Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
AIS’s Management Insight Series The Challenges of Pharmacy Star Ratings: Solutions for Part D Plan Sponsors and Their Partners Adapted from an AIS Webinar presented by Vivien Chan, Pharm.D. Director of Quality and Cost Strategies OmedaRx Kristine Walhof Program Director, Medicare Star Ratings Cambia Health Solutions, Inc. Edited by Erin Trompeter, Managing Editor, AIS GC6A02 AIS’s Management Insight Series is designed to provide practical solutions to complex business challenges with the help of the industry’s most insightful advisors and managers. See a full list of titles in this series at http://aishealth.com/marketplace/insight-series. Other Related Publications from AIS Medicare Advantage News Medicare-Medicaid Dual Eligibles Database AIS’s Medicare and Medicaid Market Data Drug Benefit News Health Plan Week Call 800-521-4323, or visit the MarketPlace at www.AISHealth.com, for a catalog of AIS books, newsletters, Webinars, Web and looseleaf services, and other information products. This publication is designed to provide accurate, comprehensive and authoritative information on the subject matter covered. However, the opinions contained in this publication are those solely of the authors and not the publisher. The publisher does not warrant that information contained herein is complete or accurate. This book is published with the understanding that the publisher is not engaged in rendering legal or other professional services. If legal advice or other expert assistance is required, the services of a competent person should be sought. ISBN: 978-1-939721-21-1 Copyright © 2016 by Atlantic Information Services, Inc. All rights reserved. No part of this publication may be reproduced, stored on a retrieval system or transmitted by any means, electronic or mechanical, including photocopying and transmittal by FAX, without the prior written permission of Atlantic Information Services, Inc. For information regarding individual or bulk purchases, contact Atlantic Information Services, Inc., 1100 17th Street, NW, Suite 300, Washington, DC 20036 (800-521-4323; 202-775-9008). Table of Contents Introduction������������������������������������������������������������������������������������������������������������������������������� 1 The Regence Approach to Stars Improvement������������������������������������������������������������������������������������ 3 Improving Medication Adherence�������������������������������������������������������������������������������������������������������� 5 Interventions and Programs������������������������������������������������������������������������������������������������������������������� 9 Q&A Session���������������������������������������������������������������������������������������������������������������������������� 15 Appendix A: AIS Coverage of Medicare Advantage Star Ratings�������������������������������� 19 Results for 2016 Plans���������������������������������������������������������������������������������������������������������������������������� 19 More MA Plans Get Four-Star Ratings for ’16, but Scores of Humana, WellCare Dip��������������������� 19 2013 to 2016 Overall Star-Rating Distribution for MA-PD Contracts��������������������������������������������� 20 Counter to Recent Stars Trend, Local PPOs Fared Better Than HMOs��������������������������������������������� 23 Plans With 4.5 Stars or More��������������������������������������������������������������������������������������������������������������� 24 Plans With 2.5 Stars or Fewer������������������������������������������������������������������������������������������������������������� 26 Percentage of Plans In Each Star Rating Level������������������������������������������������������������������������������������ 26 2016 Part D Star Ratings for Plans Serving Dual Eligibles��������������������������������������������������������������� 27 CMS Regulation and Guidance������������������������������������������������������������������������������������������������������������ 34 CMS Suspension of Automatic Star-Rating Cuts for Sanctioned Plans Stuns Observers����������������� 34 Star Ratings for Duals����������������������������������������������������������������������������������������������������������������������� 37 Inovalon Study Shows Duals’ Characteristics Yield Poor Health Outcomes�������������������������������������� 37 Personal Engagement, Communication Technology Could Boost Star Ratings and Health Outcomes for Duals Plans�������������������������������������������������������������������������������������������������������� 37 Diabetes Management Could Help Boost Stars for Struggling Duals Plans�������������������������������������� 38 Stars Memo Offers Two Complex Options to Aid Plans Serving Low-SES Members����������������������� 40 Trade Groups Cite Doubts About CMS’s Two Options for Adjusting Stars for SES������������������������� 43 Impact on Performance ������������������������������������������������������������������������������������������������������������������������ 47 Contract Enrollment for Plans Whose Star Ratings Increased in 2016���������������������������������������������� 47 Contract Enrollment for Plans Whose Star Ratings Decreased in 2016��������������������������������������������� 49 Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016����������������������������������� 51 Appendix B: CMS Fact Sheet — 2016 Star Ratings���������������������������������������������������������� 57 Appendix C: CMS Guidance and