorbidities, similar treatment duration and fewer injections, the DARB group incurred a 20% greater drug cost for amenia treatment compared with EPO in this large managed care population. www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ ASSESSING THE IMPACT OF GLYCEMIC CONTROL ON HEALTH RESOURCE UTILIZATION AND COSTS OVER A 1-YEAR PERIOD Gricar J*, Welz JA, Menditto L, Koo Y. 264 E. Broadway, #C402, New York, NY 10002 ■■ BUDGET IMPACT MODEL FOR ERLOTINIB (TARCEVA) IN ADVANCED NON–SMALL-CELL LUNG CANCER Ramsey SD*, Clarke L, Kamath TV, Veenstra DL. Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N, M2-B230, Seattle, WA 98109 INTRODUCTION: The relationship between glycemic control and health resource utilization and costs was evaluated in a cohort of managed care members with diabetes over a 1-year period. METHODS: A retrospective analysis was conducted on a nationally representative administrative claims database of managed care enrollees aged 18 to 64 years between 2000 and 2002. Patients with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code diagnosis of diabetes who were continuously enrolled for at least 365 days of follow-up were included. Glycemic control was measured by using the average score of ≥2 glycosylated hemoglobin (A1c) tests. Selected comorbidities were controlled for using multiple logistic regression techniques. Resource utilization (hospitalizations, emergency department (ED) visits, office visits, pharmacy, and laboratory tests) was assessed in the entire study cohort as well as in subpopulations with health care encounters for 7 diabetes-related conditions (cardiovascular disease, hypoglycemia, nephropathy, neuropathy, retinopathy, embolic stroke, and transient ischemic attack). RESULTS: A total of 20,914 subjects (mean age 52 years) met the inclusion criteria for analysis. Overall, a positive statistical relationship was observed between increasing A1c level and total direct medical cost over a 1-year period. In patients with an A1c level of <7, the average total direct medical cost was $1,418.69; corresponding costs in patients with A1c levels 7 to <8.5%, 8.5% to 10%, and >10% were $1,867.53, $2,382.26, and $2,904.33, respectively. The relationship extended to rates of hospitalizations, ED visits, and office visits, and the number of prescriptions for diabetes-related drugs. Similar trends were evident in the subpopulations, with a positive statistical relationship observed between increasing A1c level and both total direct medical cost and hospitalization rate in all 7 subpopulations. CONCLUSIONS: Tighter glycemic control is associated with lower medical costs during a 1-year period. By promoting ways to improve glycemic control in their members with diabetes, such as optimizing drug therapy and high-risk member diabetes case management, health insurers may benefit from averted medical costs, even in the short term. OBJECTIVE: To develop a tool that can be used by health care plans to assess the budgetary impact of covering erlotinib for treating patients, aged 18-65 years, with stage IIIB/IV non–small-cell lung cancer (NSCLC) who have failed at least 1 prior chemotherapy regimen. METHODS: An Excel-based decision model was developed to assess costs and outcomes for a formulary that includes U.S. Food and Drug Administration-approved and National Comprehensive Cancer Network guideline-recommended treatment options, comparing a formulary with erlotinib versus without erlotinib. The model considers second- and third-line treatments and includes differential outcomes that may influence direct medical costs (grade 3/4 adverse effects of treatment such as neutropenia and anemia). The incidence of advanced NSCLC and adverse effects related to treatment (all agents) are based on the Surveillance, Epidemiology, and End Results Cancer Registry and published results of clinical trials. Drug and treatment costs are obtained from publicly available sources. One-way and multiway uncertainty analyses were used to evaluate the impact of varying assumptions and data values on outcomes. RESULTS: The base case considers a health plan of 500,000 enrollees. Assuming that erlotinib represents 22% of secondline treatments and 50% of third-line, total costs of treating Stage IIIb/IV NSCLC patients over 1 year are $418,710 without erlotinib and $388,226 with erlotinib (difference: $30,484; 90% CI, $8,722-$ 90,170). Erlotinib lowers costs compared with standard chemotherapy through reductions in standard chemotherapy-related infusion costs and costs of managing adverse events. Sensitivity analyses demonstrate that the budget impact is most sensitive to the incidence of NSCLC, the proportion of patients receiving treatment, the proportion receiving erlotinib versus standard agents, the cost of erlotinib, and duration of treatment. Outcomes were relatively insensitive to the incidence of serious erlotinib-related adverse events. CONCLUSION: Adding erlotinib as a second- and third-line treatment option may reduce total health plan costs compared with standard therapy. www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 597 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ CLINICAL AND FINANCIAL IMPACT OF ERYTHROPOIETIN: A STUDY OF PRE–END-STAGE RENAL DISEASE PATIENTS IN A HIGH-RISK SPECIALTY-CASE COORDINATION PROGRAM Fein IA*, Mullany LD, Wells BA, Cumbaa V, McDermott C, Fein AB, Elkins AL. University of Florida School of Public Health at AvMed Health Park, 4300 NW 89th Blvd., Gainesville, FL 32602 OBJECTIVE: To assess the clinical and financial impact of erythropoietin (EPO) in the management of anemic patients with chronic kidney disease (CKD). METHODS: A retrospective cohort study of 4,097 predialysis CKD patients enrolled in a specialty coordination program of the AvMed Health Plans, a not-for-profit health maintenance organization, from January 2000 and December 2003. Administrative databases were used to examine the clinical characteristics and health care costs of the group. Two hundred seventy-one members received EPO, but after those with aplastic anemia or renal transplants were eliminated as well as those undergoing chemotherapy, hemodialysis, or peritoneal dialysis, the final study group had 52 patients with 52 casematched controls. Clinical and demographic characteristics of both the case and control groups were virtually identical to nationally reported data for CKD patients. RESULTS: CKD patients experienced a continuous deterioration of renal function over time (creatinine clearance fell 16.3% over 9 months, P < .0001), accompanied by a progressive 5-fold increase in global per-member-per-month (PMPM) health care expenditures over the 3-year period prior to the initiation of EPO therapy [r2 =.651, P<.0001]. Following EPO therapy, renal function was effectively stabilized, measured by creatinine clearance, and sustained over a mean follow-up time of 9 months. Mean hemoglobin levels fell from 11.52 g/dL (± 1.56 SD) to 9.90 g/dL prior to EPO therapy and were corrected to starting levels (11.52 g/dl, ± 1.44 SD) with the use of EPO. The rise in PMPM expenditures was halted and remained stable over the 9-month mean follow-up period. Health care expenditures were comprehensive, including all drug costs. CONCLUSIONS: Treatment with EPO in anemic patients with predialysis CKD is clinically effective and is associated with a stabilization of both the progression of renal disease and global PMPM costs. ■■ CMS COMPETITIVE DRUG ACQUISITION PROGRAM: SHAPING FUTURE MCO PROCESSES? Baker JJ*, McClard C. The Resource Group, PO Box 70, Pickton, TX 75471 INTRODUCTION: This case study details the Centers for Medicare & Medicaid Services (CMS) Competitive (Drug) Acquisition Program (CAP) processes that become effective January 1, 2006, and explores the potential impact on managed care organization processes. 598 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 METHODS: The primary objective of the case study was 2-fold: to develop a process model of the CMS CAP for Part B drugs as it relates to physician providers, and to explore the potential for those government processes to shape managed care organization (MCO) processes in the near future. Government sources and physician interviews and observations were utilized to create the CAP process model. MCO interviews and relevant survey results were utilized to create a likely future scenario for the MCO process changes. RESULTS: The case study identified sequential processes that are required of physician providers who will participate in the CMS CAP. Although the CAP processes were initially designated as primarily administrative in nature, the physician practice personnel involved professional clinical labor as well as clinical labor in more than 50% of practices interviewed. More than 70% of practices interviewed indicated they planned to adopt the government-mandated process as their primary model. A significant portion of physician practices indicated they intended to request specific changes from MCOs in the next contracting cycle. When the framework for potential changes, constructed utilizing physician responses, was ranked using a 5-point scoring system, 3 particular categories emerged as most likely to impact MCO processes in the near future. CONCLUSIONS: CAP stipulates a series of mandatory processes for participating physician providers. This case study’s model demonstrates that processes utilized by the government model have a significant potential for migrating to MCOs in the near future. Process changes will most likely be triggered by provider requests and/or negotiations when contract expiration dates are pending. The model also demonstrates that the processes utilized are capable of being successfully replicated. ■■ COMPARISON OF HEALTH CARE UTILIZATION AND INCIDENCE OF ADVERSE EVENTS OF LEVETIRACETAM WITH GABAPENTIN IN EPILEPTIC PATIENTS USING A RETROSPECTIVE CLAIMS ANALYSIS Grossman P.* UCB Pharma, 1950 Lake Park Dr., Smyrna, GA 30064 OBJECTIVE: To compare health care utilization and incidence of adverse events (AEs) in epileptic patients initiating levetiracetam (LEV) or gabapentin (GBP). METHODS: A retrospective cohort analysis of epileptic patients was conducted using data from a U.S. claims database. Patients without any LEV or GBP prescription for 6 months prior to therapy initiation were followed for 3 months to 1 year (July 2001-December 2003). GBP patients were matched to LEV patients by clinical characteristics, seizure, and therapy types. Comparison of health care utilizations used Wilcoxon rank-sum tests. Risk of AEs was assessed using Cox proportional hazards models. RESULTS: Treatment groups (n = 816 in each) were comparable: www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference mean age ~39 years, ~63% women, 64% generalized seizures, 65% adjunctive therapy. LEV patients refilled their prescriptions more often than GBP (mean/patient/year: 8.3 vs. 6.0, P<0.001). Utilization was significantly lower in LEV than in GBP for physician office visits (18.1 vs. 20.8, P < 0.01), emergency room visits (8.5 vs. 11.0, P < 0.01), other outpatient visits (35.7 vs. 40.6, P < 0.05), and medications other than antiepileptics (24.4% vs. 36.5, P < 0.001). Diagnostic tests and inpatient services were comparable (not significant) in both groups. Absence of AEs during follow-up was 38% for LEV versus 29% for GBP (P < 0.001). Risk of AEs was significantly lower in LEV than GBP: hazard ratio: 0.76, 95% confidence interval, 0.68-0.86, P < 0.001; median time to first AE: 42 days LEV, 28 days GBP. CONCLUSION: LEV patients showed lower utilization of most common health care services than GBP patients. Rate and risk of AEs were significantly lower in LEV despite a higher refill rate relative to GBP. ■■ COST-EFFECTIVENESS AND BUDGETARY IMPACT OF BIOLOGIC THERAPIES FOR MODERATE-TO-SEVERE PLAQUE PSORIASIS Broder M*, Laouri M, Ding L, Kwon P. Partnership for Health Analytic Research, LLC, 1950 Sawtelle Blvd., Suite 280, Los Angeles, CA 90025 OBJECTIVE: To calculate, from the payer perspective, the relative cost-effectiveness and budget impact of efalizumab and etanercept for psoriasis in a million-member health plan. METHODS: We used information on dosing and treatment-related utilization from product labeling information and subsequent clinical trials for 2 biologic drugs (efalizumab and etanercept) indicated for moderate-to-severe plaque psoriasis to construct a cost-effectiveness model that calculates cost per patient with a successful outcome. Success was defined as achieving a ≥75% improvement in the Psoriasis Area and Severity Index score (PASI 75) after 24 weeks of treatment. Cost of treatment was determined by adding costs in 4 categories: (1) drugs (average wholesale price), (2) administration, (3) monitoring (platelet counts or tuberculosis test), and (4) adverse events. For budget impact, we assumed that a million-member plan would have 2,000 adults with moderate-to-severe plaque psoriasis and 200 would use biologic therapy. To calculate cost, we assumed 100% of etanercept users stepped down their dose from 100 mg/week (weeks 1-12) to 50/mg/week (weeks 13-52) as recommended in the product labeling. RESULTS: Cost per PASI 75 responder was $20,438 for efalizumab and $24,351 for etanercept. Projected annual treatment costs were $3.8 million for efalizumab and $4.7 million for etanercept, assuming all patients step down and continue at 50 mg/week for 40 weeks. Annual treatment costs increased if www.amcp.org patients on etanercept had lower step-down rates: $5.4 million (75%); $6.1 million (50%); and $6.8 million (25%). CONCLUSIONS: Efalizumab is more cost effective than etanercept. The budget impact of etanercept is unpredictable since it is unknown how many patients will step down from 100 mg/week to 50 mg/week. Efalizumab dosing is stable and predictable. Cost-effectiveness and dosing stability may be particular advantages when efficacy and safety are comparable and the ability to accurately predict drug costs is a key consideration. ■■ COST-EFFECTIVENESS OF LINEZOLID VERSUS VANCOMYCIN IN THE TREATMENT OF COMPLICATED SKIN AND SOFT-TISSUE INFECTION DUE TO PROVEN OR SUSPECTED MRSA IN PATIENTS AGED 65 YEARS OR OLDER McCollum M*, Sorensen S, Liu L. University of Colorado at Denver and Health Sciences Center, 4200 East Ninth Ave., Box C-238, Denver, CO 80262 INTRODUCTION: We estimated costs of treating complicated skin and soft tissue infections (cSSTI) with linezolid (LZD) versus vancomycin in a subset of patients aged > 65 years enrolled in the largest clinical trial of cSSTI due to proven or suspected methicillin-resistant staphylo- coccus aureus (MRSA). METHODS: Clinical trial patients aged > 65 years and admitted to U.S. hospitals with cSSTI due to proven or suspected MRSA were included in these analyses. Infection-related costs were estimated by applying nationally representative 2003 per-diem hospital costs for days in medical/surgical, intensive care, or stepdown units. Costs of administering intravenous (IV) therapy were applied to IV treatment duration. Medications were valued at wholesale acquisition cost. Hospitalization costs and total costs (hospital plus outpatient) were estimated. Cure rates were determined from the subset of patients who were clinically evaluable. RESULTS: Of the 717 patients in the clinical trial, 163 were aged > 65 years (87 LZD, 76 vancomycin). No significant clinical or demographic differences at baseline were observed between groups. Average hospitalization costs for patients treated with LZD were $4,511 versus $6,478 for vancomycin (P < 0.001). Average total costs for patients treated with LZD was $6,009 versus $7,329 for vancomycin (P = 0.03). Average length of stay for LZD patients was 6.8 days compared with 10.3 days for patients treated with vancomycin (P<0.001). No significant differences were observed in clinical cure rates between linezolid and vancomycin (LZD = 89%, vancomycin = 81%, P = 0.2). CONCLUSION: Treatment with linezolid (IV/orally) for cSSTI due to proven or suspected MRSA in patients aged >65 years results in lower hospital and total costs compared with vancomycin. Clinical cure rates were similar in both groups. Lower costs for patients treated with linezolid are enabled by transition to oral therapy and earlier hospital discharge. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 599 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ CUMULATIVE EXPOSURE MODELS: CYCLOOXYGENASE-2 INHIBITORS AND CARDIOVASCULAR RISK Shaya FT*, Gu A, Weir MR, Skolasky R. University of Maryland School of Pharmacy, 515 West Lombard St., Room 256, Baltimore, MD 21201 INTRODUCTION: Little is known about the association of length of exposure to cyclooxygenase-2 inhibitors (COX-2s) and the risk of cardiac events. This study determines the impact of extended exposure to COX-2s among high-risk Medicaid patients and is based on a previous propensity-adjusted model that showed no added risk of cardiac events in COX-2 versus nonsteroidal antiinflammatory drug (NSAID) users in this Medicaid population. METHODS: Selecting COX-2 users alone, we analyzed all medical and prescription claims of all continuously enrolled Medicaid patients, with at least 1 prescription for a COX-2 between January 1, 2000, and January 1, 2003, and no such prescriptions in the first 6 months. We used both direct adjustment and propensity score methods and assessed length of exposure to COX-2s as a risk factor for postuse cardiac events, defining risk as a categorical variable (<30, 30-59, 60-89, 90-119, and >120 days), then as a continuous variable (divided by 30). The models are adjusted for age, gender, race, location (urban/suburban/ rural), and clinical risk factors. RESULTS: A total of 1,784 patients used COX-2s, 25% for fewer than 30 days. From the categorical analysis, there are significant increases in the likelihood of a postuse cardiac event given increased use compared with fewer than 30 days of cumulative exposure. From the analysis of exposure to COX-2s as a continuous variable, in the propensity-adjusted model, each 30-day increase in exposure corresponds to a concomitant but nonsignificant 2% increase in risk of cardiac events. For the direct adjusted model, there is a concomitant 5.5% significant increase in postuse cardiac events. CONCLUSION: Among Medicaid COX-2 users, the risk of cardiac events is associated with longer exposure to COX-2s only when exposure is categorized in 30-days increments but not when used as a continuous variable, suggesting a nonlinear relationship between exposure and events. ■■ DEPRESSION COMPLIANCE RETROSPECTIVE CASE ANALYSIS Straub PA.* Precision Healthcare Delivery, 2301 River Rd., Suite 302, Louisville, KY 40206 patients receiving drug therapy for depression during 2003. There were no time limits for treatment on drug therapy. A total of 150 patients who were identified were sent a letter by their physicians and asked to participate in a discussion about their therapy with a registered pharmacist. The patients who responded to the letter were asked to: a) provide demographic information such as age, gender, etc.; b) indicate chronic disease states, number of prescriptions taken each day; c) provide the name of medication used for depression therapy; d) identify a stressful event in the last 12 months; e) rank themselves according to a self-reported medicationtaking behavior scale; f) discuss whether they skip doses due to cost; g) discuss whether they experience embarrassment because of medication; and h) discuss whether they are receiving any counseling or group therapy. RESULTS: There were 76 patients interviewed for this study. Despite treatment recommendations, 77.6% of the patients were noncompliant due to at least one reason, whether it was side effects, cost, or forgetfulness. A patient could have had more than one reason for being noncompliant. The results also showed that there was no correlation between noncompliance and the medication being taken for depression. CONCLUSIONS: It can be concluded that, in this population, patient education regarding the value of their medications was to be encouraged. The results were shared with the physicians to increase their awareness of patients not properly taking their medications; 27.6% of patients were noncompliant due to side effects. A discussion with their physician regarding these unwanted side effects can determine if patients need to switch medications. The study showed that 51.3% of the patients admitted that they either skip doses, split their pills, or asked their physician for samples, so cost was a major issue with compliance. Patientassistance program phone numbers for assistance and other discount card information was given to the staff to distribute to these patients. The other reason given for noncompliance was forgetfulness; 31.6% of patients admitted to being forgetful about taking their medication. Morisky tear-off sheets were placed in patient waiting areas for patients to privately score themselves for medication non-adherence and encourage them to discuss their results with a staff member. OBJECTIVE: To review and document nonspecific, nonbranded drug therapy of patients being treated for depression with issues related to noncompliance. The study was to determine the causes of the noncompliance and provide recommendations for better compliance. METHODS: A managed care database was used to identify 600 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ DOSING DISTRIBUTION PATTERNS AND ASSOCIATED COSTS OF ERYTHROPOIETIC AGENTS IN PATIENTS WITH PREDIALYSIS CHRONIC KIDNEY DISEASE FROM 3 LARGE MANAGED CARE ORGANIZATIONS Barron JJ, Mody SH*, Boudreaux MY, Yang W, Groesbeck M, McKenzie RS. Ortho Biotech Clinical Affairs, LLC, 430 Route 22 East, Bridgewater, NJ 08807 INTRODUCTION: Since few reports exist describing current dosing patterns of epoetin alfa (EPO) and darbepoetin alfa (DARB) in patients with predialysis chronic kidney disease (pCKD), this retrospective, observational study was performed to analyze dosing patterns and associated costs of EPO and DARB from a managed care practice perspective. METHODS: Anemic pCKD patients who were aged ≥18 years, had >2 EPO or DARB claims, and were newly initiated on erythropoietic therapy between the fourth quarter of 2001 and the third quarter of 2004 were identified from medical claims of 3 regionally diverse health plans covering approximately 10.2 million lives. EPO and DARB use was identified via Healthcare Common Procedure Coding System (HCPCS) codes in medical claims with doses calculated using billed units. Dosing frequency, mean weekly dosing, and drug costs (using 2004 wholesale acquisition prices) were calculated for each group. RESULTS: A total of 325 EPO and 163 DARB patients met the inclusion criteria. Mean age (years, EPO 69.5±13.8, DARB 69.6 ± 12.7, P = not significant NS), gender distribution (EPO 45.0% male, DARB 45.0% male, P = [NS]), and prevalence of comorbid conditions were similar between therapeutic groups. Weekly and extended (every [Q] 2 weeks [Ws]) dosing frequency were utilized in patients receiving EPO (QW: 23.4%, Q2W: 34.8%, > Q3W: 29.6%) and DARB (QW: 5.5%, Q2W: 46.0%, > Q3W: 44.5%), with an average interval between treatments of 18.6 ± 15.9 days for EPO and 20.7 ± 8.6 days for DARB patients. The average weighted weekly dose was 8,516 units for EPO and 57 mcg for DARB, which corresponded with estimated mean weekly costs of $100 for EPO and $241 for DARB. Similar dosing patterns and cost differences were observed for patients completing 4, 8, and 12 weeks of therapy. CONCLUSIONS: Extended EPO and DARB dosing (> Q2W) was common among anemic pCKD patients. However, costs associated with these treatments differed between therapies, providing a cost advantage to those patients being administered EPO. www.amcp.org ■■ ECONOMIC BURDEN OF GOUT TO THE EMPLOYER Kleinman NL, Patel P, Brook RA*, Melkonian AK, Smeeding JE, JosephRidge N. The JeSTARx Group, 18 Hirth Dr., Newfoundland, NJ 07435-1710 OBJECTIVE: To determine the economic burden of gout associated with medical costs and work loss from an employer perspective. METHODS: Medical, pharmacy, workers’ compensation (WC), short- and long-term disability (STD, LTD), and sick leave (SL) costs in employees with gout, as identified by an International Classification of Diseases, Ninth Revision (ICD-9) code of 274.xx, were examined in a database consisting of 2001 through 2004 claims, payroll, and demographic data from more than 250,000 employees from multiple large U.S.