Documentation for 2017 Star Ratings�������������������� 75 Appendix D: Medicare 2016 Part C & D Display Measure Technical Notes������������� 121 The Challenges of Pharmacy Star Ratings • 1 Introduction Star ratings on numerous Medicare Part D pharmacy measures continue to present problems for plan sponsors, especially for stand-alone Prescription Drug Plans (PDPs), which overall fared worse for 2016 than for 2015 on the quality scores. Factors that complicate Part D measures include CMS’s changing methodologies for measuring results, the need to involve PBMs and pharmacies, and the importance of ensuring the appropriate dissemination of data among these entities. For the 2016 star ratings, the Regence Blue Cross and Blue Shield-branded Part D plans operated by Cambia Health Solutions, Inc. made significant gains in pharmacy metrics, including improvements on managing high-risk medications, and five-star ratings on cholesterol and hypertension adherence. But many plans, particularly PDPs, continue to struggle on the adherence measures. This report, The Challenges of Pharmacy Star Ratings: Solutions for Part D Plan Sponsors and Their Partners, discusses the steps Cambia and its PBM subsidiary, OmedaRx, took to achieve high ratings on these and other Part D measures. The publication provides valuable, practical intelligence and training on how insurers and PBMs can: u Use a cross-functional, holistic collaborative approach via a population-management model to manage member health and improve star ratings across measures. u Improve Part D measures for gains in star ratings through better member outcomes. u Influence Part D improvements through a focus on Part C measures. u Use incentives and partnerships with providers to drive gains. This report, The Challenges of Pharmacy Star Ratings: Solutions for Part D Plan Sponsors and Their Partners, was updated from a spring 2016 virtual conference sponsored by Atlantic Information Services, Inc., “Medicare Star Ratings: New Strategies to Match CMS’s New Goals,” AIS’s 3rd annual virtual conference on Medicare star ratings. This publication is adapted from the session, “A Coordinated PBM-Health Plan Approach to Improving Part D Star Ratings,” given by Vivien Chan, Pharm.D., director of quality and cost strategies, OmedaRx, and Kristine Walhof, program director, Medicare star ratings, Cambia. Both speakers answered tough questions from conference attendees after their session. Appendices filled with relevant background coverage of industry developments and other resources give insight into the evolving world of MA star ratings. They include: u Appendix A: AIS Coverage of Medicare Advantage Star Ratings, a collection of AIS articles featuring developments and strategies regarding star ratings from AIS’s biweekly newsletter Medicare Advantage News and data from AIS’s health data division. u Appendix B: CMS Fact Sheet - 2016 Star Ratings, a CMS document detailing findings and trends from 2016 star ratings. 2 • AIS’s Management Insight Series u Appendix C: CMS Guidance and Documentation for 2017 Star Ratings, a set of CMS reports and guidance preparing carriers for 2017 star ratings. u Appendix D: Medicare 2016 Part C & D Display Measure Technical Notes, a document that describes the metric, data source and reporting time period for each Medicare Part C or Part D display measure for 2016. Erin Trompeter Managing Editor Atlantic Information Services, Inc. The Challenges of Pharmacy Star Ratings • 3 The Regence Approach to Stars Improvement The Regence plans owned by Cambia cover four states: Oregon, Washington, Utah and Idaho. Regence’s commercial enrollment exceeds 2.4 million members, mostly located in Washington and Oregon. Its Medicare Advantage segment as of January 2016 had more than 100,000 members across all its states, but Oregon and Washington again had the largest enrollment. Cambia offers both PPO and HMO MA products, with the majority of its enrollment in PPO plans. Regence achieved 4.5 stars for its MA plans in the Oregon and Utah markets, and 4 stars in its Washington and Idaho plans. These ratings represent an improvement from 3.5 stars across all its markets for its PPO plans, since its HMO contracts were too new to be rated in 2015. Part D Improvements Regence’s prescription drug event (PDE) rates across its markets averaged a full star improvement from 2015 to 2016, moving from an average rate of 3.4 in 2015 to 4.4 in 2016. Two Part D measures, both triple weighted, achieved five stars in every plan for medication adherence for both hypertension and cholesterol. Regence also had three PDE measures that improved in every state, and four PDE measures that improved at least one full star in each market. Medication adherence for hypertension and high-risk medication are both triple-weighted measures, and Regence plans improved an average of two stars across its markets. Medication adherence for cholesterol improved at least one star in every market. Regence’s Part D improvement measure, which carries a weight of five, improved at least one star in every market, as well. Regence’s medication adherence for cholesterol and hypertension both experienced an average rate improvement of 3%. The insurer’s high-risk medication prescribing rates dropped 7%. How It Happened To drive its improvement from an insurer with its share of 3.5-star plans, Regence conducted some “soul searching.” That effort started with the understanding that its members deserved better and that, to be a high-performing plan, it needed to transform the way it managed its measures and interventions. Regence attributes its 2016 star ratings achievement to tactical and intentional shifts in how it approaches the star ratings program as a whole. The shift in its program began when it evolved its strategy to align with CMS’s National Quality Strategy. 4 • AIS’s Management Insight Series The six priorities that are built into the CMS’s National Quality Strategy are: (1) Making care safer by reducing harm caused by the delivery of care; (2) Ensuring that each person and family are engaged as partners in their care; (3) Promoting effective communication and coordination of care; (4) Promoting the most effective prevention and treatment practices for the leading causes of mortality; (5) Working with communities to promote wide use of best practices to enable healthy living; and (6) Making quality care more affordable for individuals, families, employers, and governments by developing and spreading new health care delivery models These priorities are ingrained into the star ratings program through the measures, the metrics and the methodologies by which CMS administers it. Regence built its stars strategy to focus on improving not only the quality of care members receive, but the way in which they experience that care. Regence set out to create a cross-functional, member-centric stars program in partnership with its PBM, OmedaRx, another Cambia subsidiary. Historically, Regence had an internal clinical team that focused primarily on Healthcare Effectiveness Data and Information Set (HEDIS) stars measures and relied on a pharmacy team to focus on the PDE measures. Each team would conduct its own interventions, including member and provider outreach activities. This method was not only fragmented and somewhat uncoordinated, but it caused dissatisfaction among Regence’s providers and members, who received multiple contacts and interventions. Regence’s first step was to move away from the silos and to partner with its pharmacy, clinical, quality and member services teams to build a cross-functional core dedicated to improving star measures. The company’s second step was to capitalize on this multidisciplinary expertise, which added a layer of sophistication and integration to how it tapped into the following key levers and capabilities to drive star improvement: u Disease management, u A clinical data strategy, u Provider engagement, and u Member experience. When Regence built its core team and started to integrate both Part C and Part D metrics, the capabilities that supported stars, into the program, it recognized improvements across an array of measures almost immediately. The company also witnessed a culture shift within the organization around stars, as its other departments and employees started looking at star ratings and realized that every individual within the organization can have an impact on its star ratings. The Challenges of Pharmacy Star Ratings • 5 Improving Medication Adherence A 2009 American Heart Association study showed that, out of all the prescriptions for chronic medications written, 12% were never filled, 12% never started and another 29% were not taken as directed. As a result, adherence to chronic medication is reported to be as low as 50%. Considering only 50% of patients are taking their medication as intended, stakeholders are not able to prevent morbidity and mortality as effectively as desired. Poor medication adherence leads to poor health outcomes, which increases service utilization and increases total cost of care. It is estimated that avoidable medical expenditures associated with poor medication adherence can be as high as up to $290 billion annually, with an even greater impact on productivity loss on a societal level. In the star ratings program, medication adherence metrics are triple-weighted measures on cholesterol, hypertension and diabetes medication. These medications have been proven to produce long-term health benefits, so adherence to these therapies is crucial. From 2015 to 2016, Regence took steps to ensure strategic and organizational alignment in order to achieve stars improvement on all three adherence measures . The member-focused strategy took into consideration the entire prescription life cycle. When members are new to a targeted treatment, Regence sends educational mailings to help solidify their understanding of their disease state, the long-term benefit of the medication and the importance of adherence. On an ongoing basis, Regence delivers refill reminders that are automated based on claims monitoring, as well as low adherence interactive voice responses that help identify barriers and also offer tips to overcome those barriers for members. In addition, members have the opportunity to hold a telephonic consultation with a pharmacist. During the fourth quarter of each year, Regence deploys a live outreach call to members who are at risk of ending the year with less than 80% adherence with the goal of assisting members with refills and promoting 90-day fills. In 2015, Regence partnered with a company that specializes in health care advanced analytics and is able to incorporate data into predictive models. Specifically, this model identifies members who are currently non-adherent to their medication and are also at high risk for developing poor outcomes. Using this model, Regence could pilot an adherence intervention targeting these high-risk members. In 2016, Regence is accruing data on the pilot in order to evaluate the effectiveness of this approach. The Provider’s Role in Adherence In formulating its provider engagement strategies on medication adherence, Regence’s guiding principle is to make it easy for providers to do the right thing. In 2016, Regence is launching provider notifications based on the presence of primary medication nonadherence (PMN) and low adherence indicators. 