-based employers. Regression modeling was used to measure the cost differences between employees with gout and employees without gout while controlling for age, job tenure, gender, salary, region, and other factors. RESULTS: Data were available for 1,171 employees with gout and a control group of (247,867) employees without gout. The gout group’s costs (per patient per year) were almost twice as high ($6,871 higher) summed across all direct medical and work loss measures (P < 0.0001). The individual differences in medical and pharmacy costs were $1,401 and $427, respectively (both P < 0.0001). Work absence costs had differences of $697 (WC, P < 0.0001), $358 (STD, P < 0.0001), -$25 (LTD, P < 0.0001), and $307 (SL, P < 0.0001). CONCLUSIONS: The economic impact of gout can be costly to employers not only in terms of direct health care costs but also from potential work loss due to absenteeism. Interventions focused on identifying and managing the underlying cause of gout have the potential to produce significant savings in medical and pharmaceutical costs. ■■ ECONOMIC EVALUATION OF CONTROLLED-RELEASE OXYCODONE (CRO [OXYCONTIN TABLETS]) VERSUS OXYCODONE/ACETAMINOPHEN (OXY/APAP [PERCOCET]) FOR OSTEOARTHRITIS PAIN OF THE HIP OR KNEE Marshall DA*, Strauss ME, Pericak D, Math M, Buitendyk M, Codding C, Kim SS, Torrance GW. Innovus Research Inc., 1016-A Sutton Ave., Burlington, Ontario L7L 6B8, Canada INTRODUCTION: Controlled-release oxycodone (CRO) is efficacious for persistent moderate-to-severe osteoarthritis pain, based on well-controlled trials. Additionally, decision makers require evidence of effectiveness in routine practice and cost-effectiveness compared with standard therapy. METHODS: An open-label, active-controlled, randomized, naturalistic 4-month study of analgesic effectiveness and costeffectiveness of CRO versus oxycodone/acetaminophen (oxy/APAP) was conducted. Outcomes and resource use were Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 601 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference collected by telephone. Effectiveness was measured in 485 patients as the proportion having at least 20% improvement from baseline in Western Ontario and McMaster Universities Osteoarthritis Index pain score. Quality-adjusted life-years (QALYs) were calculated from Health Utilities Index-3 scores. Cost-effectiveness was measured as cost/patient improved and QALYs gained from societal and health care perspectives using generic oxy/APAP (base case). Uncertainty was evaluated using multiple 1-way sensitivity analyses and cost-effectiveness acceptability curves. RESULTS: In the study, 62.2% versus 45.9% (P = 0.0003) of patients improved with CRO and oxy/APAP, respectively. Mean QALYs gained over 4 months with CRO compared with oxy/APAP was 0.0105 (P = 0.1673). Mean societal cost/patient over 4 months was US$6,792 versus US$6,929 (P = 0.3345) for CRO and oxy/APAP, respectively. CRO was both more effective and less costly than oxy/APAP, using the societal perspective (includes costs associated with time loss). Using a health care perspective (excludes costs associated with time loss), costeffectiveness of CRO was US$4,500/patient improved and US$69,856/QALY gained. CONCLUSIONS: From the societal perspective, CRO was both more effective and less costly than oxy/APAP. From the health care perspective, CRO, compared with generic oxy/APAP, fell within the acceptable range of cost-effectiveness if decision makers were willing to pay between US$50,000/QALY and US$100,000/QALY. These findings should be considered in decisions about treating osteoarthritis pain. ■■ EFFECTIVENESS OF A DIRECT-TO-PATIENT SPECIALTY PHARMACY COMPLIANCE PROGRAM Allen JD*, Zabriski S, Brown J. McKesson Specialty, 4343 N. Scottsdale Rd., Suite 150, Scottsdale, AZ 85251 INTRODUCTION: The therapy persistence for patients enrolled in a hepatitis C direct-to-patient specialty pharmacy compliance program was compared with a retail distribution control group in a matched-pair cohort fashion. METHODS: Patients new to hepatitis C therapy who enrolled in a Specialty Pharmacy Compliance Program (n = 279) were matched retrospectively in a 1 to 5 ratio with a control group of patients receiving therapy through a retail distribution channel and analyzed over a 1-year period. The primary end point was the percentage of patient persistence, defined as the percentage of patients in a given month of the study period who had drugs on hand as measured by prescription fill date and days supply. Secondary measures included compliance and length of therapy. Patients were excluded if they had any medication fills from a retail outlet during the study period. RESULTS: Of the 279 patients enrolled, 227 were available for analysis. Beginning at 2 months, and extending each month to 602 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 the end of the evaluation period, the specialty pharmacy cohort had significantly more persistence than the retail comparator group (P < 0.05). Specialty patients were also more compliant at 6 and 9 months and utilized 37% more units than did the control group over the same period. CONCLUSION: A specialty pharmacy compliance program, including relationship marketing, call center support, and distribution of a hepatitis C product, improves patient persistence with therapy. ■■ EFFECTIVENESS OF DEPRESSION TREATMENT PREDICTS SUBSEQUENT HEALTH SERVICES COSTS Simon GE, Khandker RK.* Wyeth Research, 500 Arcola Rd., Collegeville, PA 19426 INTRODUCTION: Numerous cross-sectional studies demonstrate a strong association between depression and use of health services, but few longitudinal studies have examined whether remission of depression is associated with decreased health services costs. METHODS: Pooled data from 7 longitudinal studies of patients beginning depression treatment were used to examine the relationship between clinical outcomes of acute-phase treatment and health services costs over the subsequent 6 months. Clinical outcomes were assessed by structured telephone interviews. Health services costs were assessed using health plan accounting records. RESULTS: Of 1,816 patients entering treatment and meeting criteria for major depressive episode, 29% had persistent major depression 3 to 4 months later, 37% were improved but did not meet criteria for remission, and 34% achieved remission of depression. Those with persistent depression had higher baseline depression scores and higher health services costs before beginning treatment. After adjustment for baseline differences, mean health services costs over the 6 months following acute-phase treatment were $2,106 (95% confidence interval [CI], $1,684$2,545) for those achieving remission, $2333 (95% CI, $1,940$2,754) for those improved but not remitted, and $2,955 (95% CI, $2,452-$3,509) for those with persistent major depression. Average costs for depression treatment (antidepressant prescriptions, outpatient visits, mental health inpatient care) ranged from $431 in the remission group to $599 in the persistent depression group. CONCLUSIONS: Clinical outcome of acute phase depression treatment predicts subsequent health services costs, and persistence of depression is associated with 40% higher costs compared with full remission. The excess costs associated with persistence of depression are nearly twice as great as spending on depression treatment. www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ EFFECTIVENESS OF ROSUVASTATIN COMPARED WITH OTHER STATIN THERAPIES ON TARGET LDL-CHOLESTEROL GOALS IN A USUAL-CARE SETTING Kamat SA*, Wertz DA, Yang GW, Bullano MF, Willey VJ. HealthCore, Inc., 800 Delaware Ave., Fifth Fl., Wilmington, DE 19801 OBJECTIVE: To evaluate current cholesterol management trends in clinical practice, low-density lipoprotein cholesterol (LDL-C) goal attainment in patients treated with rosuvastatin compared with other statins was examined using aggressive therapeutic options suggested in the 2004 National Cholesterol Education Program (NCEP) Report (Grundy et al. Circulation. 2004;110: 227-29). METHODS: Patients newly initiated on rosuvastatin, atorvastatin, simvastatin, pravastatin, lovastatin, or fluvastatin between August 1, 2003, and September 30, 2004, were identified from a Southeastern health plan’s administrative claims with integrated lab results data for this retrospective, longitudinal cohort study. Patients were excluded if they had prior dyslipidemic therapy in the 12-month period preceding their initial statin fill. Patients with at least 1 preinitiation and postinitiation LDL-C level were followed until they switched, supplemented, or discontinued their initial statin. Administrative claims were utilized to assign patients an NCEP risk status and corresponding LDL-C goal. Adjusted LDL-C goal attainment odds ratios were calculated using multivariate regression techniques after controlling for baseline differences between groups. RESULTS: From the identified cohort (N=3,139), patients receiving rosuvastatin were slightly younger and had lower mean doses compared with other statins (11 mg vs. 18-69 mg). LDL-C goal attainment was higher with rosuvastatin compared with other statins (58% vs. 29%-48%). After adjusting for age, gender, preindex LDL-C, NCEP risk status, and therapy duration, significantly (P <0.05) fewer patients achieved their LDL-C goals with atorvastatin (odds ratio [OR] = 0.66; confidence interval [CI], 0.49-0.89), simvastatin (OR = 0.53; CI, 0.380.74), pravastatin (OR = 0.22; CI, 0.15-0.33), fluvastatin (OR = 0.16; CI, 0.09-0.28), and lovastatin (OR = 0.32; CI, 0.220.46) compared with rosuvastatin. Furthermore, dose-stratified analysis revealed LDL-C goal attainment was significantly lower (P < 0.05) with 10, 20, and 40 mg atorvastatin compared with 10 mg rosuvastatin (46%-48% vs. 57%). CONCLUSION: Rosuvastatin patients were significantly more likely to attain their LDL-C goals compared with patients on other statins. These data are among the first to illustrate the effectiveness of rosuvastatin over other statins in a usual-care setting and reinforce findings from randomized controlled trials. www.amcp.org ■■ EFFECTS OF IMPLEMENTING A GENERIC MEDICATION STEP EDIT ON THE UTILIZATION AND COSTS OF ANTIDEPRESSANTS IN A MANAGED CARE ORGANIZATION Dunn JD*, Cannon HE, Burgoyne DS. Intermountain Health Care Health Plans, 4646 West Lake Park Blvd., Suite N3, Salt Lake City, UT 84120 OBJECTIVE: To evaluate the impact on utilization and costs of implementing a generic medication step edit in the antidepressant category in an integrated managed care organization (MCO). This report contains short-term results and an explanation of the methodology undertaken by this MCO. Long-term data will be presented in a follow-up abstract. METHODS: Antidepressants do not significantly differ in their ability to treat depression or in their incidence of adverse events (with some interpatient variability). Initial clinical medication choices should be made based on cost considerations, with specific treatments being prescribed at the discretion of the treating provider. With the availability of multiple generic antidepressant medications, using them first will improve the cost-effectiveness of treatment and lower the cost of treatment for patients and MCOs. Of the selective serotonin reuptake inhibitors (SSRIs) currently on the market, fluoxetine (Prozac), paroxetine (Paxil), and citalopram (Celexa) are available generically. In addition to the SSRIs, 2 other agents are available generically. Bupropion SR (Wellbutrin) is a weak inhibitor of norepinephrine and dopamine uptake while mirtazapine (Remeron) is a serotonin, alpha-adrenergic, and histamine antagonist. Some of the antidepressants are U.S. Food and Drug Administration-labeled for additional indications other than treatment of depression. The vast majority of patients will both tolerate and respond to 1 of these 5 medications. On January 1, 2005, Intermountain Health Care (IHC) Health Plans and the IHC Behavioral Health Clinical Program introduced their GenericStart! Program. Under this program, for new starts, IHC Health Plans covers brand-name antidepressants only after a trial of a generic antidepressant medication (excluding tricyclic antidepressants, or TCAs). New starts are defined as members with no claims history of antidepressant treatment within the previous 6 months. Branded antidepressants were reevaluated for formulary positioning, with different copays being applicable after the edit was met. IHC Health Plans generally has a 3-tier benefit. Tier-1 is for generic medications and has the lowest copay. Tier-2 is for branded preferred medications, while tier-3 is reserved for branded nonpreferred medications that have the highest copay. In addition, bupropion SR, citalopram, and paroxetine were added to IHC Health Plans GenericSample Program (fluoxetine had been available through this program since 2003). GenericSample is a program for IHC Health Plans members that eliminates a copay/coinsurance for the first fill of select generic prescriptions when filled at a participating retail pharmacy. If the member has not filled a prescription for the requested GenericSample drug Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 603 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference in the previous 6 months, the prescribed GenericSample drug will not require a copay/coinsurance. All subsequent refills of that drug will require the usual generic (lowest) copay/coinsurance. Also, a recommendation for prescribing medications for the treatment of depression was sent to physicians that focused on the following messages: generic antidepressants offer a dramatic improvement in cost-effectiveness over their brand-name equivalents; they have an excellent efficacy and safety profile and low expense; they should be considered as the initial choice in a patient presenting with depression; most people respond within the first 4 to 6 weeks of treatment, but a substantial minority will respond after 8 to 12 weeks on an anti-depressant; and because the possibility of a suicide attempt is inherent in major depressive disorder and may persist until significant remission occurs, close supervision of high-risk patients should accompany drug therapy. RESULTS: All generic medications were made available at tier-1, with 4 being available with a zero copay the first time they are filled. Paxil CR (paroxetine), Wellbutrin XL (bupropion), and Effexor XR (venlafaxine) remained at tier-2. Lexapro (escitalopram) and Cymbalta (duloxetine) remained at tier-3, while Zoloft (sertraline) moved from tier-2 to tier-3. Ingredient cost, the number of prescriptions, the ingredient cost per prescription, and the per-member-per-month (PMPM) cost for the antidepressants remained fairly consistent throughout 2004 (prior to the implementation of the GenericStart! Program). The ingredient cost, the number of prescriptions, the ingredient cost per prescription, and the PMPM costs for first-quarter 2004 were $5,403,976, $67,435, $80.14, and $4.22, respectively. For fourth-quarter 2004, the comparable numbers were $5,371,476, $67,028, $80.14, and $4.14. During first-quarter 2005, the first-quarter after the GenericStart! Program was implemented, the ingredient cost, the number of prescriptions, the ingredient cost per prescription, and the PMPM costs all decreased substantially. The respective values were $4,853,841, $62,689, $77.43, and $3.76. This was despite the fact that the average ingredient cost per prescription of the branded antidepressants continued to increase over the 5 reported quarters. As expected, the market share shifted more toward the generics during firstquarter 2005 compared with 2004. During first-quarter 2005, 3,928 prescriptions for generic antidepressants were filled through the GenericSample Program at an expense to the MCO of $22,891. This expense is very small compared with the savings generated through the increased use of generic antidepressants. Generic citalopram became available during fourth-quarter 2004. It had no impact on overall spend or utilization during that quarter. As the price of citalopram and the other generic antidepressants continue to decrease (and an MCO is able to place a maximum allowable cost) and the prices of the branded antidepressants continue to increase, there will be more of an impact on the category, and savings will continue to grow. The overall estimated annualized savings for this MCO 604 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 due to this program exceeds $2 million. CONCLUSIONS: Antidepressants are generally a top 5 mostutilized category for most MCOs. As more antidepressants become available generically, MCOs can implement utilization controls to improve formulary compliance and reduce costs to both the MCO and its members. The implications for the medical side should be negligible, while the long-term savings for the MCO should be significant. As more medications become available generically, this process can be expanded and built upon, especially in large disease categories such as the proton pump inhibitors or HMG-CoA reductase inhibitors (statins). ■■ EFFECTS OF IMPLEMENTING EDITS TO CORRECT DOSING INEFFICIENCIES AMONG THE ATYPICAL ANTIPSYCHOTIC MEDICATIONS IN A MANAGED CARE ORGANIZATION Dunn JD*, Cannon HE, Burgoyne DS. Intermountain Health Care Health Plans, 4646 West Lake Park Blvd., Suite N3, Salt Lake City, UT 84120 INTRODUCTION: The ability of managed care organizations (MCOs) to balance high-quality pharmaceutical care with improved cost efficiency is becoming increasingly more challenging because of a variety of issues. Certain drug categories are exhibiting inefficiencies regarding appropriate utilization and dosing regimens. With the atypical antipsychotics (aripiprazole [Abilify], olanzapine [Zyprexa], quetiapine [Seroquel], risperidone [Risperdal], and ziprasidone [Geodon]), there is substantial off-label use and suboptimal dosing. Dose optimization is one method of addressing the rising costs associated with the use of atypical antipsychotics. An example of atypical antipsychotic dose optimization would be recommending the administration of a single 10 mg tablet in place of two 5 mg tablets if a patient was prescribed 10 mg/day of olanzapine. Another example would be a claims edit that would ensure that one 4 mg tablet was dispensed if a patient was receiving four 1 mg risperidone tablets per day. The keys to implementing a dose-optimization program include (1) maintenance medication being available in multiple strengths, (2) clinical evidence (pharmacokinetics, study data) supporting once-daily administration being available, and (3) similar average wholesale price (AWP) among the different dosage strengths of each drug. This quality-based cost-containment approach ensures that patients still receive the same medication at the same daily dosage; however, the dosing regimen is simplified, which may improve compliance. Medically necessary exceptions (which are clinically supported) to this rule are always allowed. OBJECTIVE: To evaluate the impact of inefficient dosing of atypical antipsychotics and the success of implementing a pharmacy claims edit in an integrated MCO. This report contains preliminary data and an explanation of the methodology www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference undertaken by this MCO. Long-term data and atypical antipsychotic class results will be presented in a follow-up abstract. METHODS: In 2003, Intermountain Health Care (IHC) Health Plans began an investigation of the potential impact of implementing a dose optimization program involving the atypical antipsychotics. A prescribing efficiency analysis was conducted to identify the extent to which inefficient prescribing of olanzapine and risperidone was occurring at IHC. Based on the findings of this analysis, a pharmacy claims edit was put in place for all the atypical antipsychotics. This edit was designed to flag prescriptions at the pharmacy level for the atypical anti-psychotics in which inefficient dosing was present. Pharmacists were required to automatically optimize the dosing regimens based on predetermined algorithms. RESULTS: The average age of patients prescribed olanzapine and risperidone in this MCO was 38 years and 28 years, respectively. There were 880 members receiving olanzapine. The most commonly prescribed daily doses for olanzapine (representing 95% of the total sample, or 836 patients) that were incorporated into this cost analysis were 2.5 mg, 5 mg, 7.5 mg, 10 mg, 15 mg, and 20 mg. Of the 46 members using 7.5 mg/day, only 35 (76%) were receiving one 7.5 mg tablet. Eight members (17%) were using three 2.5 mg tablets and three members (7%) were using one and a half 5 mg tablets. Of the 203 members using olanzapine 10 mg/day, 147 (73%) were receiving one 10 mg tablet, 47 (23%) were using two 5 mg tablets, 8 (4%) were using four 2.5 mg tablets, and 1 (≤1%) was using one 7.5 mg tablet in combination with one 2.5 mg tablet. The average number of days on medication per year was estimated at 230 for olanzapine-treated patients. The average cost per patient per day for olanzapine was $9.10, based on AWP. If the dosing was optimized to the appropriate strength of 1 tablet once daily, the average cost per patient per day would be $8.32, resulting in a savings of $0.78 per patient per day (or 8.5%). Multiplying this estimated savings per patient per day by the number of days on medication per year by an estimated 70% dose conversion would result in an estimated $112,000 annual savings, based on AWP. In comparison, there were 473 members receiving risperidone. The most commonly prescribed daily doses for risperidone (representing 80.5% of the total sample, or 381 patients) that were incorporated into this cost analysis were 1 mg, 2 mg, 3 mg, 4 mg, and 6 mg. Of the 144 members using risperidone 1 mg/day, 108 (75%) were receiving one 1 mg tablet, 32 (22%) were using two 0.5 mg tablets, and 4 (3%) were using four 0.25 mg tablets. Of the 18 members using risperidone 4 mg/day, only 6 (33%) were using one 4 mg tablet, while 9 (50%) were using two 2 mg tablets and 3 (17%) were using four 1 mg tablets. The average days on medication per year for risperidone was estimated at 187. Multiplying a savings of $1.11 per patient per day if the dosing were optimized to the appropriate strength of 1 tablet once daily ($5.71 minus $4.61) by the number of days on medication per year by an estimated www.amcp.org 70% dose conversion would result in an estimated $55,000 annual savings, based on AWP. By simply optimizing the dosing of olanzapine and risperidone, this MCO estimated a potential savings of more than $167,000. Therefore, effective July 1, 2004, IHC Health Plans implemented a dose optimization program involving the atypical antipsychotics. The data do not include diagnosis information, prior or concomitant medication use, net costs (rebates), or prescriber (physician) information. CONCLUSIONS: Atypical antipsychotics are becoming an increasingly difficult category of medication to manage because of substantial off-label use and suboptimal dosing. Programs to optimize dosing can have a positive impact on the cost of these medications and may improve compliance. One-year data from this MCO’s dose optimization program will be available in the fall of 2005. ■■ EFFECTS OF MANAGING HIGH-QUANTITY LONG-ACTING OPIOID UTILIZATION IN A MANAGED CARE ORGANIZATION Dunn JD*, Cannon HE, Burgoyne DS. Intermountain Health Care Health Plans, 4646 West Lake Park Blvd., Suite N3, Salt Lake City, UT 84120 OBJECTIVE: To evaluate the impact of implementing quantity limits, requiring medication management agreements (MMA), and case managing members receiving high quantities of long-acting opioids in an integrated managed care organization (MCO). METHODS: Effective May 1, 2003, Intermountain Health Care (IHC) Health Plans limited the amount of OxyContin a member may receive from 4 tablets/day up to a maximum of 160 mg/day. Exceptions to this limit are evaluated on a case-by-case basis and require each member using more than 160 mg/day to sign an MMA with IHC Health Plans. It requires the member to commit to using 1 physician and 1 pharmacy for his or her pain medications. The purpose of this agreement is to protect the health care providers’ interests while ensuring that members are receiving the most appropriate treatment possible. Also, IHC Pharmacy Services works closely with an IHC case manager, who is available to assist where necessary. This same protocol has been extended to other long-acting opioids (i.e., Duragesic, Kadian, Avinza) and people using multiple physicians and pharmacies for short-acting opioids. Prior to the changes being implemented, letters were sent to members, and letters and clinical information including specific lists of patients on high quantities of long-acting opioids or patients using other additional physicians to obtain pain medications or using multiple pharmacies were sent to health care professionals. In addition, pharmacists from the MCO educated health care professionals via clinical in-office presentations. RESULTS: For the 2 quarters prior to the changes, the number of OxyContin prescriptions totaled 2,417 and 2,495. By thirdquarter 2004, this number had been decreased to 2,072. The daily average consumption for OxyContin 10 mg was reduced Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 605 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference from a high of 3.16 prior to the change to a low of 2.78 after the change. The changes for OxyContin 20 mg, 40 mg, and 80 mg were 2.92 to 2.65, 3.12 to 2.91, and 5.37 to 3.08, respectively. Total ingredient cost for all quarters beginning with first quarter 2003 were $542,780, $540,862, $586,543, $599,669, $558,895, $539,674, and $567,567. This relative flat trend in ingredient cost was despite an average 15% price increase in OxyContin over the same time period. The number of members filling a prescription for OxyContin during first quarter 2003 was 964l in third quarter 2004, 863 members filled a prescription for OxyContin. Some members switched to other newer longacting opioids such as Avinza or Kadian. Overall, the number of members using a long-acting opioid did not change. By third quarter 2004, more than 400 members had signed an MMA. Pharmacy Services and Case Management were actively coordinating care for nearly half of the members with an MMA. This coordinated effort ensures that costs are simply not shifted to facilities such as emergency rooms or to other agents. CONCLUSIONS: The requirement of an MMA, in conjunction with quantity limits and case management, improves the appropriate use of potentially abused and very expensive opioids. Quantity limits aid in the identification of potential opioid misuse, and utilization numbers become more consistent with recommended guidelines. Members are not denied medication. They are only required to more appropriately use the health care system. Theoretically, overall care is improved by using 1 physician and 1 pharmacy. Overall costs are also decreased. ■■ EMPLOYED RHEUMATOID ARTHRITIS PATIENTS EXPERIENCED DECREASED WORK LOSS ON ETANERCEPT TREATMENT Paulus HE, Weaver AL, Yu EB, Xia HA, Chiou CF*, Louie J. Amgen, Inc., 1 Amgen Center Dr., MS 28-3-B, Thousand Oaks, CA 91320 OBJECTIVE: To assess the effects of etanercept on work loss in rheumatoid arthritis (RA) patients. METHODS: Patients aged ≥18 years with active RA who started on or switched to etanercept could enroll in RADIUS 2, a prospective, multicenter, observational study. Data on employment status and the number of times patients missed work for half a day or more in the past month were collected at every visit. Only patients who were on etanercept monotherapy (ETA) or etanercept combination therapy with methotrexate (ETA+MTX) were included in this analysis. Missing data were imputed using last observation carried forward. For patients who reported that they were “full-time,” “part-time,” “temporarily employed,” or “other” at baseline, the number of missed workdays was analyzed comparing baseline with month 6 and month 12 using the Wilcoxon Signed Rank Test. This analysis was based on data available through November 5, 2004. RESULTS: At baseline, there were 1,146 patients on ETA and 606 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 1,659 patients on ETA+MTX. At baseline, 41%, 11%, 1%, and 4% of ETA patients and 45%, 10%, 1%, and 4% of ETA+MTX patients reported that they were “full-time,” “part-time,” “temporarily employed,” or “other,” respectively. The number of times that patients had missed work for half a day or more in the past month because of their RA declined significantly from baseline (mean for ETA = 1.74, mean for ETA+MTX = 1.43) to month 6 (mean for ETA = 0.75, mean for ETA+MTX = 0.67) and to month 12 (mean for ETA = 0.91, mean for ETA+MTX = 0.71) (P <0.0001 at both time points for both groups). CONCLUSION: RA patients on etanercept, either with or without methotrexate, reported significantly fewer missed days of work after 6 or 12 months of treatment. These improvements may be translated into savings in indirect costs to society. ■■ ESZOPICLONE COADMINISTERED WITH FLUOXETINE FOR INSOMNIA ASSOCIATED WITH MAJOR DEPRESSIVE DISORDER: EFFECTS ON SLEEP AND DEPRESSION Skolly S*, Fava M, Buysse D, Rubens R, Wessel T, Caron J, Amato D, Roth T. Sepracor, Inc., 84 Waterford Dr., Marlborough, MA 01752 INTRODUCTION: Insomnia and major depressive disorder (MDD) coexist; currently, no treatment standards exist that address hypnotic administration during antidepressant therapy. This study evaluated the efficacy of eszopiclone on insomnia in patients administered fluoxetine for MDD. METHODS: Patients who met Diagnostic and Statistical Manual of Mental Disorders—Fourth Edition (DSM-IV) criteria for both MDD and insomnia received 10 weeks of fluoxetine QAM (every morning) and were randomized to nightly eszopiclone 3 mg (n= 270) or placebo (n=275) for 8 weeks. Subjective sleep and daytime function were assessed weekly by telephone. Depression was assessed with the Hamilton Depression Rating Scale (HAMD17) every 4 weeks and with the Clinical Global Impression Improvement (CGI-I) and Severity scales (CGI-S). Depression response is ≥50% decrease from baseline HAMD17, and remission is HAMD17 ≤7. RESULTS: Completion rates were similar. Compared with placebo, eszopiclone was associated with significantly decreased WASO (wake time after sleep onset) and increased TST (total sleep time) at each treatment week (P <0.05), significantly decreased latency at all time points (P <0.05) except week 8, higher ratings across the treatment period in sleep quality and depth (P <0.005), and higher ratings of daytime alertness, ability to concentrate, and physical well-being (P <0.05). Patients coadministered eszopiclone experienced significant reductions in HAMD17 total scores compared with placebo coadministration at Week 4 (P = 0.01) and Week 8 (P = 0.002). After removing HAMD17 insomnia items, differences remained significant at Week 8 (P <0.05). At Week 8, significantly more eszopiclone patients were responders (P <0.0011) and remitters (P <0.03). www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference CGI-I and CGI-S scores were significantly improved with eszopiclone coadministration at all time points beyond week 1 (P <0.05). Treatment was well tolerated, with similar adverse event and dropout rates. Unpleasant taste was more common with eszopiclone. CONCLUSIONS: In this study, patients coadministered eszopiclone and fluoxetine experienced statistically significant improvements in several insomnia and depression measures. Eszopiclone was well tolerated and associated with significantly improved sleep and daytime function in patients with insomnia and coexisting MDD. ■■ EVALUATION OF COMORBIDITIES AND COSTS IN PATIENTS WITH BENIGN PROSTATIC HYPERPLASIA AND CARDIOVASCULAR DISEASE Curtice TG*, Shah MB, Bramley TJ. Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury Rd., Ridgefield, CT 06877 INTRODUCTION: To evaluate common comorbidities and annual costs in patients with benign prostatic hyperplasia (BPH) also having cardiovascular disease (CVD). METHODS: A retrospective case-control analysis was conducted in a large, national, managed care database to assess comorbidities and costs for patients with BPH and CVD. The study population (cases) consisted of males, aged 35 years or older, having evidence of CVD within 6 months prior to the index BPH date (first identified BPH diagnosis) and having 18 months of continuous eligibility between January 1997 and June 2004. A control cohort was matched by age and index CVD date (first indication of CVD) at a maximum of 4:1 versus cases. Comorbidities, total costs, and CVD-related costs (adjusted to 2004 dollars) were evaluated over the 12-month period following the index BPH date (using the matched case’s index BPH date for the controls). Log-transformed costs were modeled using regression analyses, and differences in comorbidities between cohorts were assessed using chi-square tests. RESULTS: 82,828 patients having BPH with CVD met inclusion criteria and were matched with 271,440 controls. Hyperlipidemia (60% vs. 47%), diabetes (23% vs. 20%), and erectile dysfunction (10% vs. 4%) were more common with cases (patients having BPH and CVD) than controls (all P <0.0001). Average annual total costs were $7,810 for cases and $5,335 for controls. Average annual CVD costs were $2,204 for cases and $1,613 for controls. Controlling for potential confounders, average costs in the case cohort were 45% higher (P <0.001) for total costs and 46% higher for CVD-related costs compared with the control cohort (P <0.0001). CONCLUSIONS: Patients having BPH with CVD are more frequently diagnosed with comorbidities and have higher costs when compared with a CVD-matched cohort. Although patients with CVD commonly have increased health care costs www.amcp.org associated with their disease, this study highlights additional incremental costs among patients with BPH and CVD. ■■ EVALUATION OF MEDICATION EFFICACY IN A MANAGED CARE ORGANIZATION WHEN USING 3 ORAL ANTIDIABETIC MEDICATIONS VERSUS METFORMIN PLUS INSULIN IN THE TREATMENT OF TYPE 2 DIABETES Mitchell M*, Hanson D, Cannon EH, Burgoyne DS, Dunn JD. Intermountain Health Care Health Plans, 4646 West Lake Park Blvd., Suite N3, Salt Lake City, UT 84120 INTRODUCTION/BACKGROUND: Controlling hyperglycemia reduces the risk of microvascular disease and alleviates the classic symptoms of diabetes mellitus including, but not limited to, polydipsia, polyphagia, and polyuria. As part of an integrated health system, health care providers work together to ensure that cost-effective treatment is part of a care process model for diabetic members. Direct member data need to be evaluated often in order to guarantee that the best service possible is being provided to health plan members. OBJECTIVE: The primary objective of this study was to evaluate glycemic control in type 2 diabetic members on 3 oral antidiabetic medications versus members using insulin plus oral metformin therapy. METHODOLOGY: The health system’s electronic medical record system was used to identify a subset of type 2 diabetic members starting January 1, 2004, through December 31, 2004. These members were then assigned to 2 groups according to their medication regimen. The first group consisted of all members taking 3 different oral anti-diabetic medications. Members of this group needed to have at least 6 prescriptions adjudicated through the system throughout the calendar year. A combination medication (e.g., glyburide/metformin 2.5/500) counted as 2 different medications for the data pull. The second group consisted of all members using any class of insulin (long-acting, short-acting, or combination) therapy plus oral metformin. Members in this group required at least 6 prescriptions for metformin and at least 6 prescriptions for insulin. In order to properly compare these 2 groups, adherence was also reviewed in terms of “Total Length of Therapy,” “Medication Possession Ratio,” and “Persistence.” The primary objective was a comparison of the efficacy of treatment determined by A1c values between the 2 groups. A1c values were retrieved from the electronic medical record system and were reviewed and evaluated by 3 pharmacists and 1 computer data analyst. Results were tested for normality using the AndersonDarling Normality test and found not to be normalized. The Mann-Whitney test was used to determine statistical difference between the groups. RESULTS: Glucose control: The oral group (members taking 3 oral medications) contained 191 members. The insulin/ Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 607 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference metformin group contained 138 members. The oral group showed better glucose control (median A1c = 7.37) compared with the metformin/insulin group (median A1c = 8.02). This difference was statistically different (P = 0.0004). Adherence and demographics of medication use: The total length of therapy is the time from the first fill to the last fill, including the days supply from the last fill. The percentage of time the patient actually possesses the medication during the course of therapy is the medication possession ratio. It can be reported by the length of therapy of the patient or by a specific time period, such as 90, 180, or 365 days. The possession ratio reported here is the length of therapy (LOT), 180 days, and 365 days. (Remember that members were required to have at least 6 months of therapy.) Insulin is not reported in the insulin/ metformin group with the medication possession ratio or the persistence measure because the correct insulin length of therapy is often hard to determine and/or is misrepresented when adjudicated at the point of sale. “Persistence” over time is one of the best measures of adherence; it measures how many patients continue to have their medications filled. Even if there is a gap in therapy, a patient is considered persistent until his last fill of medication. CONCLUSION: In this retrospective analysis, hyperglycemia was better controlled, according to A1c values, in members taking 3 different oral medications compared with patients taking metformin and insulin together. This is despite similar adherence measures. The next step in this evaluation of these treatment regimens is to consider the cost-effectiveness of each group. ■■ EXTENT OF BLOOD PRESSURE CONTROL AMONG DIABETIC PATIENTS WITHIN THE MANAGED CARE SETTING Doyle JJ*, Frech FH, Marshall D. Novartis Pharmaceuticals Corporation, 59 Route 10, East Hanover, NJ 07936 OBJECTIVE: Diabetes is associated with a 2- to 10-fold increased risk of coronary events. Approximately 75% of cardiovascular disease in diabetic patients may be attributed to hypertension. The objective of this study is to determine and compare blood pressure (BP) control in patients with diabetes among 4 health plan populations. METHODS: A retrospective chart review of a random sample of diabetic patients (identified using Health Plan Employer Data and Information Set [HEDIS] specifications for the Comprehensive Diabetes Care measure) was conducted within 4 commercial health plans during 2004. The level of BP control was stratified by 4 criteria according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: (1) diastolic (DBP) level >80 mm Hg, (2) systolic (SBP) level >130 mm Hg, (3) SBP >130 or DBP >80 mm Hg, and (4) BP <130/80 mm Hg, defined as controlled. 608 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 RESULTS: Of the 1,260 usable patient records, 16.9% of the patients were aged 18-44 years, 73.6% were aged 45-64 years, and 9.5% were aged >65 years; 50.3% were female. More than 50% of diabetic patients in all plans had an SBP that met or exceeded 130 mm Hg (range 50.6%-57.7%), and more than 50% of diabetic patients in 3 of 4 plans had a DBP level that met or exceeded 80 mm Hg (range 34.8%-56.6%). In 2 of the 4 plans, fewer than one third of diabetic patients met the recommended BP target of <130/80 mm Hg (range 28.0%-42.5%). CONCLUSIONS: Diabetic patients are frequently not treated to their BP goal, putting them at higher risk for cardiovascular complications. More aggressive treatment by physicians and education by health plans are needed to achieve BP goals for the diabetic population. ■■ FINANCIAL IMPACT OF A PROVIDER AND PHARMACY OUTREACH PROGRAM DESIGNED TO CONVERT MEMBERS FROM PRESCRIPTION TO OVER-THE-COUNTER PROTON PUMP INHIBITORS Devlin KA*, Seneviratne V, Milevich GE, Belazi DT, Tegenu M. AmeriHealth Mercy/PerformRx, 200 Stevens Dr., Philadelphia, PA 19113 INTRODUCTION: An outreach program was initiated to encourage the use of therapeutically equivalent, cost-effective over-thecounter (OTC) drugs. METHODS: Proton pump inhibitors (PPIs) are one of the most heavily promoted and prescribed therapeutic categories of medications in the United States. The introduction of Prilosec OTC in September of 2003 presented health plans with yet another opportunity to utilize therapeutically equivalent OTC products as a cost-effective alternative to available prescription medications. A “therapeutic alternative” outreach program was initiated in a Medicaid health plan where state regulations did not allow for the use of copays or tiers as incentives for formulary compliance among members. Medical providers were given the names of their patients who were currently being treated with prescription PPIs and asked to consider Prilosec OTC as a therapeutic option. The health plan’s pharmacy network was also notified of the program and requested to reevaluate inventory to help ensure product availability and to remind providers of the availability of Prilosec OTC as a formulary alternative when discussing patient therapy options. The PPI prior authorization requirements were lifted for Prilosec OTC during the first 6 months of therapy but remained in place for the prescription PPIs. Providers were given updated PPI utilization reports on their patients on a quarterly basis throughout 2004. RESULTS: Prior to implementing this program, the average cost for a one-month supply of a PPI was approximately $120. As the program progressed, that average cost decreased to approximately $52. Annual savings attributed to the program www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference amounted to more than $3.6 million. CONCLUSIONS: Despite product availability issues during 2004, a provider and pharmacy outreach program to encourage the use of Prilosec OTC was extremely successful. The promotion of OTC products as cost-effective alternatives to prescription products should be considered by health plans when identifying costcontainment solutions. ■■ FINANCIAL IMPACT OF A PROVIDER EDUCATION PROGRAM ON THE APPROPRIATE USE AND DOSING OF ATYPICAL ANTIPSYCHOTICS Seneviratne V*, Devlin KA, Milevich GE, Belazi DT, Tegenu M. AmeriHealth Mercy/PerformRx, 200 Stevens Dr., Philadelphia, PA 19113 INTRODUCTION: The impact, from a payer’s perspective, of a provider education program on the appropriate use and dosing of atypical antipsychotics in a Medicaid population is evaluated. METHODS: Atypical antipsychotics tend to be one of the most expensive classes of medications used by predominately Medicaid populations, if not the most expensive. Compounding resource issues faced by such health plans is the reality that atypicals are often not used in accordance with U.S. Food and Drug Administration labeling and recognized treatment guidelines, such as those created by the Texas Medication Algorithm Project (TMAP). Unjustified combination therapy with 2 or more atypical antipsychotics has become a common practice even though guidelines do not recommend this type of combination therapy until all other therapeutic options have been exhausted. A provider education program was developed to encourage prescribing patterns that are in accordance with the TMAP guidelines. Providers were faxed an illustrative algorithm of the guidelines and a report of their patients receiving duplicate therapy with atypical antipsychotics. They were asked to evaluate the patients’ current regimens and determine if there were opportunities for changes that would improve the patients’ care. A clinical pharmacist was available via telephone should the provider have any questions or would like to discuss an action plan for making changes to a patient’s regimen. Throughout 2004, these reports were sent every quarter, and results were evaluated routinely. RESULTS: Prior to initiating the program, approximately 925 members were receiving combination therapy with 2 or more atypicals each month. By the end of the year, that number decreased to 634, and the total cost savings attributed to the program was approximately $1.1 million. CONCLUSION: A provider education program on the appropriate use and dosing of atypical antipsychotics was successful in improving practice patterns and should be considered by health plans that experience similar prescribing patterns in their population. www.amcp.org ■■ FINANCIAL SAVINGS OF A GENERIC DRUG SAMPLING PROGRAM IN A MANAGED HEALTH CARE SETTING Young SC*, Pervanas HC. Anthem Prescription Management, Anthem BC/BS, PO Box 7101, IN25A-549, Indianapolis, IN 46204 INTRODUCTION: A program was designed and implemented by a pharmacy benefits management (PBM) company to counter increasing drug costs and encourage the use of generic medications by its members. METHODS: The program was available to health plan members with eligible pharmacy benefits in the participating states. PBM clinical pharmacists provided physician offices with sample request forms listing the 10 generic medications offered in the program. Physicians provided these forms to patients who had medical conditions that could be treated by one of the listed generic medications. The patient presented the request form along with a prescription to his or her pharmacy to receive a complimentary supply (up to $10) of the generic medication. RESULTS: The retrospective analysis of pharmacy claims data from January 1, 2004, through December 31, 2004, was conducted to approximate the total savings associated with the program. A total savings of $1,101,675.61 was attributed to this program for 2004, with 18,866 participants. The average ingredient cost savings per prescription was $70.81 for the health plan and $26.86 in copayment dollars for the member. CONCLUSION: Development and implementation of this generic drug sampling program provided a significant cost savings to the health plan and saved copay dollars for the members who continued to take these medications. ■■ HEALTH CARE COSTS OF PATIENTS WITH PERSISTENT ASTHMA Colice G, Crivera C*, Varghese S, Faruqi R, Wu EQ, Birnbaum H, Daher M, Marynchenko MB. sanofi-aventis, 200 Crossing Blvd., M/S BX2-6600D, Bridgewater, NJ 08807-0890 INTRODUCTION: Health care costs of persistent asthma patients were compared with health care costs of individuals without asthma and across asthma severity levels within asthma patients. METHODS: A persistent asthma patient sample (<65 years) was selected from an administrative database (1999-2003). Patients were included in the sample if they met the following criteria: (1) had 1 asthma diagnosis (International Classification of Diseases, Ninth Revision [ICD-9]: 493.xx), (2) had no diagnosis of chronic obstructive pulmonary disease (COPD), (3) satisfied the 2005 Health Plan Employer Data and Information Set (HEDIS) criteria of persistent asthma between July 2002 and June 2003. The asthma sample was subsequently divided by asthma severity according to Leidy’s Reliever and Oral Steroid Method, and recommended inhaled corticosteroids (ICS) dosage by asthma severity in the 2004 Global Initiative for Asthma guidelines for Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 609 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference asthma management. A matched sample of individuals without asthma (2 controls:1 patient) was randomly selected based on patient demographic characteristics. Annual health care costs, measured as payments to medical providers and pharmacies, were compared between asthma and nonasthma control samples, within the overall asthma sample by severity level, and limited to patients receiving long-term ICS treatment. Descriptive statistics were evaluated for statistical significance using paired t tests. Costs were adjusted to 2004 dollars using the medical Consumer Price Index. RESULTS: An average asthma patient uses $6,452 in health care costs annually, whereas the cost for an average control individual with matching characteristics is $2,040 (difference: $4,412, P <0.01). Severe asthma patients, compared with moderate and mild patients, had higher total direct health care costs ($7,933 vs. $6,314 and $4,840, respectively; P <0.01). Health care cost of persistent asthma patients receiving long-term ICS treatment (severe: $7,635, moderate: $5,594, mild: $3,679) were lower compared with the average asthma patients of the same severity. The cost reduction ($1,161) was greatest in mildpersistent patients. CONCLUSIONS: Persistent asthma is an expensive, chronic condition across all severity levels, with health care costs increasing as severity worsens. Long-term ICS treatment can reduce health care costs of asthma, especially in the mild-persistent patients. ■■ HEALTH OUTCOMES OF MS PATIENTS MANAGED BY A SPECIALTY PHARMACY Duong LP*, Cantwell AE, Pennybacker BJ, Cocolicchio CJ, Lewis CA. CuraScript Pharmacy Inc., 6272 Lee Vista Blvd., Orlando, FL 32822 INTRODUCTION: Specialty pharmacies have well-established programs for providing multiple sclerosis (MS) medications and monitoring patients’ therapy adherence. Our specialty pharmacy expanded its program to include the assessment and evaluation of health outcomes. METHODS: A prospective, open-label study was implemented with an Institutional Review Board-approved protocol in February 2003. The objective was to compare health outcomes between patients beginning therapy for MS with interferons (IFNs) and those beginning therapy for MS with the noninterferon, glatiramer acetate (GA). Consecutive MS patients were invited to participate in a telephone survey assessing general health status, depression (Center for Epidemiologic Studies Depression Scale), and fatigue (Fatigue Severity Scale and Daily-Fatigue Impact Scale). Verbal consent was provided for a baseline survey prior to beginning therapy; a mailed written informed consent form had to be returned prior to initiation of follow-up surveys at months 1, 3, 6, and 12. Included in this mailing was the Guy’s Neurological Disability Scale, which assessed the patients’ disabilities in 12 domains commonly affected by MS. 610 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 Other variables thought to potentially impact the outcome measures were collected: concomitant medications, compliance with MS therapies, relapses, adverse events, and demographic characteristics. RESULTS: A total of 103 patients completed 12 months of follow-up as of February 2005; 56 began therapy with IFN and 47 began with GA. Mean age was 45.9 years (range: 28 to 65 years); 88% were female and 85% were white. There were no statistically significant differences between the treated groups in disability, depression, or fatigue at baseline. However, there were significant or near-significant declines in fatigue severity, the impact of fatigue, and depression over the course of the 1-year study period. CONCLUSIONS: Assessing health outcomes as part of a specialty pharmacy program allowed us to advise patients and their physicians about problematic symptoms. In addition, we tracked symptom change, showing reduced fatigue and depression over the 12-month period. ■■ THE HIDDEN VALUE OF PRIOR AUTHORIZATION (PA): REIMBURSEMENT OUTCOMES AND ESTIMATED SAVINGS FOLLOWING A POINT-OF-SALE REJECTION FOR VARIOUS PA DRUGS Cowan C*, Ma J, Semelman S, Grootendorst P. ESI Canada, 5770 Hurontario St., 10th Fl., Mississauga, Ontario L5R 3G5, Canada INTRODUCTION: Although prior authorization (PA) is a popular drug- cost-containment tool, the approval rate of PA requests has been reported to be as high as 90%. Little is known, however, about the rate of successful PA application following a point-of-sale rejection and its attendant effect on drug use and costs. This study analyzed claims data for 8 drugs commonly requiring PA on Canadian employer-sponsored drug plans to determine the impact of PA on patient reimbursement outcomes and savings to the plan. METHODS: Study drugs were grouped by therapy class: osteoarthritis (celecoxib, rofecoxib), erectile dysfunction (sildenafil, tadalafil), rheumatoid arthritis (etanercept, infliximab), and obesity (orlistat, sibutramine). All plans administered by ESI Canada, a large pharmacy benefit manager, with at least one of the study drugs managed on a PA program were included. Retrospective claims data for all patients (n = 4,510) with a point-of-sale PA rejection from October 2003 through September 2004 were examined. Patients reimbursed for the PA drug after the point-of-sale rejection were compared with those with no subsequent paid claims. The estimated decrease in drug costs was based on utilization patterns of patients reimbursed over the study period. RESULTS: Overall, 27.5% of patients (1,240) with a point-of-sale PA rejection pursued reimbursement and were eventually approved. Approval rates varied by therapy class: 16.5% (erecwww.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference tile dysfunction), 20.7% (osteoarthritis), 30.4% (obesity), and 72.5% (rheumatoid arthritis). The estimated decrease in drug costs within each therapy class was calculated to be 48.6% (erectile dysfunction), 44.3% (osteoarthritis), 44.5% (obesity), and 9.5% (rheumatoid arthritis). CONCLUSIONS: The value of PA programs in managing drug utilization should be measured from the point-of-sale rejection. The rate at which patients seek reimbursement following a PA rejection varies by therapy class and has been found to be relatively low compared with the reported approval rate. This translates into savings for employer plans. ■■ IMPACT OF A CARDIOVASCULAR RISK REDUCTION PROGRAM ON LIPID GOAL ATTAINMENT Burns N*, Carter T, Hall K, Evans TS. Southeastern Indiana Health Organization, 417 Washington St., Columbus, IN 47202 INTRODUCTION: A cardiovascular risk-reduction program was evaluated by comparing baseline Health Plan Employer Data and Information Set (HEDIS) cholesterol measures with comprehensive follow-up chart review. METHODS: HEDIS 2004 low-density lipoprotein cholesterol (LDL-C) control was determined. The plan’s medical management nurses employed provider- and patient-directed interventions to address cardiovascular risk reduction and to increase awareness of the importance of cardiovascular medication adherence. Interventions included cardiovascular health screenings at employer health fairs, monthly patient education programs by mail, employer educational lunches, pharmacist-patient communication programs, and communication program for the plan’s staff and nurses. A follow-up chart review determined member cardiovascular risk factors and National Cholesterol Education Program (NCEP) goal attainment. RESULTS: Baseline HEDIS measurements were 63.3% for LDL-C <130 mg/dL and 36.7% for LDL-C <100 mg/dL. Chart review was conducted on 356 members with a diagnosis of hyperlipidemia. LDL-C <100 mg/dL was documented in 51.0% of members with coronary heart disease or CHD risk equivalents, and LDLC <130 mg/dL was documented in 73.2% of members with 2 or more cardiovascular risk factors. Overall NCEP goal attainment was achieved in 66.6% of members evaluated. CONCLUSIONS: Provider- and patient-directed cardiovascular risk reduction interventions had a positive impact on member lipid goal attainment and may positively impact future HEDIS measures. Continued use of this multifaceted approach may result in overall improved cardiovascular health for the plan’s member population. www.amcp.org ■■ IMPACT OF A TARGETED DISEASE INTERVENTION ON HMG-COA REDUCTASE INHIBITOR UTILIZATION IN PATIENTS WITH DIABETES Tjioe D*, Chang E, Nguyen CD, Stroup JR. Prescription Solutions, 3515 Harbor Blvd., Costa Mesa, CA 92626 INTRODUCTION: In May 2004, Prescription Solutions implemented the HMG-CoA Reductase Inhibitor (Statin) Use in Diabetes Targeted Disease Intervention (TDI) to promote the proper use of these agents in patients with diabetes. METHODS: This program was an interactive provider-based pharmacy intervention designed to increase awareness of diabetic dyslipidemia treatment recommendations from the National Cholesterol Education Program Adult Treatment Panel III and the American Diabetes Association. This program involved analysis of medical claims data to identify members who were diabetic or had a diabetes-related diagnosis. Providerspecific reports were generated listing patients who might benefit from lipid-lowering therapy, and on May 4, 2004, these reports were sent to physicians in the intervention cohort. The statin use of patients under the care of intervention and control physicians was then compared. It was hypothesized that members who were included in the intervention mailing would be more likely to initiate statin therapy after the intervention compared with those in the control group. RESULTS: The Statin Use in Diabetes TDI was associated with a nearly 2-fold increase in the percentage of members who initiated statin therapy during the postintervention period. The statin initiation rate was significantly higher (one-sided test with P value = 0.0497) in the intervention cohort compared with the control cohort (13.2% vs. 7.7%), with an adjusted odds ratio of 1.8. CONCLUSIONS: Patients whose physicians received TDI components were more likely to initiate statin therapy than patients in the control cohort. These data suggest that the Statin Use in Diabetes TDI had a positive impact on appropriate statin utilization within this high-risk patient population. ■■ IMPACT OF A TARGETED DISEASE INTERVENTION TO OPTIMIZE ANGIOTENSIN-CONVERTING ENZYME INHIBITOR AND ANGIOTENSIN RECEPTOR BLOCKER UTILIZATION WITHIN A MANAGED CARE ORGANIZATION Jahansouz F, Lew KH,* Vanderplas AM, Jarrar M, Stroup JR. Prescription Solutions, 3515 Harbor Blvd., Costa Mesa, CA 92626 INTRODUCTION: An educational initiative was implemented to optimize angiotensin-converting enzyme inhibitor (ACEI) and angiotensin receptor blocker (ARB) therapy in patients with diabetes. METHODS: This pharmacy intervention highlighted the benefits of blood pressure control in patients with diabetes to physician providers. Educational materials presenting evidence from Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 611 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference published clinical trials and recommendations from the American Diabetes Association were sent to primary care physicians providing care to patients with diabetes. The mailing also included a report listing the provider’s patients with diabetes who had no history of ACEI or ARB utilization within the previous 6 months. Twenty percent of identified physicians were randomly selected as a control group and did not receive the intervention. Six months following the intervention, ACEI and ARB utilization of the previously identified patients under the care of intervention and control physicians was compared to evaluate the program’s impact. Patients not continuously enrolled through the postintervention review period were excluded from the evaluation analysis. RESULTS: A total of 2,621 physicians were identified who provided care to 29,224 patients with diabetes not previously receiving an ACEI or ARB. A total of 24,997 patients were available for follow-up analysis. The ACEI and ARB educational initiative was associated with a significant increase in the percentage of members initiating ACEI or ARB therapy during the postintervention period. Within the intervened cohort, 3,635 of 20,045 patients (18.1%) initiated either ACEI or ARB therapy compared with 821 of 4,952 patients (16.6%) from the control group (P = 0.01). CONCLUSIONS: Physicians who received educational materials and a list of members with diabetes were more likely to initiate ACEI or ARB therapy in their identified patients than control physicians. These data suggest that the intervention had a positive impact on optimizing ACEI and ARB therapy within this high-risk patient population where clinical evidence supports it use. ■■ IMPACT OF CLINICIAN EDUCATION, COLD CARE KITS, AND CLINICIAN PEER REVIEW ON CLINICIAN PRESCRIBING OF ANTIBIOTICS FOR THE TREATMENT OF UPPER RESPIRATORY TRACT INFECTIONS AT A STAFF-MODEL HEALTH MAINTENANCE ORGANIZATION Brzozowski DF*, Durvasula R, Murphy G. Yale University Health Services, 17 Hillhouse Ave., New Haven, CT 06520 INTRODUCTION: The impact of clinician education, cold care kits, and clinician peer review on clinician prescribing of antibiotics for upper respiratory tract infections (URIs) was determined at our staff-model health maintenance organization (HMO). METHODS: Starting in October 2003, the Department of Medicine and Department of Pharmacy used clinician education, cold care kits, and clinician peer review as methods to promote appropriate utilization of antibiotics for upper respiratory infections (URIs) in 3 clinical departments: Internal Medicine, Student Medicine, and Pediatrics. First, clinicians at our staffmodel HMO were educated on appropriate use of antibiotics for URIs at mandatory staff education programs in which treatment 612 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 guidelines for URIs and barriers to appropriate use of antibiotics were discussed. Second, the Department of Pharmacy stocked the Internal Medicine, Student Medicine, and Pediatrics departments with cold care kits to be given to patients presenting to the clinic with URIs. Thirdly, the Department of Medicine and Department of Pharmacy developed a clinician peer review program. On a quarterly basis, the department head provided each clinician with a letter containing their specific antibiotic utilization for the treatment of URIs versus the department average. The results of the audit are being used as part of the clinician performance review, and clinicians are encouraged to seek further education on appropriate use of antibiotics as necessary. RESULTS: Overall, antibiotic prescriptions for treatment of URIs from October to December 2002 compared with October to December 2004 were reduced from 33.3% to 26.5%, a 20.4% reduction in antibiotic utilization. Antibiotic utilization in Student Medicine dropped from 45% to 26.6% (41% reduction), Internal Medicine dropped from 30.2% to 24% (20.5% reduction), and Pediatrics increased from 26.8% to 28.2% (5.2% increase). From October to December 2004, our clinicians dispensed 196 cold care kits to health plan members. A total of 110 cold care kits were dispensed in Student Medicine, 55 in Internal Medicine, and 31 in Pediatrics. CONCLUSIONS: A multipronged approach (clinician education, cold care kits, clinician peer review) is effective at lowering utilization of antibiotics for treatment of URIs. At our staff-model HMO, this method had the most significant impact in our Student Medicine department. ■■ IMPACT OF PHYSICIAN INTERVENTION ON THE MANAGEMENT OF CONGESTIVE HEART FAILURE PATIENTS Geierman JL*, Uchida KM. Beaver Medical Group, 242 Cajon St., Redlands, CA 92374 INTRODUCTION: The impact of a physician intervention program targeting congestive heart failure (CHF) patients was determined through retrospective analysis of patient charts. METHODS: CHF is a chronic and progressively debilitating disease and a major public heart concern in the United States. Despite overwhelming evidence to support their efficacy, angiotensinconverting enzyme inhibitors (ACEI) and β-adrenergic blocker use in CHF patients is often suboptimal. Chart reviews of 775 patients were analyzed to establish a baseline of CHF management at Beaver Medical Group. Following implementation of a physician-based intervention program, a subsequent chart review of 125 of the patients was conducted to determine the impact of the intervention program on CHF management and outcomes. RESULTS: 53 of 125 (42.4%) patients were taking a β-blocker, and 87 (69.6%) patients were taking an ACEI or angiotensin receptor blocker (ARB) at baseline. Following physician education www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference programs, 66 of 110 (60.0%) patients and 75 of 110 (68.2%) patients were taking a β-blocker and ACEI or ARB, respectively. Medication data were not available for 15 patients. In addition, of the 125 patients, only 30 (24.0%) of the patients remained on the same β-blocker regimen and 16 (12.8%) remained on the same ACEI/ARB regimen. Almost 25% and 35% of the patients had their β-blocker and ACEI/ARB, respectively, changed to another drug within the same class during the study period. CONCLUSIONS: Physician education may result in an improvement in management of patients with CHF. Physicians change drug regimens frequently. More study is required to determine whether these prescribing changes are due to formulary issues, financial concerns of patients, clinical reasons, or other confounding variables. ■■ IMPROVING PATIENT HEALTH OUTCOMES USING PHARMACY CLAIMS DATA: A PHYSICIAN-TARGETED POLYPHARMACY PROGRAM Lee G*, Stroup J, Tjioe D, Chang EY, Lew KH. Prescription Solutions, 3515 Harbor Blvd., Costa Mesa, CA 92626 INTRODUCTION: A physician-targeted polypharmacy program was implemented within an employer population to reduce inappropriate use of pharmaceuticals; we then evaluated the program’s impact. METHODS: Members with a medication profile meeting defined criteria for polypharmacy (duplicate therapies, drug-drug or drug-disease interactions, and drugs not recommended for use in the elderly) were identified from pharmacy claims data over a 3-month review period for a 35,000+ member employer. The last prescriber of the polypharmacy agent was notified by mail of the polypharmacy incident and relevant clinical information. A second mailing was issued to those not responding to the initial report. Pharmacy claims were analyzed retrospectively 3 months following the intervention to evaluate the polypharmacy program’s impact. Members who became ineligible prior to the end the postintervention review period were excluded from analysis. Baseline polypharmacy cases not present during the postintervention period were considered resolved. RESULTS: Among 2,924 members with 3,988 polypharmacy incidents identified at baseline, a total of 2,217 members with 2,932 polypharmacy incidents were available for evaluation. The most frequently identified polypharmacy incident was for use of drugs not recommended for the elderly (61.7%), followed by drug-disease interactions (30.1%) and drug-drug interactions (14.8%). Following the intervention, 54.8% of polypharmacy cases were resolved. Resolution was highest for the polypharmacy categories “drug-drug interactions” (61.6%) and “drugdisease interactions” (58.7%) and lowest for “duplicate therapies” (50.4%). Resolution of at least one incident, and all baseline polypharmacy incidents, occurred in 59% and 52.4%, www.amcp.org respectively, of identified members. CONCLUSIONS: The physician-targeted polypharmacy program was effective in eliminating a majority of polypharmacy incidents identified over a 3-month period, reducing the potential for adverse complications in 1,310 of 2,217 members examined during the postintervention period. An ongoing program may help follow up on unresolved polypharmacy incidents and also ensure that all new polypharmacy cases are quickly identified and evaluated for resolution. ■■ ORAL DIABETIC MEDICATION ADHERENCE IS CORRELATED WITH A1C GOAL ATTAINMENT IN A MANAGED CARE DIABETES DISEASE MANAGEMENT PROGRAM Lawrence DB*, Long LB, Israel G, Mohn LA, Parris ES. Pfizer, Inc., 175 Beresford Creek St., Charleston, SC 29492 INTRODUCTION: Poor medication adherence is a significant barrier to positive clinical outcomes. The purpose of this evaluation is to determine the relationship between adherence with oral diabetic therapy and glycosylated hemoglobin (A1c) goal attainment in a diabetes disease management program. METHODS: This was a retrospective, descriptive evaluation of patients enrolled in a managed care diabetes disease management program. A dataset analysis containing demographic, enrollment, pharmacy claims, and clinical lab data was performed. Continuously enrolled patients with a documented A1c obtained at least 90 days after the initial sulfonylurea and metformin prescription index dates were included. The medication possession ratio (MPR) was calculated from the prescription claims records and correlated with the A1c value. RESULTS: Forty-two percent of patients on sulfonylurea therapy and 46% of metformin were reaching an A1c goal of <7.0, and the average MPRs were 0.76 (+0.31) and 0.69 (+0.3), respectively. The average MPR for sulfonylurea-utilizing patients reaching and not reaching the A1c goal was 0.82 (+0.29) and 0.72 (+0.31), respectively (P <0.0001, students t test). The average MPR for metformin patients reaching and not reaching the A1c goal was 0.77 (+0.3) and 0.62 (+0.3), respectively (P <0.0001, students t test). A Pearson correlation showed significant positive associations between A1c and the MPR (sulfonylureas: r= -0.295, P <0.001; and metformin: r = -0.285, P <0.001). CONCLUSIONS: This evaluation shows that oral diabetic medication adherence is a significant factor in A1c goal attainment for diabetes patients. Nonadherence to medications should be considered and evaluated when a patient is not reaching a clinical goal. Barriers to adherence should be assessed and interventions to improve adherence implemented. Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 613 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference ■■ OUTCOMES OF A PROTON PUMP INHIBITOR FORMULARY COMPLIANCE PROGRAM Nelson CL*, Williams CE, Handke LJ. Blue Cross and Blue Shield of Nebraska, 7261 Mercy Rd., Omaha, NE 68180 OBJECTIVE: The goal of the study was to determine if utilization of formulary proton pump inhibitor (PPI) agents could be increased by providing formulary and benefit design information along with formulary brand PPI mail-in rebates to targeted members with claims for non-formulary PPIs. METHODS: A prospective cohort study was conducted. The intervention cohort Blue Cross and Blue Shield of Nebraska (BCBSNE) was compared with a nonintervention control Blue Cross and Blue Shield plan (BCBS). Formulary options for PPIs were identical for both health plans. Members met study criteria if they had continuous enrollment, 3-tier drug copay benefit, and at least one nonformulary PPI claim during the 3-month period prior to mail date November 22, 2004. The analysis period was November 29, 2004, to March 1, 2005. Inclusion criteria were met by 3,379 BCBSNE members (~400,000 lives) and 2,028 BCBS comparison group members (~320,000 lives). The BCBSNE intervention detailed the 3-tier benefit by formulary status of PPIs and copay differentials. Formulary-brand PPI mail-in rebates were included, which reimbursed the member up to $30 in first prescription rebate dollars. The primary end point measured was the incidence of generic or formularybrand PPI claims during the analysis period. Statistical 2-sided chi-square analyses performed with SPSS version 13.0 (SPSS, Inc., Chicago, IL). RESULTS: There were 946 (28%) BCBSNE members and 473 (23%) BCBS control members excluded because they had no PPI claims during the follow-up period. In the 3 months after the letter, 265 of 2,433 (11%) intervention group members and 33 of 1,518 (0.02%) control group members had a claim for a generic or formulary PPI agent claim, P <0.001. In the intervention group, 130 of 265 (49%) of claims were for a generic PPI. Of the intervention group members, 232 (86%) maintained generic or preferred-brand formulary agents throughout the follow-up period, P = 0.028. CONCLUSION: Targeted formulary education regarding formulary PPI agents and brand-formulary mail-in rebate incentives directed toward members utilizing nonpreferred PPI agents were associated with a statistically significant switch to formulary agents. ■■ PATTERNS OF ANTIDEPRESSANT USE AND COST IMPLICATIONS OF PRODUCT SWITCHING Kruzikas D, Khandker RK*, McLaughlin T, Tedeschi M. Wyeth Research, 500 Arcola Rd., Collegeville, PA 19426 INTRODUCTION: The study examines patterns of antidepressant use, including drug switching and related resource utilization. 614 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 METHODS: Using retrospective claims of managed care enrollees using a national database (PharMetrics), the study follows newly diagnosed depression patients (aged 18+ years) with newly prescribed anti-depressants. We identified the proportion of switchers from commonly prescribed selective serotonin reuptake inhibitors (SSRIs: fluoxetine, citalopram, sertraline, and paroxetine) to serotonin and norepinephrine reuptake inhibitors (SNRIs: venlafaxine), and vice versa. We then aggregated health care costs for a 1-year period following diagnosis for various switcher groups. Multivariate regression analyses determined predictors of switching and factors influencing overall and depression-related costs while controlling for confounding factors. RESULTS: Of the 48,950 patients included in the study population, 89% were treated with SSRIs and 11% with SNRIs. Between 12% to 15% of patients switched antidepressants. Of the SSRI switchers, 29% switched to an SNRI. Increased likelihood of switching was associated with female gender, Medicaid coverage, prior anxiolytic use, treatment by a psychiatrist or psychologist, and paroxetine as the index medication. Compared with SSRI nonswitchers, costs for SSRI switchers were 36% higher for all causes and 58% higher for depression-related causes. In contrast, compared with SNRI nonswitchers, costs for SNRI switchers were 27% higher for all causes and 5% higher for depressionrelated causes. Thus, relatively more costly patients are switching from SSRIs to SNRI than vice versa. In addition, among SSRI patients switching to SNRI, costs increased with the number of switches. Mutlivariate analyses confirmed that switching was associated with higher overall and depression-related costs. CONCLUSIONS: Switching among antidepressants is quite frequent among depression patients. Switchers incur significantly higher overall and depression-related costs, and, in general, more costly SSRI patients end up switching antidepressants. ■■ PERSISTENCE WITH ANTIHYPERTENSIVE MONOTHERAPY IN A MANAGED CARE SETTING Patel BV*, Remigio-Baker RA, Thaker DJ, Preblick R. MedImpact Healthcare Systems, Inc., 10680 Treena St., 5th Fl., San Diego, CA 92131 INTRODUCTION: Persistence in taking frequently prescribed classes of anti-hypertensive (AHY) therapy was assessed through a longitudinal, retrospective analysis of pharmacy records. METHODS: Pharmacy claims data (2001-2003) from the MedImpact database were used to identify patients with (1) at least 1 prescription claim from January 1, 2001, to December 31, 2003 (index date), (2) 6 months of negative medication history for AHY therapy, and (3) continuous benefit-eligibility 6 months preindex and 1-year post-index date. Patients were followed for 1 year to assess persistence and medication ownership ratio (MOR: the proportion of members with therapy at end of follow-up). Multiple variable linear regression was used to www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference assess differences in MOR, adjusted for age, gender, business segment, comorbidity, and concurrent cardiovascular-related medication utilization. Pair-wise comparisons were performed. RESULTS: The study cohort consisted of 304,818 patients who initiated AHY monotherapy: 103,064 on β-blockers (BBs), 99,154 on angiotensin-converting enzyme inhibitors (ACEIs), 47,229 on calcium channel blockers (CCB), 41,651 on diuretics (D), and 13,720 on angiotensin receptor blockers (ARBs). A higher proportion of initial ARB patients (53.3%) remained persistent at 12 months past the index date compared with ACEI (49.3%), BB (42.0%), CCB (41.7%), and D (29.9%). Adjusted MORs were greatest for ARB patients (44.8%), followed by ACEI (43.8%), BB (37.3%), CCB (36.5%), and D (27.0%). All pair-wise comparisons were significant (P <0.0001), except BB versus CCB, for persistence. CONCLUSIONS: Initiating AHY with ARBs versus other AHY classes tends to have a greater persistence and proportion of members on therapy after 1 year. These findings have important implications for managed care since optimal medication-taking behavior influences the extent of blood pressure control and the reduction of long-term cardiovascular risks. Further research is needed to assess whether clinical differences exist among AHY therapies as well as the relationships between utilization and health outcomes. ■■ PHARMACOECONOMIC ANALYSES OF ANGIOTENSIN RECEPTOR BLOCKER THERAPY IN PATIENTS WITH MILD-TO-MODERATE HYPERTENSION Joish VN, Brixner DI.* Pharmacotherapy Outcomes Research Center, University of Utah, 421 Wakara Way, Salt Lake City, UT 84108 INTRODUCTION: Seven angiotensin II receptor blockers, also known as sartans (candesartan, eprosartan, irbesartan, losartan, olmesartan, telmisartan, and valsartan), have been approved for the treatment of hypertension. The purpose of this study was to conduct a cost-efficacy analysis (CEA) based on clinical outcomes and costs to determine the least costly agent per treatment success from a managed care perspective. METHODS: Efficacy data were derived from published head-tohead clinical studies or meta-analyses that evaluated angiotensin receptor blockers (ARBs) in mild-to-moderate hypertension patients. The primary efficacy outcome was diastolic blood pressure reduction in mm Hg at week 6 to 8 from baseline. Average efficacy rates were determined by calculating averages of the efficacy rates in the individual studies. Costs to treat were averaged for multiple dosing regimens. Costs included were 2005 average wholesale prices from FirstData Bank. RESULTS: Eighteen published studies were identified, and efficacy data from these studies were used in the CEA model. Exclusions were applied for using doses or durations that differed from the primary efficacy outcome. The highest average diastolic blood www.amcp.org pressure reduction from baseline was for losartan hydrochlorothiazide 100 mg/25mg (17.5 mm Hg), and the lowest was for valsartan 160 mg (5.3 mm Hg). The lowest average cost per diastolic blood pressure reduction was for eprosartan, at an average of $0.14, and highest for candesartan and valsartan, at an average of $0.24. CONCLUSION: Using this simple straightforward pharmacoeconomic method, eprosartan mesylate had the lowest cost-efficacy ratio compared with other sartans in mild-to-moderate hypertensives. ■■ PREVALENCE AND FACTORS ASSOCIATED WITH VACCINATION RATES AMONG U.S. ADULTS AT HIGH RISK OF VACCINE-PREVENTABLE HEPATITIS Cantrell CR*, Chen H, DeBartlo C. GlaxoSmithKline, Five Moore Dr., Research Triangle Park, NC 27709 INTRODUCTION: This study sought to estimate the prevalence of adults at high risk of vaccine-preventable hepatitis (hepatitis A virus [HAV] and hepatitis B virus [HBV]) in the United States and their vaccination rates. Secondly, the study investigated the association between vaccination rates and demographic (age, gender, location of birth, and race) and social economic (annual household income, education level, and marital status) characteristics. METHODS: Four years (1999-2002) of publicly available National Health and Nutrition Examination Survey data were utilized. Survey participants aged 20 to 59 years were selected. Survey participants were considered at high risk of vaccinepreventable hepatitis if they belonged to a “risk population group” where their situation and/or behavior placed them at a higher risk of contracting hepatitis, as identified by the Centers for Disease Control and Prevention. All prevalence estimates were weighted to represent the total U.S. population, using 4-year interview and examination weights. Logistic regression was utilized to identify demographic and social economic factors associated with vaccination rates. RESULTS: The study included 6,237 survey participants who represent 153,919,438 adults aged 20 to 59 years in the United States. Of those adults, 12,347,634 (8.0%) were at high risk of HAV, 18,849,536 (12.3%) were at high risk of HBV, and 2,141,907 (1.4%) were at high risk of both HAV and HBV. The vaccination rates among these high-risk groups were 13%, 23.6%, and 13.4%, respectively. The most prevalent risk groups were persons with sexually transmitted diseases and persons using illegal drugs. Within the higher-risk population, single males between the ages of 20 and 29 years were significantly (P <0.05) less likely to be vaccinated than their counterparts. CONCLUSION: Among persons identified at high risk of vaccinepreventable hepatitis (HAV, HBV, or both), only a small proportion Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 615 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference of this population had evidence of hepatitis vaccination (13%, 23.6%, and 13.4%, respectively). ■■ REAL-WORLD DOSING BEHAVIORS FOR SELF-ADMINISTERED ANTITUMOR NECROSIS FACTORS Vishalpura TV*, Grogg A, Abeyta K. Applied Health Outcomes, 4114 Woodlands Pkwy., Suite 500, Palm Harbor, FL 34685 OBJECTIVE: To assess real-world dosing patterns for self-administered anti-tumor necrosis factors (TNF). METHODS: Retrospective analyses of National Data Corporation’s database were conducted. Patients with a pharmacy claim for etanercept (25 mg, 50 mg) or adalimumab (40 mg) for various therapeutic areas, including inflammatory, gastroenterology, and dermatology, from November 2004 through January 2005 were analyzed. Prescription claims were stratified into 3 cohorts: new, continuing, or switched. Patients in the new prescription cohort were those without a claim 3 months preceding the study period. Patients in the continuing prescription cohort were those with a claim 3 months preceding the study period. Patients in the switched prescription cohort were defined as those switched from 25 mg to 50 mg etanercept. Average dose per patient and percentage deviation from reference dosing were calculated for etanercept (25 mg, 50 mg) and adalimumab for each cohort and therapeutic area. RESULTS: For both drugs, patients in the new cohort had greater dose deviations and higher average doses than those in the continuing cohort across all therapeutic areas. Among patients in the new cohort, both agents had doses well above the reference dose. Dermatology patients utilized the highest average dose for both drugs. The average dose was 94% higher than the onceweekly reference dose among patients receiving the etanercept 50 mg formulation. Etanercept patients in the 25 mg continuing cohort received doses that were slightly below the twice-weekly reference dose, and those in the 50 mg continuing cohort received doses that were 25% above the once-weekly reference dose. In the adalimumab continuing cohort, doses were 10% above the 40 mg every-other-week reference dose. Specifically, for gastroenterology, adalimumab had the highest average dose, at 66% above the reference dose. CONCLUSION: Overall, in the cohorts examined through a retrospective review of claims data, this study illustrates that doses well above reference doses are commonly prescribed for the self-administered anti-TNFs. ■■ RESOURCE USE AND ANEMIA TREATMENT COSTS AMONG CANCER PATIENTS TREATED WITH EPOETIN ALFA OR DARBEPOETIN ALFA Harley C, Muser E*, Nelson M, McKenzie RS, Piech CT. Ortho Biotech Clinical Affairs, LLC, 1061 Brook Mont Dr., O’Fallon, MO 63366 INTRODUCTION: This study examines resource use and treatment costs in cancer patients receiving epoetin alfa (EPO) or darbepoetin alfa (DARB) in a managed care setting. METHODS: A retrospective analysis was conducted using medical claims from a large U.S. health plan from January 1, 2002, through December 31, 2003. Cancer patients aged 18 years and older with at least 2 doses of EPO or DARB were eligible for inclusion. Patients with renal failure or who were switched between erythropoietic agents were excluded. Treatment episode was defined as the period from first EPO/DARB claim to last claim (with no gaps in treatment greater than 30 days) with a 3-month maximum duration of treatment. Resource use and treatment costs were examined for each group. RESULTS: 4,753 EPO and 1,601 DARB patients met inclusion criteria. Patient age was similar between groups with a higher proportion of men in the EPO group compared with the DARB group (EPO 34%, DARB 29%, P <.001). Treatment duration (approximately 8 weeks) and proportion of patients transfused were similar. While the number of hemoglobin determinations and cancer-related outpatient visits overall were similar, the number of anemia-related outpatient visits were lower in the DARB group (EPO 6.92 vs. DARB 4.89, P<.001). Mean anemiarelated costs (per patient per episode) were significantly lower for the EPO group than for the DARB group, respectively ($3,943 vs. $5,723, P <.001). CONCLUSIONS: In this managed care population, mean anemiarelated costs were 45% higher for the DARB group than for the EPO group. A similar number of cancer-related office visits and hemoglobin determinations were observed between groups despite less-frequent DARB dosing, suggesting that factors other than frequency of erythropoietic agent administration determine resource utilization in this population. ■■ SYNCHRONICITY: A STUDY TO EVALUATE THE EFFECTIVENESS OF DARBEPOETIN ALFA AT 300 MCG EVERY 3 WEEKS ON CLINICAL OUTCOMES IN CANCER PATIENTS WITH ANEMIA DUE TO CHEMOTHERAPY Boccia R, Silberstein P, Tchekmedyian S, Tomita D, Rossi G*, Otterson GA. Amgen, Inc., One Amgen Center Dr., Thousand Oaks, CA 91320 OBJECTIVE: To assess the effectiveness of darbepoetin alfa administered at 300 mcg every 3 weeks (Q3W) in achieving and maintaining hemoglobin (Hb) levels recommended by current evidence-based guidelines. 616 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference METHODS: This interim analysis of a multicenter, open-label, 16-week study of patients with chemotherapy-induced anemia (CIA) includes 1,225 patients who received at least 1 dose of darbepoetin alfa. Study end points included the proportion of patients who achieved and maintained Hb levels within the range recommended by current evidence-based guidelines (11 to 13 g/dL), the proportion of patients who required red blood cell (RBC) transfusions, and patient-reported changes in symptoms of fatigue measured using the Functional Assessment of Cancer Therapy-Fatigue (FACT-F) scale. RESULTS: Seventy-nine percent of patients achieved the target Hb range, of which 72% maintained Hb levels within the range of 11 to 13 g/dL, consistent with current evidence-based guidelines. The mean (95% confidence interval) change in Hb from baseline to the end of the treatment period was 1.5 g/dL (1.3, 1.6) (last-value-carried-forward approach). Darbepoetin alfa reduced the proportion of patients who required RBC transfusions from 11% in the first month to 4% in the last month of the study. Increases in Hb levels from the start to the end of the study period were associated with clinically significant improvements in FACT-F. CONCLUSIONS: Darbepoetin alfa administered at 300 mcg Q3W is well tolerated by cancer patients with CIA. Most patients achieved and maintained Hb levels within the range recommended by the current evidence-based guidelines, required fewer RBC transfusions, and demonstrated improvements in symptoms of fatigue. Administering darbepoetin alfa Q3W may simplify the treatment of CIA in the oncology practice by synchronizing darbepoetin alfa therapy with chemotherapy, thereby reducing disruption to the life of patients and their caregivers. resource use were compared between the preperiod and postperiod. Statistical tests were conducted using McNemar’s tests for categorical variables and Wilcoxon tests for continuous variables. The most common TPM-associated adverse events observed in pivotal trials included paresthesia, cognitive impairment, fatigue, anorexia, and nausea. RESULTS: The study included 1,749 TPM prophylaxis patients. The mean ± SD age was 43 ± 10 years, and 90% were female. The mean ± SD number of TPM claims per patient in the 6-month postperiod was 5.5 ± 1.8. Changes in preacute versus postacute medication and resource use were as follows: 15% decrease in triptan prescriptions (P <.0001); 7% decrease in other abortive medications (P <.0001); 8% decrease in outpatient services (P = .045), which included a 45% decrease in emergency room services (P<.0001) and a 50% decrease in procedures such as computed tomography scan and magnetic resonance imaging (P =.0003); and 52% decrease in hospital admissions (P =.001). CONCLUSION: This is the first study to examine the impact of TPM prophylaxis on the resource use of migraineurs in a realworld setting. In this setting, use of TPM was associated with significantly lower acute medications and health care resource use. Future studies should examine the longer-term impact of TPM prophylaxis. ■■ TOPIRAMATE IN THE PROPHYLAXIS OF MIGRAINE: IMPACT ON RESOURCE UTILIZATION Vermilyea JA, Feliu AL, Rupnow MFT*, Carter CT, Blount A, Ollendorf DA. PharMetrics, Inc., 1125 Trenton-Harbourton Rd., Titusville, NJ 08560 OBJECTIVE: To describe the demographic and pharmacy/medical resource utilization profile of patients diagnosed with migraine who received topiramate (TPM) prophylaxis. METHODS: The Pharmetrics Patient-Centric Database was used to obtain pharmacy and medical claims data from January 1, 1999, to September 30, 2004. Migraine patients were required to have continuous enrollment 6 months prior to and after an index TPM prescription. In the 6 months immediately preceding the TPM index prescription, patients had to have ≥1 triptan prescription with a minimum of 6 and a maximum of 180 triptan equivalents (TEs). Patients with claims for other U.S. Food and Drug Administration (FDA)-approved prophylaxis therapy in the postperiod were excluded. Demographics, comorbidities, prescriber specialty, and migraine-related inpatient and outpatient resource use were assessed using descriptive statistics. RESULTS: The study included 1,749 TPM prophylaxis patients. Mean ± SD age was 43 ± 10 years, with 83% of patients aged between 25 and 54 years. Ninety percent were female and 81% were commercially insured. Neurology providers initiated 54% of TPM prescriptions. The most common comorbidities were OBJECTIVE: To evaluate the effect of topiramate (TPM) prophylaxis on acute medication use and medical resource utilization among patients diagnosed with migraine. METHODS: The Pharmetrics Patient-Centric Database was used to obtain pharmacy and medical claims data. The first TPM claim between January 1, 1999, and September 30, 2004, was considered the index date. Triptan use was standardized to triptan equivalents (TE) based on the maximum daily dose allowed. Analysis included patients continuously enrolled for at least 6 months preindex and postindex date, with ≥6 TEs (but ≤180 TEs) during 6 months preindex date and ≥3 claims for TPM prescription. Patients with use of other U.S. Food and Drug Administration-approved prophylaxis therapy in the postperiod were excluded. Migraine-related inpatient and outpatient www.amcp.org ■■ TOPIRAMATE IN THE PROPHYLAXIS OF MIGRAINE: PATIENT PROFILES FROM A PAYER DATABASE Vermilyea JA, Feliu AL, Rupnow MFT*, Carter CT, Blount A, Dodd S, Ollendorf DA. PharMetrics, Inc., 1125 Trenton-Harbourton Rd., Titusville, NJ 08560 Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 617 Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference depression and psychosis (16%), allergy (11%), gastrointestinal disorders (8%), and anxiety (6%). Mean ± SD Charlson Comorbidity score was 0.20 ± 0.6. Prior to receiving TPM, 78% had a prescription for FDA- and non–FDA-approved prophylaxis therapy and 72% used abortive prescription medications in addition to triptans. The mean ± SD number of TEs per person was 23.2 ± 19.4. Six months prior to receiving TPM, these migraineurs had a mean ± SD of 0.64 ± 3.32 emergency room visits, 2.12 ± 2.99 physician office visits, 0.08 ± 1.72 procedures, and 0.41 ± 1.72 lab tests, cumulating to 4.20 ± 8.03 migrainerelated outpatient services. Additionally, they had 4 hospitalizations and 18 days hospitalized per 100 migraineurs. CONCLUSION: In the 6 months prior to TPM initiation, migraineurs showed patterns of extensive acute and prophylactic pharmacy resources as well as health services utilization. This is the first study to examine the TPM patient profile from a claims perspective. Future studies should examine the impact of TPM prophylaxis in reducing migraine-related resource consumption. ■■ TREATMENT RATES AND GLYCEMIC CONTROL IN U.S. ELDERLY TYPE 2 DIABETIC PATIENTS, 2001-2002. Suh DC, Shin H, Wogen J, Plauschinat C.* Novartis Pharmaceuticals Corporation, One Health Plaza, East Hanover, NJ 07936 OBJECTIVE: To characterize the elderly type 2 diabetes (T2DM) population in the United States, we evaluated patient characteristics, treatment rates, and glycemic control in a U.S.-representative, weighted sample of T2DM patients aged ≥65 years. METHODS: The National Health and Nutrition Examination Survey 2001-2002 (n = 11,039) was used to identify persons with T2DM aged ≥65 years, defined by physician diagnosis at >30 years or current or past use of insulin and/or oral agents. Body mass index was used to classify patients as normal weight (<25 BMI), overweight (25-29 BMI), or obese (>30 BMI). Data were analyzed using SAS and SUDAAN statistical software. RESULTS: Age-adjusted prevalence was 14.22% (SE = 0.86), estimating that 4.3 million elderly persons in the United States have been diagnosed with T2DM. Of these, 56.9% were aged 65 to 74 years, 34.3% 75 to 84 years, and 8.8% >85 years. Sixty percent were female, and 80% were non-Hispanic white, 9% non-Hispanic black, and 9% Hispanic. Approximately 40% were overweight and 46% obese. Mean glycosylated hemoglobin (A1c) for elderly T2DM patients was 6.99%; 42.3% had A1c >7%. 83.7% of patients were treated with an oral agent and/or insulin; half (49.9%) of treated patients had A1c >7%. 