6 • AIS’s Management Insight Series PMN occurs when a member is prescribed a new medication but never picks it up. Regence understands that it won’t be able to capture all cases of PMN, but if the prescription is sent to the pharmacy, processed and later reversed, the insurer can use it as a proxy for PMN. With the increased adoption of e-prescribing, Regence is hopeful it will be able to capture more cases of PMN. Kelly: A Member Case Study In applying Regence’s adherence strategy to individual members, the insurer using the example of Kelly, who is newly diagnosed with diabetes and needs to start on a prescription of Metformin. Kelly’s doctor sends the prescription to the pharmacy, she picks it up, and a few days later, she receives educational materials about Metformin in the mail. If Kelly hadn’t picked up her prescription, her doctor would have received a fax from the plan, and the office’s nurse would call Kelly to urge her to the pharmacy the next day to pick up her prescription. Kelly’s pharmacy offers steep discounts on generic medications, and she has in the past paid cash for some pharmaceuticals. When the pharmacy runs her Metformin as a cash transaction, it receives an alert that says, “Kelly is in the Regence plan that offers competitive generic cost share.” The pharmacy runs the prescription through Regence and finds the cost share to be affordable. From this point, Kelly receives automated messages to remind her about refills. If her adherence drops below 80%, she also receives an automated message to help her understand what is stopping her from refilling and what she can do to get back on track. In fact, during a phone call to discuss her low adherence, she describes side effects to a pharmacist, who tells her about a long-acting Metformin tablet that can help reduce these issues. She is able to have her prescription changed. In the fourth quarter, Regence calls Kelly because she needs one more refill to keep her adherence above 80% for that year. The pharmacy technician contacts Kelly after seeing that she’s been filling a 30-day supply of the Metformin at a time and informs her about Regence’s 90-day retail benefit. With this benefit, she pays two copays for a three-month supply. Kelly’s example incorporates a number of health plan considerations: the 90-day retail benefit, cash claim conversions, tactical strategies, and provider and member strategies that come together to help the patient identify opportunities to be more adherent to medications. Integration with care management is also key. Case management nurses can talk to a member about issues driving non-adherence such as post-discharge medication reconciliation. This collaboration enables health plans to leverage different touch points with members. The Challenges of Pharmacy Star Ratings • 7 High-Risk Medication Measure The high-risk medication (HRM) measure is another important metric in driving stars performance. HRMs are those identified on the Beers Criteria list, later adopted by CMS. These drugs have been linked to falls, fractures and other adverse events. It is recognized, though, that there are unique clinical situations where an HRM is the only reasonable option for an elderly member. The goal of this metric is not to drive utilization to zero, but to minimize use and mitigate associated risks. The way the metric is calculated, the first fill of a high-risk medication in a calendar year doesn’t count negatively to a plan. But once the patient obtains a second fill, then the member becomes part of this numerator. In this case, a lower rate means better performance. From 2015 to 2016, Regence plans’ performance on the HRM measure improved by one to two stars. Much like with the adherence measures, the improvement was the result of a combination of strategies. Unlike the approaches that Regence took for improving adherence, which more heavily emphasize motivating the member, the HRM strategies depend more upon provider engagement. Regence uses medication therapy management programs not only for retrospective drug utilization reviews, but also in its quality incentive program for providers. The incentive program is supported by a provider pool that allows the sharing of utilization and progress reports. At regularly scheduled provider group meetings, pharmacies participate in sharing performance measures and best practices. Regence also works with pharmacy departments that are embedded in provider groups and health systems, sharing best practices and providing them with information that can help them identify which Regence members to prioritize in addressing a trend. Since these pharmacy departments have a much closer working relationship with their providers, they often carry out the actual interventions. In 2015, Regence tried a couple of new approaches, such as adding live outreach to providers when the first fill of a targeted HRM is detected. The calls’ intent is to bring attention to the HRM used and suggest a safer alternative. Another strategy Regence tried in 2015 was to be more strategic in targeting certain members and providers in letters and campaigns. For example, for estrogen, Regence identified patients who were on both anticoagulants and estrogen and sent letters to the prescribers of the anticoagulants to ask them to consider making therapy changes. When it comes to estrogen use in general, it is often the member who