56.1% of patients treated with only oral agents, 33.7% of patients treated with only insulin, and 38.5% of patients treated with oral agents and insulin had A1c <7%. CONCLUSIONS: Though 84% of elderly T2DM patients were treated with oral agents and/or insulin, glycemic control rates in this population were strikingly poor, as 49.9% of treated patients 618 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 had A1c >7%. Almost half (46%) of elderly T2DM patients were obese. This study highlights a significant unmet need for more aggressive treatment of elderly T2DM patients to achieve American Diabetes Association-recommended A1c goal. ■■ TREATMENT SATISFACTION AND PATIENT FUNCTIONING WITH OXYBUTYNIN OR TOLTERODINE IN PATIENTS WITH OVERACTIVE BLADDER Wan GJ, Chen A*, Bolge SC. Ortho Urology, Ortho-McNeil Pharmaceuticals, Inc., 1000 Route 202, Raritan, NJ 08869 OBJECTIVE: To compare satisfaction and patient functioning with immediate-release (IR) and extended-release (ER) formulations of oxybutynin or tolterodine in patients with overactive bladder (OAB). METHODS: Cross-sectional data were obtained from the Consumer Health Sciences 2004 National Health and Wellness Survey, a nationally representative sample of a noninstitutionalized U.S. civilian population. Patients were currently taking either IR or ER oxybutynin (n = 162) or IR or ER tolterodine (n = 277) to treat OAB. Patient satisfaction with treatment (PST) was measured using a 5-point scale (1 = not at all satisfied to 5 = extremely satisfied). Satisfaction rates were computed as the percentage of patients reporting a 4 or 5 on the PST scale. Patient functioning was assessed using the mental and physical component summary scores of the 8-item Short-Form Health Survey and the activity impairment score of the Work Productivity and Activity Impairment Questionnaire. A P value <0.05 in a 2-tailed test was considered significant. RESULTS: The mean age of patients was 58 years, and 84% were female. More patients reported that they were satisfied with oxybutynin (66%; 107 of 162) versus tolterodine (48%; 132 of 277; P <0.001), with an adjusted odds ratio of 1.98 (95% CI, 1.31, 3.00; P <0.001). Those satisfied with their studied OAB medication reported statistically significantly better patient functioning compared with those who were dissatisfied. CONCLUSIONS: In this study, significantly more patients were satisfied with oxybutynin versus tolterodine. Satisfaction with OAB therapy and patient functioning are important factors to consider in the selection of OAB medication by health care professionals and payers. ■■ TREATMENT WITH ETANERCEPT IMPROVES PATIENT-REPORTED OUTCOMES IN PATIENTS WITH MODERATE-TO-SEVERE PSORIASIS Gottlieb AB, Chiou CF*, Woolley JM, Lalla D, Jahreis A. Amgen, Inc., One Amgen Center Dr., MS 28-3-B, Thousand Oaks, CA 91320 OBJECTIVE: To evaluate the effect of etanercept therapy on patientreported outcomes in patients with moderate-to-severe psoriasis. www.amcp.org Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference METHODS: A 12-week, double-blind, multicenter clinical trial was conducted in the United States and Canada to confirm the efficacy and safety of etanercept 50 mg twice weekly in patients with psoriasis. Patients with stable moderate-to-severe psoriasis were randomized to receive subcutaneous etanercept 50 mg twice weekly or placebo. Patient-reported outcomes included the Dermatology Life Quality Index (DLQI), a disease-targeted health-related quality-of-life questionnaire. The DLQI has 10 questions that cover 6 life areas (symptoms and feelings, daily activities, leisure, work and school, personal relationships, and treatment satisfaction). Lower scores indicate better health status; a 0 score means that a patient is “not at all” bothered by psoriasis in the 6 life areas. Patients who received at least one dose of study drug and provided baseline DLQI assessments (N = 308 for etanercept; N = 304 for placebo) were included. RESULTS: Relative to patients on placebo, patients on etanercept achieved statistically significantly greater percentage improvements in DLQI total score by week 1 (P < 0.05). At week 12, the mean percentage improvement in DLQI was 69.1% for patients receiving etanercept compared with 22% for patients on placebo (P < 0.0001). Furthermore, at week 12, the improvements in each of the 6 DLQI life areas for patients receiving etanercept were statistically significantly superior to those for patients receiving placebo (P < 0.001). At week 12, significantly more patients on etanercept achieved a 0 score on the DLQI (28% on 50 mg twice weekly and 3% on placebo; P < 0.0001). CONCLUSIONS: Treatment with etanercept improves patient-reported outcomes in patients with moderate-to-severe psoriasis. Combined with clinical efficacy, these results support the benefit of etanercept in patients with moderate-to-severe psoriasis. ■■ UTILIZATION OF CHOLINESTERASE INHIBITORS IN THE TREATMENT OF ALZHEIMER’S DISEASE Mucha L*, Cuffel B, McCrae T, Mark T, Wang S. Thomson Medstat, Inc., 125 Cambridge Park Dr., Cambridge MA 02140 days on therapy, and medication possession ratio. METHODS: Data were from MarketScan’s Medicare Claims Database from 2001 through 2003 (n = 3,177). Inclusion criteria were (1) age 65 years or older, (2) at least 1 claim with an International Classification of Diseases, Ninth Revision (ICD-9) code for Alzheimer’s disease (331.0), (3) at least 1 cholinesterase inhibitor prescription preceded by a 6-month period without any such prescription, and (4) at least 18 months of continuous enrollment. A 30-day gap between prescription end date and fill date indicated discontinuation. Cohorts were defined by their starting medication (index therapy) and were similar in most patient characteristics. Logistic regression models tested for cohort differences in discontinuation and switching, controlling for demographics, region of the country, type of insurer, and Charlson index. RESULTS: Fewer patients started on galantamine (73%) and rivastigmine (79%) reached an effective dose than those started on donepezil (99%, P <.0001). Donepezil patients had higher rates of medication persistence starting at month 5 compared with rivastigmine, and at month 7 compared with galantamine (P <.05). Switching rates were higher for rivastigmine (odds ratio [OR]=1.57, P<.0001) than donepezil patients but were not different for galantamine patients. Similarly, discontinuation rates were higher for rivastigmine (OR = 1.22, P <.05) than donepezil patients but were not different for galantamine patients. Donepezil patients spent more days on index therapy than galantamine (215 days) or rivastigmine (206 days) patients. Cohorts did not differ in medication possession ratio. CONCLUSIONS: Choice of initial therapy affects medication adherence in a cognitively impaired Medicare population. Patients initiated on donepezil were more likely to reach an effective dose, persist on therapy, and have more days on the initial therapy than patients started on other cholinesterase inhibitors. INTRODUCTION: This study compared use of cholinesterase inhibitors by patients diagnosed with Alzheimer’s disease on time-to-effective-dose, persistence, discontinuation, switching, www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 619 RESIDENCIES, FELLOWSHIPS, AND OTHER PROGRAMS Managed Care Pharmacy Residencies, Fellowships, and Other Programs T he following is a partial list of available managed care pharmacy residency and fellowship programs and other programs compiled as of August 2005. The residencies listed were submitted by AMCP members in response to AMCP’s call for residency program listings. This is not a comprehensive list of all available programs. AMCP provides it solely as a service to its readers. This list does not imply AMCP’s endorsement of any particular program nor does AMCP guarantee the availability of any of the programs listed. AMCP does not assume responsibility for any errors that may appear in these listings. If you are aware of additional residency and fellowship programs not listed here, please contact AMCP at (800) TAP-AMCP. ■■ AMERICA SERVICE GROUP Managed Care Accredited: Yes Length of Program: 1-year, 2-year option Number of Positions: 1 Affiliation: Various universities Application Deadline: February 15 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: Graduate of accredited school of pharmacy Fringe Benefits: 3 weeks paid personal leave, paid sick leave, full medical and dental, tuition assistance toward specified degree programs Special Features: America Service Group is a closed-model HMO that provides comprehensive at-risk health care; Secure Pharmacy Plus is the pharmacy benefits manager. Contact Information: Peter Mikhail Vice President, Clinical Services American Service Group/Secure Pharmacy Plus 416 Mary Lindsay Polk Dr., Suite 515 Franklin, TN 37067 (615) 771-1457 (615) 771-4557 (fax) [email protected] pharmacist license or eligibility for licensure Fringe Benefits: Medical, dental, vision, holidays, vacation, and attendance at national conference(s) Special Features: Activities the resident will be involved in include, but are not limited to, P&T committee participation and presentations; formulary management and review; one-onone physician correspondence on evidence-based medicine with current clinical studies reviewed and presented, focusing on actual patients who may benefit from this information; new drug review and its placement with available therapies; clinical participation and set-up; client summary report write-up and delivery; participation in various education conference and clinical studies. Contact Information: Nazly Westernoff American Health Care 3001 Douglas Blvd., #320 Roseville, CA 95661 (916) 773-7227 (916) 773-7210 (fax) [email protected] ■■ AMERICAN HEALTH CARE Clinical Therapeutics / Managed Care Pharmacy Accredited: Being pursued Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: Open Starting Date: July 1 Estimated Stipend: $43,000 and up Onsite Interview: Yes Educational/Special Requirements: PharmD with a California 620 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 ■■ AON CONSULTING Pharmacy Benefits Consulting Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 3 Starting Date: July 5 Estimated Stipend: $36,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, managed care course work preferred Fringe Benefits: 2 weeks paid vacation, paid holidays, medical/dental insurance, travel budget, professional meetings Special Features: Aon Consulting is the second largest employee benefits consulting firm nationwide. This unique program provides residents with the opportunity to help www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs employers and health plans better manage and control the prescription drug benefit for their employees or members. Residents will be exposed to all aspects of pharmacy benefit management consulting: plan design modeling, formulary analysis, clinical programming, disease management, audits, regulatory, trends and forecasting through proprietary actuarially based models. Residents will interact with benefit administrators, major PBMs, PPOs, HMOs, and disease management firms nationally. Currently, there are 2 resident positions available, based in Chicago, IL, and Philadelphia, PA. Contact Information: Connie Perry Vice President Aon Consulting 330 East Kilbourn Ave., Suite 450 Milwaukee, WI 53202 (414) 225-5345 (414) 276-3929 (fax) [email protected] ■■ APPLIED HEALTH OUTCOMES Health Outcomes Research Fellowship Accredited: No Length of Program: 2 years Number of Positions: 1 Affiliation: University of South Florida, College of Public Health Application Deadline: January 3 Starting Date: July 1 Estimated Stipend: Contact program Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent (residency preferred) Fringe Benefits: Competitive salary, health insurance, vacation, 401(k), tuition for required classes, and travel expenses to 1 national meeting per year Special Features: This 2-year, degree-granting fellowship provides a unique research and education experience in an outcomes consulting environment. Research activities include, but are not limited to, quality improvement programs, database analysis, economic modeling, and development of researchbased manuscripts. In addition, the fellow will obtain an MSPH or MPH degree from the University of South Florida, College of Public Health. Contact Information: James H. Jackson IV, PharmD, MPH Senior Consultant Applied Health Outcomes 4114 Woodlands Pkwy., Suite 500 Palm Harbor, FL 34685 (727) 771-4100 www.amcp.org (727) 771-4145 (fax) [email protected] Web: http://www.applied-outcomes.com ■■ BIOSCRIP Managed Care Pharmacy Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: February 1 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, licensure in the state of Ohio Fringe Benefits: 10 vacation days, 5 holidays, 4 floating holidays, 2 personal days, health benefits, and professional travel allowance Special Features: This 12-month residency program is designed to provide a firm knowledge base of managed care from the pharmacy benefit perspective and allows residents to excel in clinical services, formulary management and research, participation in formulary development and medication management, and provision of clinical services, including drug information, clinical program development, pharmacoeconomic assessment, and disease state management as well as the strengthening of written and verbal communication skills. Contact Information: Suzanne Tschida, PharmD, BCPS Director, Clinical Services BioScrip 10900 Red Circle Dr. Minnetonka, MN 55343 (800) 444-5951, ext. 3621 (952) 352-6785 (fax) [email protected] ■■ BLUE CROSS AND BLUE SHIELD OF ALABAMA Managed Care Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 15 Starting Date: July 1 (flexible) Estimated Stipend: $32,000 Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent experience Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 621 Managed Care Pharmacy Residencies, Fellowships, and Other Programs Fringe Benefits: Paid vacation, personal holiday leave, health/dental insurance, no on-call responsibilities Special Features: The program provides the resident with the opportunity to experience a true integrated medical and pharmacy system. The areas of focus will include pharmaceutical care, drug information, formulary management, clinical program management, disease state management, and outcome studies. The resident also will complete a research project suitable for publication. This program will incorporate communication and time management skills. Contact Information: Jerry Wong, PharmD, MBA Residency Director Blue Cross and Blue Shield of Alabama 450 Riverchase Pkwy., East Birmingham, AL 35244 (205) 220-6526 (205) 220-2939 (fax) [email protected] ■■ BLUE CROSS AND BLUE SHIELD OF NEBRASKA Managed Care Pharmacy Accredited: Being pursued Length of Program: 12 months Number of Positions: 1 Affiliation: University of Nebraska Medical Center Application Deadline: January 16 Starting Date: July 1 Estimated Stipend: $36,000 Onsite Interview: Required Educational/Special Requirements: PharmD degree from an ACPE-accredited college of pharmacy. Licensed pharmacist in the United States or eligibility for licensure (successful candidate must be fully licensed at the start of residency). Application requirements include: application, writing sample, 3 professional letters of recommendation, and onsite interview with short formal presentation. Fringe Benefits: 10 days paid vacation, holidays, health insurance, paid travel and registration to AMCP Educational Conference and Annual Meeting Special Features: The Blue Cross and Blue Shield of Nebraska residency will offer exposure to various aspects of managed care pharmacy through health plan and PBM exposure as well as direct patient care through clinical rotations. The residency will offer participation and exposure to the following areas: pharmacy benefit design, utilization management, formulary management, pharmacy trend management, drug information, health economic/outcomes research, pharmacy benefit management, disease state management quality assurance initiatives. This unique program also allows the resident the option to participate in teaching or didactic coursework through the 622 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 University of Nebraska Medical Center. Contact Information: Lee Handke Vice President of Pharmacy and Wellness Blue Cross and Blue Shield of Nebraska 7261 Mercy Rd. Omaha, NE 68180 (402) 398-3884 (402) 548-4683 (fax) [email protected] ■■ BLUE SHIELD OF CALIFORNIA Managed Care Pharmacy Systems Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: UCSF Application Deadline: February 25 Starting Date: July 1 Estimated Stipend: $46,000 Onsite Interview: Yes Educational/Special Requirements: PharmD degree from an accredited school of pharmacy, completion of a pharmacy practice residency or equivalent experience, 3 letters of recommendation, letter of intent, and onsite interview Fringe Benefits: Health/dental/vision benefit; 20 days of paid time off, including professional leave (with travel allowances); and 9 holidays; no on-call responsibilities Special Features: This residency instills the philosophy that health care outcomes need to be considered from all relevant perspectives (patient, provider, and payer). Residents participate in the development of drug policy, clinical guidelines, pharmacy benefits, and population-based disease management, pharmaceutical contracting support/analysis, and quality improvement. This program teaches residents to conceptualize, integrate, and transform accumulated experiences and knowledge into improved drug therapy for managed care patients. Contact Information: Tara Abrams Senior Clinical Pharmacist, Quality Improvement Blue Shield of California Pharmacy Services 50 Beale St., 22nd Fl. San Francisco, CA 94105 (415) 229-6424 (415) 229-6011 (fax) [email protected] www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs ■■ CAREMARK, INC. Managed Care Accredited: Length of Program: Number of Positions: Affiliation: AMCP/ASHP 12 months 3 (1 each in Maryland, Texas, and Arizona) University of Maryland, University of Arizona Application Deadline: January 1 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent experience Fringe Benefits: 2 weeks vacation, health insurance, free parking, professional meetings, and other management and pharmaceutical industry experience Special Features: Off-site rotations, university affiliation at MD and AZ sites, ambulatory care clinic, disease management, industry experience Contact Information: Melissa Jay Caremark, Inc. 750 West John Carpenter Fwy. Irving, TX 75039 (469) 524-5832 (469) 524-5858 (fax) [email protected] ■■ CAREMARK, INC. Managed Care Specialty-Analytics and Outcomes Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: University of Illinois at Chicago; Midwestern University-Chicago College of Pharmacy Application Deadline: January 3 Starting Date: July 1 Estimated Stipend: $38,000 Onsite Interview: Yes Educational/Special Requirements: PharmD with experiential or internship-based experience in managed care/PBM industry Fringe Benefits: Comprehensive medical, dental, and life insurance plan; 2-week paid vacation; holidays; employee stock purchase program; flexible spending program; travel budget Special Features: Caremark is a leading pharmaceutical services company, providing comprehensive drug benefit services to approximately 24 million participants throughout the United States. Caremark's clients include corporate health plans, managed care organizations, insurance companies, unions, government agencies, and other funded benefit plans. The www.amcp.org Analytics and Outcomes Residency will provide the resident a unique opportunity to work on initiatives that foster pro-active management of pharmaceutical and overall health care costs. It offers the ability to work with large data sets and perform various pharmaceutical cost analyses such as plan design modeling, formulary analysis, and clinical outcomes. As part of a core sales and account management team, the resident will have the opportunity to interact directly with clients, consultants, and various other benefit providers. While the focus is on analytics, the resident will be exposed to various areas in pharmacy benefit management such as clinical program development and implementation, operations, sales, account management, clinical sales support, marketing and communications, trade relations, pharmaceutical services, and therapeutic services. Contact Information: Anita Allemand Director, Client Analytic Services, Inc. Caremark, Inc. 2211 Sanders Rd. Northbrook, IL 60062 (847) 559-3923 (847) 559-5475 (fax) [email protected] ■■ CLINICAL PHARMACOLOGY SERVICES, INC. Ambulatory Care/Clinical Research Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: February 1 Starting Date: June 1 Estimated Stipend: $34,000 Onsite Interview: Yes Educational/Special Requirements: None Fringe Benefits: Sponsorship to professional meeting and Southwestern Residency Conference Special Features: None Contact Information: Daniel Buffington Director Clinical Pharmacology Services, Inc. 6285 E. Fowler Ave. Tampa, FL 33617 (813) 983-1500 (813) 983-1501 (fax) [email protected] Web: http://www.cpshealth.com Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 623 Managed Care Pharmacy Residencies, Fellowships, and Other Programs ■■ COVENTRY HEALTH CARE OF KANSAS, INC. Pharmacy Benefits Management Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: University of Missouri–Kansas City (UMKC) Application Deadline: January 6 Starting Date: July 1 Estimated Stipend: $32,000 Onsite Interview: Yes Educational/Special Requirements: PharmD degree from an ACPE-accredited college of pharmacy, licensed pharmacist in the United States or eligibility for licensure (successful candidate must be fully licensed at the start of residency), application, letter of intent, curriculum vitae, writing sample, official transcripts, 3 professional letters of recommendation, onsite interview with short formal presentation in Kansas City Fringe Benefits: 2 weeks paid vacation; holidays; health insurance; paid travel and registration to AMCP Educational Conference, AMCP Annual Meeting, and ASHP Midyear Meeting Special Features: Residency opportunities in all managed care core competencies: pharmacy benefit management, utilization management, formulary management, clinical consultation service, drug information, new technologies assessment, provider network relations, Medicare plan interventions, disease state management, health and wellness initiatives, quality assurance/improvement activities, marketing and sales, contracting, health economics/outcomes, rotations with pharmaceutical industry Contact Information: Shawn Burke Director of Pharmaceutical Services Coventry Health Care of Kansas, Inc. 8320 Ward Pkwy. Kansas City, MO 64114 (866) 795-3995 (866) 795-3992 (fax) [email protected] ■■ DEPARTMENT OF VETERANS AFFAIRS MEDICAL CENTER Pharmacy Practice-primary care emphasis Accredited: ASHP Length of Program: 12 months Number of Positions: 3 Affiliation: None Application Deadline: January 13 Starting Date: July 1 Estimated Stipend: $35,000 plus benefits Onsite Interview: Yes 624 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 Educational/Special Requirements: PharmD or equivalent experience Fringe Benefits: Vacation, paid holidays, sick days, and administrative time off for selected meetings Special Features: ASHP-accredited pharmacy practice residency with large emphasis in direct patient care in a primary care setting. The pharmacy resident will work under a collaborative practice agreement with a medical team to facilitate achievement of therapeutic goals through evidence-based disease state management. In addition, the resident will have learning experiences in critical care, internal medicine, practice management, drug policy development, and education and teaching. Upon completion of this residency program, the pharmacy resident will have achieved a variety of advanced practice skills that will enable the graduate to feel confident to function effectively in multiple health care environments and roles. Contact Information: Jo-Ann Caudill Residency Program Director Dept. of Veterans Affairs Medical Center 3200 Vine St. Cincinnati, OH 45220 (513) 475-6322 (513) 475-6981 (fax) [email protected] ■■ GROUP HEALTH COOPERATIVE Managed Care Pharmacy Practice Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 10 Starting Date: July 1 Estimated Stipend: $40,000 Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent experience. Fringe Benefits: Full medical coverage ($35/month paid by the employee) and dental coverage (cost varies depending on the plan chosen by employee) for the resident, 7 days of vacation; paid registration and some fees to attend professional meetings is provided Special Features: The residents are trained in the role of the pharmacist in the development and implementation of clinical practice guidelines, formulary development and management, and drug use policy development. In addition, residents are trained to function as leaders in implementing pharmaceutical care plans for specific patients in a managed care setting. Contact Information: Jim Carlson www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs Director, Pharmacy Administration Group Health Cooperative 12400 E. Marginal Way S. Seattle, WA 98168 (206) 901-4420 (206) 901-4410 (fax) [email protected] Web: http://www.ghc.org/about_gh/employ/rxresidency.jhtml ■■ HARVARD VANGUARD MEDICAL ASSOCIATES Pharmacy Practice with Emphasis in Managed Care Accredited: ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: Massachusetts College of Pharmacy and Health Sciences Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $32,000 Onsite Interview: Yes Educational/Special Requirements: BS in pharmacy or PharmD Fringe Benefits: Comprehensive medical plan, 2 weeks paid vacation Special Features: Academic appointment: instructor of pharmacy practice Contact Information: William McCloskey Massachusetts College of Pharmacy & Health Sciences 179 Longwood Ave. Boston, MA 02115 (617) 732-2167 (617) 732-2244 (fax) [email protected] Web: http://www.mcphs.edu ■■ HEALTHPARTNERS Managed Care Pharmacy Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: Competitive Onsite Interview: Yes Educational/Special Requirements: PharmD Fringe Benefits: Health insurance, vacation, and holidays Special Features: Travel/registration for 1 national meeting www.amcp.org Contact Information: Vyvy Vo Clinical Pharmacy Program Manager HealthPartners 8100 34th Ave. South PO Box 1309 Minneapolis, MN 55440 (952) 967-5133 (952) 883-5875 (fax) [email protected] ■■ HENRY FORD HEALTH SYSTEM Managed Care Pharmacy Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: February 15 Starting Date: July 1 Estimated Stipend: Competitive Onsite Interview: Yes Educational/Special Requirements: PharmD, pharmacy practice residency desirable Fringe Benefits: Health care, 2 weeks paid vacation, travel to 1 meeting Special Features: The resident will design system enhancements and participate in ongoing utilization management, disease management, and compliance intervention programs. He/she will participate in formulary management and design clinical indications of effectiveness for guidelines. The resident will formulate and answer a research question, using scientific principles, with a strong emphasis on outcomes, pharmacoeconomics, and quality-of-life research. Experiences will include exposure to our HMO (Health Alliance Plan) and involvement with the Center for Clinical Effectiveness, health system studies, quality improvement center, and clinical pharmacy services. Direct patient care responsibilities in one of our ambulatory clinics will be ongoing throughout the year. Contact Information: Vanita Pindolia VP, Pharmacy Care Management Henry Ford Health System 30100 Telegraph Rd., Suite 200 Bingham Farms, MI 48025 (248) 723-0206 (248) 642-6094 (fax) [email protected] Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 625 Managed Care Pharmacy Residencies, Fellowships, and Other Programs ■■ HORIZON NJ HEALTH Managed Care Pharmacy Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: Horizon BC Application Deadline: February 1 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent experience; eligibility for New Jersey state licensure Fringe Benefits: Paid vacation, full medical/dental/retirement benefits Special Features: Medicaid managed care HMO, unique focus on government programs, pharmacy case management, formulary and disease state management, development of clinical policies, outcomes research, assist with PharmD student oversight, ambulatory care experience, professional development courses, attendance to at least 1 national conference Contact Information: Samuel Currie Director, Clinical Pharmacy Programs Horizon NJ Health 210 Silvia St. West Trenton, NJ 08628 (609) 538-0700 (609) 538-1698 (fax) [email protected] ■■ HUMANA INC. Managed Care Pharmacy Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 6 Starting Date: July 1 Estimated Stipend: $37,000 Onsite Interview: Yes Educational/Special Requirements: Graduate of an accredited school of pharmacy, PharmD preferred, minimum GPA of 3.2 (on a 4.0 scale), and eligibility for pharmacy licensure Fringe Benefits: Health, dental, 401(k), 3 weeks vacation, 2 floating holidays, relocation allowance, and travel expenses paid for 2 professional meetings Special Features: Humana Inc. has a broad. range of programming within the pharmacy management department. The residency will offer exposure to various aspects of managed care pharmacy as well as direct patient care through clinical rotations at the Veterans Affairs hospital. The residency will offer 626 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 participation in corporate pharmacy and therapeutics committees, manufacturer relations, pharmaceutical contracting and rebating, pharmacy benefit design, consumer relations, legislative and drug policy issues relating to managed care and health benefit design, ePharmacy initiatives, and outcomes analysis. Contact Information: Jane Stacy Clinical Advisor Humana Inc. 500 W. Main St., 16th Fl. Louisville, KY 40202 (502) 580-1591 (502) 508-1591 (fax) [email protected] ■■ IBA HEALTH PLANS/BLUE CARE NETWORK Managed Care Pharmacy Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: Ferris State University/Kalamazoo Center for Medical Studies/Pfizer Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $32,000 Onsite Interview: Yes Educational/Special Requirements: PharmD Fringe Benefits: Health/dental, insurance, 2 weeks vacation, paid holidays Special Features: Travel/registration reimbursement for 1 national conference and the Great Lakes Residency Conference (pharmacy residents conference); no weekend or evening shifts Contact Information: Teresa Klepser Residency Director MSU-KCMS/Ferris 1000 Oakland Dr. Kalamazoo, MI 49008 (269) 337-6392 (269) 337-4474 (fax) [email protected] ■■ INTERMOUNTAIN HEALTH CARE (IHC) HEALTH PLANS Managed Care Pharmacy Practice Accredited: Anticipated, July 2005 Length of Program: 1 year Number of Positions: 1 Affiliation: None Application Deadline: January 10 Starting Date: June 30 www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs Estimated Stipend: $40,000 Onsite Interview: Required Educational/Special Requirements: Candidates must graduate from an ACPE-accredited pharmacy program with a doctor of pharmacy degree with a minimum GPA of 3.0 on a 4.0 scale. Candidate must obtain Utah pharmacist licensure within the first 60 days of the program. Fringe Benefits: Health, dental, and life insurance; staff discounts; 10 days of vacation leave; and 10 holidays Special Features: Travel benefits to either the ASHP Midyear Clinical Meeting or the AMCP Educational Meeting and the Western States Conference Contact Information: Jeffrey Dunn Formulary and Contract Manager IHC Health Plans 4646 W. Lake Park Blvd., Suite N3 Salt Lake City, UT 84120 (801) 442-7984 (801) 442-3006 (fax) [email protected] ■■ KAISER PERMANENTE Managed Care Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 1 Starting Date: July 1 Estimated Stipend: $41,220 (2005-2006) Onsite Interview: Yes Educational/Special Requirements: This is a specialty residency in managed care with emphasis on pharmacy practice management; BS/PharmD Fringe Benefits: Medical benefits, 10 days sick leave/vacation, 5 holidays, $500 for national meeting plus expenses paid for Western States Residency Conference Special Features: Emphasis in administration/practice management Contact Information: Susan Downard Area Pharmacy Manager Kaiser Permanente 16601 E. Centretech Pkwy. Aurora, CO 80011 (303) 326-6764 (303) 739-3574 (fax) [email protected] www.amcp.org ■■ KAISER PERMANENTE Mental Health (Specialty) Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 1 Starting Date: July 5 Estimated Stipend: $41,220 (2005-2006) Onsite Interview: Yes Educational/Special Requirements: PharmD required; completion of pharmacy practice residency or possession of equivalent clinical experience preferred Fringe Benefits: Health insurance, holidays, travel assistance Special Features: Group-model HMO setting Contact Information: Daniel Dugan Clinical Pharmacy Specialist in Mental Health Kaiser Permanente of Colorado 16601 E Centretech Pkwy. Aurora, CO 80011 (303) 467-5776 (303) 467-5805 (fax) [email protected] ■■ KAISER PERMANENTE Primary Care (Specialty) Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 1 Starting Date: July 5 Estimated Stipend: $41,220 (2005-2006) Onsite Interview: Yes Educational/Special Requirements: PharmD required; completion of pharmacy practice residency or possession of equivalent clinical experience preferred Fringe Benefits: Health insurance, holidays, travel assistance Special Features: Group-model HMO setting Contact Information: Rachana Patel Clinical Pharmacy Specialist Kaiser Permanente 1375 East 20th Ave. Denver, CO 80205 (303) 764-4479 (303) 861-3668 (fax) [email protected] Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 627 Managed Care Pharmacy Residencies, Fellowships, and Other Programs [This listing has not been not updated; contact the program for current information.] ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— CALIFORNIA Pharmacy Practice and Drug Information Practice Accredited: ASHP Length of Program: 12 months Number of Positions: 26 Affiliation: None Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: Applicants must be a graduate of an accredited college of pharmacy and be licensed, or be eligible for licensure, in California. Good communication skills required. Resident will develop a project with targeted care outcomes and present at the annual Western States Conference. Fringe Benefits: Medical, dental, and optical insurance; holidays; vacation/sick leave Special Features: Hospital and ambulatory care experiences in the nation’s largest health maintenance organization, preventative and disease state management in an integrated managed care setting Contact Information: Elaine Watanabe Pharmacy Services Manager, Recruitment Kaiser Permanente Pharmacy Operations Services, California Division 9521 Dalen St. Downey, CA 90242 (714) 796-4809 (714) 796-4826 (fax) [email protected] ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— INLAND EMPIRE SERVICE AREA General Pharmacy Practice Accredited: ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: Graduate of an accredited college of pharmacy and be licensed, or be eligibile for licensure, in California; pharmacy school transcript; 3 letters of recommendation; letter of intent; and curriculum vitae 628 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 Fringe Benefits: 2 weeks paid vacation; health benefits, including dental/optical; paid holidays; office space; reimbursement for off-site experiences Special Features: Kaiser Permanente is the nation's largest nonprofit health plan, serving more than 8.4 million members in 9 states. The Inland Empire Service Area consists of 2 medical centers and 14 satellite medical offices for 598,000 members in the San Bernardino and Riverside counties of southern California. Contact Information: Patricia Gray Clinical Operations Manager Kaiser Permanente Inland Empire Service Area 9310 Sierra Ave. Fontana, CA 92335 (909) 427-3838 (909) 427-3830 (fax) [email protected] ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— LOS ANGELES MEDICAL CENTER Pharmacy Practice Accredited: ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 10 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: Applicants must be a graduate of an accredited college of pharmacy and be licensed, or be eligible for licensure, in California. Good communication skills required. Resident will develop a project with targeted care outcomes and present at the annual Western States Conference. Fringe Benefits: Medical, dental, optical insurance; holidays; vacation/sick leave Special Features: The Kaiser Permanente Los Angeles Medical Center is the tertiary care center for Kaiser Permanente in southern California and provides comprehensive inpatient, outpatient, and ambulatory care services to Kaiser Permanente members. This residency program provides development and training for recently graduated pharmacists, with an emphasis on pharmaceutical care and leadership to a diverse community. This program will allow the residents to become familiar with pharmacy practice in an integrated health care program. Contact Information: Steve Litsey Pharmacy Leader-Metro Service Area Kaiser Permanente Pharmacy Operations Services www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— TRI-CENTRAL SERVICE AREA 1515 N. Vermont Ave., Suite 237 Los Angeles, CA 90027 (323) 783-8306 (323) 783-7609 (fax) [email protected] ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— METRO LOS ANGELES Managed Care Accredited: Yes Length of Program: 12 Number of Positions: 2 Affiliation: None Application Deadline: January 10 Starting Date: July 1 Estimated Stipend: 35,000 Onsite Interview: Yes Educational/Special Requirements: Applicants must be a graduate of an accredited college of pharmacy and be licensed, or be eligible for licensure, in California. Good communication skills required. Resident will develop a project with targeted care outcomes and present at the annual Western States Conference. Fringe Benefits: Paid insurance (medical, dental, vision), holidays, vacation/sick days Special Features: The Kaiser Permanente Los Angeles Medical Center is the tertiary care center for Kaiser Permanente in southern California and provides comprehensive inpatient, outpatient, and ambulatory care services to Kaiser Permanente members. This residency program provides development and training for recently graduated pharmacists, with an emphasis on pharmaceutical care and leadership to a diverse community. This program will allow the residents to become familiar with managed care pharmacy practice in an integrated health care program. Contact Information: Elizabeth Oyekan, PharmD Residency Progam Coordinator Kaiser Permanente Medical Care Program—Metro Los Angeles Pharmacy Operations 1515 N. Vermont Ave., Suite 237 Los Angeles, CA 90027 (323) 783-8306 (323) 783-7609 (fax) [email protected] www.amcp.org Managed Care Organization Accredited: ASHP Length of Program: 1 year Number of Positions: 2 Affiliation: None Application Deadline: January 9 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Yes Educational/Special Requirements: Recent graduate of an accredited school of pharmacy and eligibility for California licensure, college transcript(s), letter of intent, 3 letters of recommendation, and curriculum vitae Fringe Benefits: 2 weeks paid vacation; 6 paid holidays; sick leave; health benefits, including dental/optical (also dependents); uniforms; office space; and reimbursement for off-site experiences Special Features: The program provides the resident with an opportunity to participate in the provision of health care in a large managed care environment. Resident will gain experience in various practice areas: acute care, ambulatory care, physician drug education, drug information, outpatient pharmacy, and practice management; ambulatory care practice includes anticoagulation, heart failure, hypertension, integrated CVD, oncology, etc. Residents completing our program will have a strong foundation for future pharmacy practice. Contact Information: John Sie Pharmacy Practice Residency Coordinator, Ambulatory Care Clinical Pharmacy Services Supervisor Kaiser Permanente Medical Care Program Tri-Central Service Area, Pharmacy Operations 1011 Baldwin Park Blvd. Baldwin Park, CA 90706 (626) 851-5602 (626) 851-5813 (fax) [email protected] ■■ KAISER PERMANENTE MEDICAL CARE PROGRAM— WEST LOS ANGELES Pharmacy Practice Accredited: Length of Program: Number of Positions: Affiliation: Application Deadline: Starting Date: Estimated Stipend: Onsite Interview: Vol. 11, No. 7 September 2005 ASHP 12 months 1 None January 15 July 1 $35,000 Yes JMCP Journal of Managed Care Pharmacy 629 Managed Care Pharmacy Residencies, Fellowships, and Other Programs Educational/Special Requirements: Applicants must be a graduate of an accredited college of pharmacy and be licensed, or be eligible for licensure, in California. Good communication skills required. Resident will develop a project with targeted care outcomes and present at the annual Western States Conference Fringe Benefits: Medical, dental, and optical insurance; holidays; vacation/sick leave Special Features: Hospital and ambulatory care experiences in the nation’s largest integrated care organization, preventative and disease state management in an integrated managed care setting. Flexible program molded to the resident’s interests. Contact Information: Michael Cinnamond, PharmD Inpatient Pharmacy Director, Residency Program Director Kaiser Permanente Medical Care Program at West Los Angeles 6041 Cadillac Ave., B 310 Los Angeles, CA 90034 (323) 857-2044 (323) 857-2870 (fax) [email protected] ■■ KAISER PERMANENTE MID-ATLANTIC STATES REGION Managed Care Pharmacy Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 6 Starting Date: July 5 Estimated Stipend: $39,500 Onsite Interview: Yes Educational/Special Requirements: PharmD preferred, pharmacy licensure eligibility in DC, Maryland, or Virginia Fringe Benefits: Medical benefits, selected holidays, sick, vacation and education leave for Midyear Clinical Meeting and Eastern States Conference Special Features: In addition to helping develop new and innovative programs, the resident will participate on the Pharmacy & Therapeutics Committee, teach patient education classes, provide pharmacy staff continuing education, assist in educating pharmacy students, and complete residency projects and presentations. Contact Information: Katrin Fulginiti Director, Managed Care Pharmacy Practice Residency Program Kaiser Permanente-Mid-Atlantic 12201 Plum Orchard Dr. Silver Spring, MD 20904 (301) 572-3330 (301) 572-3399 (fax) 630 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 [email protected] Web: http://www.kaiserpermanente.org ■■ KAISER PERMANENTE OF CALIFORNIA— SPECIALTY RESIDENCY IN DRUG INFORMATION Medical Care/Drug Information Accredited: ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: March 31 Starting Date: July 1 Estimated Stipend: $70,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, pharmacy practice residency program or equivalent experience, excellent communication skills, eligibility for California licensure, 3 letters of recommendation Fringe Benefits: Medical, dental, optical insurance; 10 days time off; attendance at 1 pharmacy conference Special Features: None Contact Information: Mirta Millares Kaiser Permanente 12254 Bellflower Blvd., Suite 106 Downey, CA 90242 (562) 658-3630 (562) 658-3758 (fax) [email protected] ■■ KAISER PERMANENTE OF COLORADO Cardiology/Managed Care Accredited: No Length of Program: 12 months Number of Positions: 3 Affiliation: None Application Deadline: January 1 Starting Date: July 1 Estimated Stipend: $41,220 Onsite Interview: Yes Educational/Special Requirements: PharmD required; prior pharmacy practice residency or equivalent preferred Fringe Benefits: Health benefits, travel support to ASHP Midyear meeting and Western States Residency Conference Special Features: This cardiology specialty residency provides unique opportunities to acquire advanced knowledge and skills in ambulatory management of anticoagulation, cardiac risk reduction, and heart failure in a managed care setting. Numerous educational opportunities exist. Experience in evaluating clinical and financial outcomes will be emphasized. www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs Completion of a residency research project required. Contact Information: Anne Denham, PharmD Clinical Pharmacy Specialist Kaiser Permanente of Colorado Clinical Pharmacy Cardiac Risk Service 16601 E Centretech Pkwy. Aurora, CO 80011 (303) 326-7663 (303) 326-7670 (fax) [email protected] ■■ KAISER PERMANENTE OF COLORADO Mental Health/Managed Care Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 1 Starting Date: July 1 Estimated Stipend: $41,220 Onsite Interview: Yes Educational/Special Requirements: PharmD required, prior pharmacy practice residency or equivalent practical experience preferred Fringe Benefits: Health benefits, travel support to ASHP Midyear Clinical Meeting and Western States Residency Conference Special Features: Experiences include inpatient and outpatient psychiatric pharmacy practice sites among a full range of behavioral health conditions as well as neurology and chemical dependence. Evaluation of clinical outcomes is emphasized and completion of residency research is required. Contact Information: Daniel Dugan Clinical Pharmacy Specialist in Mental Health Kaiser Permanente of Colorado 16601 E Centretech Pkwy. Aurora, CO 80011 (303) 467-5776 (303) 467-5805 (fax) [email protected] ■■ KAISER PERMANENTE OF COLORADO Primary Care/Managed Care Accredited: No Length of Program: 12 months Number of Positions: 3 Affiliation: None Application Deadline: January 1 www.amcp.org Starting Date: July 1 Estimated Stipend: $41,220 Onsite Interview: Yes Educational/Special Requirements: PharmD required; prior pharmacy practice residency or equivalent experience preferred Fringe Benefits: Health benefits, travel support to ASHP Midyear Clinical Meeting and Western States Residency Conference Special Features: This primary care/managed care specialty residency provides the unique opportunity to acquire advanced knowledge and skills in ambulatory care pharmacotherapy in a managed care setting. Practice sites include primary care clinics, centralized services (anti-coagulation, cardiac risk), and specialty services (diabetes, cardiology, infectious disease, mental health, nephrology). Residents participate in formulary management, pharmacoeconomic evaluations, and target drug programs. Numerous educational opportunities exist. Experience in evaluating clinical and financial outcomes will be emphasized. Completion of residency research project required. Contact Information: Rachana Patel Clinical Pharmacy Specialist Kaiser Permanente 1375 East 20th Ave. Denver, CO 80205 (303) 764-4479 (303) 861-3668 (fax) [email protected] ■■ KAISER PERMANENTE OF GEORGIA Managed Care Pharmacy Practice Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 13 Starting Date: July 1 Estimated Stipend: $34,000 Onsite Interview: Yes Educational/Special Requirements: PharmD preferred; eligibility for Georgia licensure Fringe Benefits: Medical/dental/vision insurance, holidays, vacation/ sick leave Special Features: Ambulatory care, acute care, drug information, and administration Contact Information: Suzanne Booth Residency Program Coordinator Kaiser Permanente 200 Crescent Centre Pkwy. Tucker, GA 30083 Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 631 Managed Care Pharmacy Residencies, Fellowships, and Other Programs (770) 496-3653 (770) 496-3487 (fax) [email protected] ■■ KELSEY-SEYBOLD CLINIC/UNIVERSITY OF HOUSTON Pharmacy Practice with an Emphasis on Managed Care Accredited: ASHP/AMCP Length of Program: 1 year Number of Positions: 2 Affiliation: University of Houston Application Deadline: January 5 Starting Date: July 1 Estimated Stipend: 30,000 Onsite Interview: Yes Educational/Special Requirements: PharmD or equivalent experience Fringe Benefits: State of Texas benefits, 10 days vacation, support to attend AMCP, ASHP, TSHP, and either Alcalde or Mid-West Residency Conference Special Features: Ambulatory-care-focused administrative and clinical responsibilities, collaborations with other UH residents Contact Information: Sarah Lake-Wallace, PharmD Kelsey-Seybold Clinic Pharmacy Administration 8900 Lakes at 610 Dr. Houston, TX 77054 (713) 442-6248 (713) 442-5253 (fax) [email protected] ■■ MEDCO HEALTH SOLUTIONS, INC. Managed Care Pharmacy Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: January 5 Starting Date: June 26 Estimated Stipend: $34,000 Onsite Interview: Yes Educational/Special Requirements: PharmD degree, eligibility for New Jersey licensure, managed care rotation or experience preferred Fringe Benefits: Complete medical coverage, paid holidays and vacation Special Features: Professional development trainings, travel and attendance at educational meetings 632 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 Contact Information: Doris Fishman Senior Director, Clinical Therapeutics Medco Health Solutions, Inc. 100 Parsons Pond Dr. Mail Stop F2-3 Franklin Lakes, NJ 07417 (201) 269-6159 (201) 260-1035 (fax) [email protected] ■■ NOVARTIS PHARMACEUTICALS CORPORATION Health Economics & Outcomes Research Fellowship Accredited: No Length of Program: 2 years (Rutgers site, flexible) Number of Positions: 3 Affiliation: Duke University; Scott & White Health Plan/University of Texas at Austin; Rutgers University Application Deadline: December 31 Starting Date: July 1 Estimated Stipend: $35,000-$44,000 Onsite Interview: Yes Educational/Special Requirements: Advanced degree in health services research, public health, health policy, pharmacy, economics, medicine, or other related areas, with some experience in outcomes research Fringe Benefits: Medical insurance, vacation Special Features: The fellows will gain familiarity with outcomes research principles/application and experience in designing research studies that examine economic, clinical, and humanistic outcomes. The first year is spent at an academic/managed care institution and the second year with Novartis’s Health Economics & Outcomes Research Department. Contact Information: Feride Frech Director, Health Economics & Outcomes Research Novartis Pharmaceuticals Corporation One Health Plaza East Hanover, NJ 07936-1080 (862) 778-5094 (973) 781-3018 (fax) [email protected] ■■ OPTIMA HEALTH PLAN/ SENTARA HEALTHCARE Managed Care Accredited: Not at this time Length of Program: 12 months Number of Positions: 1 Affiliation: None www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs Application Deadline: February 15 Starting Date: July 1 Estimated Stipend: $31,720 Onsite Interview: Yes Educational/Special Requirements: PharmD from an accredited school of pharmacy or equivalent experience, onsite interview, and eligibility for Virginia licensure Fringe Benefits: 2 weeks of paid vacation and uninterrupted stipend during minor illness, a health insurance plan, travel assistance for continuing education events and other professional activities Special Features: Resident will have exposure to an integrated health care system and system-wide pharmacy services, including a drug information center. The resident will have the opportunity to precept students from Virginia Commonwealth University and Hampton University schools of pharmacy and work with medical residents from the Eastern Virginia Medical School. Contact Information: Elizabeth Brusig, PharmD Clinical Pharmacist Optima Health Plan 4417 Corporation Lane Virginia Beach, VA 23464 (757) 552-7519 (757) 552-7516 (fax) [email protected] ■■ ORTHO-MCNEIL JANSSEN SCIENTIFIC AFFAIRS, LLC Pharmaceutical Industry, Department of Outcomes Research Accredited: No Length of Program: 24 months Number of Positions: 1 Affiliation: Thomas Jefferson University Application Deadline: December 31 Starting Date: June 2006 (flexible) Estimated Stipend: Competitive Onsite Interview: Yes Educational/Special Requirements: Contact program Fringe Benefits: Contact program Special Features: Fellows spend the first year of the program working on outcomes research projects in an academic setting and the second year in the pharmaceutical industry. Fellows have the opportunity to take coursework in biostatistics, epidemiology, economics, and other outcomes-related subjects. Contact Information: Julie Locklear, PharmD, MBA Associate Director, Outcomes Research Ortho-McNeil Janssen Scientific Affairs, LLC 1125 Trenton-Harbourton Rd. Titusville, NJ 08560 (609) 730-3664 (609) 730-2556 (fax) [email protected] ■■ ORTHO-MCNEIL JANSSEN SCIENTIFIC AFFAIRS, LLC Drug Information Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: None Application Deadline: December 31 Starting Date: July 1 Estimated Stipend: Competitive Onsite Interview: Yes Educational/Special Requirements: Contact program Fringe Benefits: Contact program Special Features: Contact program Contact Information: Will Zachok Associate Director, Medical Communications Ortho-McNeil Janssen Scientific Affairs, LLC 1000 Route 202 S Raritan, NJ 08869 (908) 218-6283 (908) 722-6402 (fax) [email protected] ■■ OUTCOMES PHARMACEUTICAL HEALTH CARE Medication Therapy Management Firm Accredited: ASHP/AMCP Length of Program: 1 year Number of Positions: 1 Affiliation: University of Iowa Application Deadline: December 31 Starting Date: July 5 Estimated Stipend: $33,000 Onsite Interview: Yes Educational/Special Requirements: Candidates for residency must possess a doctor of pharmacy degree from a school of pharmacy accredited by the American Council on Pharmaceutical Education (ACPE) and be eligible for licensure in Iowa. Fringe Benefits: For additional information, contact the Outcomes office at (515) 237-0001 or [email protected]. Special Features: Outcomes’ clinical services residency is a 12-month managed care systems residency program. This residency offers a unique learning experience on how pharmacists can impact the quality and rising cost of health care by providing patient-oriented, care-based services. Residency activities include account management, marketing and communications, pharmacoeconomic data collection and analysis, claims assessment, development and implementation of disease management www.amcp.org Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 633 Managed Care Pharmacy Residencies, Fellowships, and Other Programs programs, development and administration of research projects, and training of pharmacists and support staff. The resident also serves as a preceptor to PharmD clerkship students. Contact Information: Patty Kumbera Chief Operations Officer Outcomes Pharmaceutical Health Care 601 E. Locust, Suite 200 Des Moines, IA 50309 (515) 237-0001 (515) 237-0002 (fax) [email protected] Web: http://www.getoutcomes.com ■■ PHARMACARE Pharmacy Benefits Management Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: University of Pittsburgh School of Pharmacy Application Deadline: January 12 Starting Date: July 1 Estimated Stipend: $31,000 Onsite Interview: Yes Educational/Special Requirements: PharmD Fringe Benefits: Comprehensive benefits package, no weekends/holidays, attend professional managed care meetings Special Features: The University of Pittsburgh School of Pharmacy and EHS, one of the nation’s largest pharmacy chainbased prescription management firms, offers an opportunity to practice in a dynamic PBM environment and gain a clinical and administrative perspective in managed pharmacy benefit plans for a wide variety of clients. Multifaceted experience will include DUR criteria development, clinical intervention activities, P&T activities, clinical systems development, and new business development/client services/ marketing support. Contact Information: Teddi Gianangeli Senior Clinical Management Specialist Pharmacare 620 Epsilon Dr. Pittsburgh, PA 15238 (412) 967-2300 ext. 5045 (412) 968-2676 (fax) [email protected] 634 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 [This listing has not been not updated; contact the program for current information.] ■■ PHARMACEUTICAL CARE NETWORK Managed Care Pharmacy Practice Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: None Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $37,500 Onsite Interview: Yes Educational/Special Requirements: Graduate of an accredited college of pharmacy and licensed, or be eligible for licensure, in California Fringe Benefits: Paid vacation and sick leave, health/dental/ vision benefits, and educational support to attend professional meetings Special Features: This residency will provide training in formulary management, provider drug therapy education, drug benefit design, outcomes analysis, and prior authorization. The resident will participate in drug utilization review, P&T committee presentations, and interface with pharmacists from a variety of professional fields. The resident will learn to use PCN’s MedIntelligence software to identify drug therapy problems and make appropriate interventions. This program includes rotations in direct patient care and drug information and an opportunity to rotate through the California Pharmacists Association. Contact Information: Philip Parsatoon Vice President, Professional Services Pharmaceutical Care Network 9343 Tech Center Dr., Suite 200 Sacramento, CA 95826-2563 (916) 361-4450 (916) 414-4650 (fax) [email protected] ■■ PHARMACY MANAGEMENT CONSULTANTS/ OKLAHOMA UNIVERSITY Managed Care Pharmacy Accredited: ASHP-Managed Care Pharmacy Systems Length of Program: 12 months Number of Positions: 1 Affiliation: Oklahoma University College of Pharmacy Application Deadline: February 1 Starting Date: July 1 Estimated Stipend: $35,000 Onsite Interview: Preferred www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs Educational/Special Requirements: Pharmacy degree Fringe Benefits: 10 days vacation, health and dental coverage, faculty appointment at the level of clinical instructor, sponsorship to at least 1 professional meeting Special Features: Evaluate health plan formularies, develop clinical guidelines, evaluate retrospective and prospective DUR criteria, develop educational intervention programs, serve as preceptor for pharmacy students; program includes experience working with dual Medicare/Medicaid eligibles and with the state PDP plan. Contact Information: Elgene Jacobs Pharmacy Management Consultants 1122 N.E. 13th St. ORI W-4403 Oklahoma City, OK 73117 (405) 271-9039, ext. 47354 (405) 271-2615 (fax) [email protected] ■■ PRESCRIPTION SOLUTIONS Managed Care Accredited: ASHP Length of Program: 12 months Number of Positions: 1 or 2 Affiliation: None Application Deadline: January 7 Starting Date: July 5 Estimated Stipend: $40,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, license to practice in California Fringe Benefits: 2 weeks paid vacation, paid holidays, sick days, travel reimbursement Special Features: Drug therapy management, developing clinical guidelines for appropriate drug use, formulary management, legal and regulatory affairs; professional meetings and seminars: required to attend the ASHP Midyear Clinical Meeting, CSHP Seminar, AMCP Educational Conference, and Western States Conference Contact Information: Alex Gilderman Director, Clinical Pharmacy Services c/o Prescription Solutions 3515 Harbor Blvd. Costa Mesa, CA 92626 (714) 825-6768 (714) 825-3742 (fax) [email protected] www.amcp.org ■■ RUTGERS UNIVERSITY/HORIZON BLUE CROSS BLUE SHIELD OF NEW JERSEY Managed Care Organization Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: Ernest Mario School of Pharmacy, Rutgers University, State University of New Jersey Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $30,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, eligibility for New Jersey state license Fringe Benefits: Full medical coverage, dental and retirement benefits Special Features: No weekend or staffing requirements, teaching and preceptoring, mandatory, graduate courses offered at Rutgers University, industry and PBM perspectives of managed care Contact Information: Saira A. Jan Associate Professor, Rutgers University Associate Director, Pharmacy Management Horizon Blue Cross Blue Shield of NJ Three Penn Plaza East, PP-13Q Newark, NJ 07105 (973) 466-4575 (973) 466-6266 (fax) [email protected] [This listing has not been not updated; contact the program for current information.] ■■ SCOTT & WHITE Pharmacy Practice Managed Care Accredited: No Length of Program: 1 year Number of Positions: 1 Affiliation: University of Texas at Austin Application Deadline: February 15 Starting Date: July 1 Estimated Stipend: $30,000 Onsite Interview: Yes Educational/Special Requirements: PharmD, eligibility for Texas state license Fringe Benefits: Full medical coverage and dental Special Features: Minimal weekend staffing requirements; teaching and preceptoring, if desired; industry and PBM perspectives of managed care; marketing and sales exposure Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 635 Managed Care Pharmacy Residencies, Fellowships, and Other Programs (520) 629-4700 (fax) [email protected] Contact Information: John Jackimiec Director of Pharmacy Managed Care Scott & White 2601 Thornton Lane, Suite A Temple, TX 76502 (254) 742-3144 (254) 742-3109 (fax) [email protected] [This listing has not been not updated; contact the program for current information.] ■■ SOUTHERN ARIZONA VA HEALTH CARE SYSTEM Pharmacy Practice Accredited: ASHP since 1982 Length of Program: 12 months Number of Positions: 7 Affiliation: University of Arizona Application Deadline: January 16 Starting Date: July 1 Estimated Stipend: $33,000 plus benefits Onsite Interview: Interview required, onsite preferred Educational/Special Requirements: Curriculum vitae, 3 letters of recommendation, college transcripts, 1 writing sample, U.S. citizenship Fringe Benefits: Residents accumulate 13 days of paid vacation time during the residency year. Sick leave is accrued at the rate of 4 hours every 2 weeks. Educational leave is provided to attend the Arizona Society of Health-Systems Pharmacists Annual Meeting, the ASHP Midyear Clinical Meeting, and the Western States Conference for Pharmacy Residents, Fellows, and Preceptors. Travel funds are available to offset some of the expenses for these educational meetings. Residents have access to the same medical and dental insurance plans that are offered to full-time employees of the VA. These include a wide range of HMO and PPO health plans. Tucson also has a wide range of outdoor activities and excellent weather year round. Special Features: This comprehensive residency includes a balance of inpatient and outpatient clinical pharmacy experiences. Residents work in the areas of internal medicine, cardiology, neurology clinics, primary care, geriatrics, hospice, surgery/nutritional support, mental health, practice management, DUE, Pharmacy and Therapeutics Committee activities, drug literature evaluation, drug policy development, teaching, research, and enhancing communication abilities. Two months of elective experience are available. Contact Information: William Jones, MS, RPh Southern Arizona Veterans Administration Health Care System 3601 South Sixth Ave. Pharmacy Service 5-119 Tucson, AZ 85723 (520) 792-1450, ext. 6721 636 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 ■■ UNITED DRUGS Managed Care Accredited: Not at this time Length of Program: 1 year Number of Positions: 1 Affiliation: AMCP Application Deadline: June 15 Starting Date: September 1 Estimated Stipend: $33,000 Onsite Interview: Yes Educational/Special Requirements: Completion from an accredited pharmacy school, PharmD preferred; ability to work independently Fringe Benefits: Understanding of managed care from the perspective of a PBM, product development opportunities, and ability to live in Arizona for a year Special Features: Work hours negotiable Contact Information: Jean Brown Director of Clinical Services United Drugs 7227 North 16th St., Suite 160 Phoenix, AZ 85020-5256 (602) 678-1179, ext. 229 (602) 678-0772 (fax) [email protected] [This listing has not been not updated: contact the program for current information.] ■■ UNIVERSITY AT BUFFALO Ambulatory Care Accredited: Pending Length of Program: 1 year Number of Positions: 2 Affiliation: University Application Deadline: October 1 Starting Date: January 1 Estimated Stipend: Contact program Onsite Interview: Required Educational/Special Requirements: PharmD or equivalent experience required Fringe Benefits: Health, dental, vision insurance; paid vacation; conference travel funds Special Features: Unique experience designed to further refine skills in pharmaceutical care in addition to developing skills in www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs program development and personnel and resource management. Resident is involved in coordination of clinical activities in a high-volume lobby-based pharmacy and a health clinic based ambulatory care pharmacy. The resident will participate in medication histories, adherence counseling, and education programs. Development and implementation of disease management initiatives, patient education, medical informatics and supervision of PharmD students are also significant aspects of the program. This residency will allow ample latitude for the resident to explore interests and further develop skills as a practitioner. The resident will also be appointed as a clinical instructor at the University at Buffalo, School of Pharmacy and Pharmaceutical Sciences. Contact Information: Gene Morse, PharmD University at Buffalo School of Pharmacy and Pharmaceutical Sciences 311 Hochstetter Hall Buffalo, NY 14260 (716) 645-2828 (716) 645-2886 (fax) [email protected] ■■ UNIVERSITY OF ILLINOIS AT CHICAGO & WALGREENS HEALTH INITIATIVES Fellowship— Outcomes Research Accredited: No Length of Program: 2 years Number of Positions: 1 Affiliation: University of Illinois & Walgreens Health Initiatives Application Deadline: February 1 Starting Date: July 1 Estimated Stipend: $36,000 Onsite Interview: Educational/Special Requirements: Applicants should have a PharmD or MD (or equivalent) and have completed a pharmacy practice or managed care residency Fringe Benefits: _________________________________ Special Features: This is a 2-year fellowship jointly offered by Walgreens Health Initiatives and the Center for Pharmacoeconomic Research at the University of Illinois at Chicago. The aim of the program is to train clinical pharmacists to conduct research in drug therapy outcomes and pharmacoeconomics in the managed care setting. Knowledge and experience will be gained in the use of research tools to evaluate economic, humanistic, and clinical outcomes of drug therapy. Presentation and publication of research findings in peer-reviewed venues is expected. The fellowship is designed to facilitate career opportunities in managed care, health provider organizations, consulting, academia, or the pharmaceutical www.amcp.org industry. Contact Information: Glen Schumock, PharmD, MBA University of Illinois at Chicago Center for Pharmacoeconomic Research 833 S. Wood St. (MC 886) Chicago, IL 60612 (312) 996-7961 [email protected] ■■ UNIVERSITY OF MARYLAND SCHOOL OF PHARMACY/ CAREFIRST BLUECROSS BLUESHIELD Managed Care Pharmacy Accredited: No Length of Program: 12 months Number of Positions: 1 Affiliation: University of Maryland Application Deadline: January 5 Starting Date: July 1 Estimated Stipend: 31,500 Onsite Interview: Yes Educational/Special Requirements: Graduate degree in pharmacy Fringe Benefits: Health insurance, parking, support for national meeting attendance and poster presentation Special Features: Appointment as a clinical instructor at the University of Maryland School of Pharmacy, ambulatory care clinics at HMO, office with computer/references at managed care organization Contact Information: Catherine Cooke Clinical Assistant Professor University of Maryland School of Pharmacy 5106 Bonnie Branch Rd. Ellicott City, MD 21043 (410) 480-5012 (410) 480-5296 (fax) [email protected] ■■ UNIVERSITY OF TEXAS MEDICAL BRANCH CORRECTIONAL MANAGED CARE Pharmacy Practice with an Emphasis in Managed Care Accredited: ASHP Length of Program: 12 months Number of Positions: 1 Affiliation: UTMB Application Deadline: February 15 Starting Date: July 1 Estimated Stipend: Contact program Onsite Interview: Yes Educational/Special Requirements: Pharmacy degree from Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 637 Managed Care Pharmacy Residencies, Fellowships, and Other Programs accredited college of pharmacy, Texas license or eligibility for Texas licensure Fringe Benefits: UTMB is an AA/EO employer; closed major holidays and weekends; generous vacation, holiday, and sick leave; competitive benefits Special Features: Program strengths include automated technology, telemedicine technology, ambulatory care and managed care Contact Information: Stephanie Zepeda Assistant Director of Pharmacy UTMB Correctional Managed Care 2400 Ave. I Huntsville, TX 77340 (936) 437-5363 (936) 437-5311 (fax) [email protected] ■■ UPMC HEALTH PLAN Managed Care Accredited: No Length of Program: 12 months Number of Positions: 2 Affiliation: UPMC Health System Application Deadline: May 20 Starting Date: July 1 Estimated Stipend: $32,000 Onsite Interview: Preferred Educational/Special Requirements: Applicants must have completed a doctor of pharmacy degree and be eligible for Pennsylvania licensure. A pharmacy practice residency is recommended but not required for consideration. Please include letter of intent, curriculum vitae, list of references (3) with contact information, and 3 letters of recommendation. Fringe Benefits: Health, dental, eye care, life, and disability insurance is available; vacation time is allotted and professional travel and stipend is available Special Features: UPMC Health Plan is the second largest health insurer in western Pennsylvania, covering commercial, medical assistance, and Medicare populations; it has integrated resources from Community Care Behavioral Health, University of Pittsburgh School of Pharmacy, and the UPMC Health System. Contact Information: Rae Ann Maxwell Director, Pharmacy Services UPMC Health Plan One Chatham Center, 5th Fl. 112 Washington Pl. Pittsburgh, PA 15219 (412) 454-5276 638 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 (412) 454-5295 (fax) [email protected] ■■ VA SAN DIEGO HEALTHCARE SYSTEM Managed Care Pharmacy Systems Accredited: Application submitted Length of Program: 12 months Number of Positions: 1 Affiliation: University of the Pacific School of Pharmacy Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $38,000 Onsite Interview: Yes Educational/Special Requirements: PharmD plus first-year residency or equivalent experience, U.S. citizenship, personal statement, curriculum vitae, transcripts, and 3 letters of recommendation Fringe Benefits: 12-13 vacation and sick leave days, 11 federal holidays, health and life insurance, free parking, paid leave, tuition and travel to required events, office with up-to-date computer systems Special Features: This second-year residency will provide the skills necessary for the practical application of pharmacoeconomic principles to formulary management and outcomes research in integrated health care systems. Education will include formal pharmacoeconomics training classes and hands-on application of principles. Work activities will encompass the VASDHS, Veterans Integrated Service Network 22 (VISN 22) pharmacy benefits management (i.e. Southern California Regional VA), and VA national formulary tasks. Out-of-state and in-state travel is required. A university-affiliated, teaching, integrated healthcare system with 100% computerized medical records, cutting-edge patient safety, pharmacy-managed clinics, pharmacist specialty practices, and a dedicated Pharmacy Health Outcomes Division with 4 full-time pharmacoeconomists. The pharmacy service has outstanding leadership with a long positive track record for innovation and excellence at the local, state, and national levels and a well-trained, well-published staff, most with residencies. This is an exciting opportunity to learn, and to start your career. Contact Information: Anthony P. Morreale, PharmD, MBA, BCPS Chief, Pharmacy Service VA San Diego Healthcare System Pharmacy Service (119) 3350 La Jolla Village Dr. San Diego, CA 92161 (858) 552-8585, ext.. 3026 (858) 552-4391(fax) www.amcp.org Managed Care Pharmacy Residencies, Fellowships, and Other Programs [email protected] Web: http://www.san-diego.med.va.gov [This listing has not been not updated; contact the program for current information.] ■■ WALGREENS HEALTH INITIATIVES Pharmacy Benefit Manager Accredited: AMCP/ASHP Length of Program: 12 months Number of Positions: 2 Affiliation: University of Illinois, Midwestern University–Chicago College of Pharmacy Application Deadline: January 7 Starting Date: July 5 Estimated Stipend: $36,000 Onsite Interview: Yes Educational/Special Requirements: PharmD Fringe Benefits: Medical plan, 2-week vacation, holidays, travel expense budget Special Features: This managed care pharmacy residency program is designed to allow the residents to work within the various departments of a pharmacy benefits management firm, including, but not limited to, care management, drug use policy, PBM operations, clinical sales, and specialty pharmacy. The residents will gain practical experience and develop skills related to disease management, health outcomes, medication management strategies, formulary management, drug utilization review, drug information, and other clinical services. Additionally, residents will have the opportunity to gain exposure to the pharmaceutical industry, be involved in professional organizations, and precept pharmacy students. Contact Information: Susie Min Manager, Clinical Education and Shared Faculty Walgreens Health Initiatives 1417 Lake Cook Rd., MS# L457 Deerfield, IL 60015 (847) 964-6740 (847) 374-2669 (fax) [email protected] www.amcp.org ■■ WELLPOINT PHARMACY MANAGEMENT Managed Care Pharmacy Accredited: ASHP/AMCP Length of Program: 12 months Number of Positions: 1 Affiliation: Blue Cross of California, PrecisionRx, University of Southern California Application Deadline: January 15 Starting Date: July 1 Estimated Stipend: $50,000 Onsite Interview: Yes Educational/Special Requirements: PharmD from an ACPEaccredited college of pharmacy or equivalent experience, eligibility for California licensure, good academic standing, excellent written and verbal communication skills Fringe Benefits: Health insurance; 2 weeks paid vacation; paid holiday and sick days; attendance at the Western States Conference, at a national pharmacy organization meeting, and at a WellPoint Pharmacy Management National P&T meeting Special Features: The program is designed to provide the resident with an overall managed care experience. The resident rotates through several areas within the PBM, including drug information, therapy management, clinical account management, clinical product development, clinical intervention and prior authorization centers, mail-order and specialty pharmacy, health informatics and policy research, and pharmaceutical contracting and industry relations. The program also includes rotations at the Blue Cross of California health plan and the University of Southern California direct-patient-care sites. Contact Information: Krista Yokoyama Residency Program Director WellPoint Pharmacy Management 8407 Fallbrook Ave. MS AF-7 West Hills, CA 91304 (818) 313-5082 (818) 313-5110 (fax) [email protected] Vol. 11, No. 7 September 2005 JMCP Journal of Managed Care Pharmacy 639