drives that decision, so Regence decided to identify which members might be the most receptive to the idea of discontinuing estrogen. It targeted members over age 75 or those who have multiple concurrent medications, with the message to discuss with their providers whether there is still a clinical need to continue estrogen therapy. One of the most effective strategies to manage HRMs Regence initiated is the addition of prior authorization, which began in 2014. With prior authorization, the pre- 8 • AIS’s Management Insight Series scriptions of the targeted HRMs must be reviewed and approved before coverage is determined. In addition, HRMs are moved to non-preferred formulary tiers, aligning member incentives to favor the safer alternative. When leveraging analytics, Regence also learned that most of the first and second fills of HRMs occur in the first quarter of the year, so if the insurer wants to start a new HRM program, it should have it up and running before the start of the year. The decision to use an HRM is ultimately between the prescriber and the patient, but through the programs, Regence can strive to support that shared decision-making by introducing the right information, tools and incentives. Regence has paid attention to targeting the right members to intervene, and has been leveraging the integration of pharmacy and medical data. Looking ahead into the new quality measures for star ratings, Regence will continue to improve its data integration. For 2018, CMS has identified potentially adding post-discharge medication reconciliation and statin use in diabetics as star measures. The asthma management measure is also under consideration as a star measure in 2018. The Challenges of Pharmacy Star Ratings • 9 Interventions and Programs Regence has several interventions and programs in place that traditionally were focused on Part C measures, but actually have a greater influence on Part D when managed holistically with OmedaRx. These programs include: Disease management. The goal of Regence’s disease management program is to optimize clinical outcomes for participating members through education and care coordination. The company targets candidates for disease management programs based on condition, looking for members with diabetes, congestive heart failure, hypertension and chronic obstructive pulmonary disease (COPD). The insurer has an internal team of disease management nurses who engage members to influence behavior modification through education and counseling, in addition to coordination with the member’s primary care provider (PCP). The nurses integrate pharmacy education into these discussions specifically around medication management and reconciliation, medication adherence, and the potential side effects of high-risk medications. Regence’s nurses want to ensure members understand how to take their prescriptions, and that they are actually being taken as prescribed. By incorporating Part D measures into its disease management scripts and protocols, Regence has discovered it can better identify factors or barriers that could have a negative impact on a member’s ability to successfully manage his or her condition. One example might be a member who lacks transportation to see his PCP, or another who can’t afford her prescription medication. When Regence can identify those factors and barriers and remedy them, it can see an increase in Part C and Part D gap closure, which drives improved outcomes. Another intervention Regence conducted is its readmission prevention program. The company has a clinical team dedicated to conducting proactive member outreach post-discharge. This outreach is performed telephonically to help manage care transitions. The team makes sure the member or his or her caregiver fully understands the discharge plan. The team also aims to perform the following tasks: u Ensure the member has access to prescriptions, and that those prescriptions are being taken as prescribed. u Make sure any ancillary services or supplies that were ordered have been delivered, or at least scheduled, including any durable medical equipment, home health or physical therapy services. u Perform a medication reconciliation review and direct coordination with the PCP to schedule post-discharge follow-up. Regence has also instituted a fall prevention program, which can include a medication review and reconciliation as a key part of reducing the risk of falling. 10 • AIS’s Management Insight Series In addition, Regence has a comprehensive in-home assessment program for chronically complex members. In this program, the insurer sends providers into the home to conduct full health risk assessments, including looking at medications, reconciling medication lists and conducting a full risk assessment. The provider sends the assessment and any alerts to the health care plan and to the member’s PCP for ongoing care coordination. Provider Engagement Engagement with the PCP is a critical piece of most of Regence’s disease management programs and is a key lever for its stars improvement strategy as well. It’s been estimated that up to 60% to 70% of a plan’s ratings are driven by the provider. It is imperative for plans to align their incentives and manage the quality of care delivered within the network. Myriad incentives exist in the market, and they come in all different shapes and sizes. Regence has had numerous programs in place, as well, and they are designed to drive improved quality outcomes and value. Regence has learned in its provider engagement strategy that it needs to partner with its providers to make sure its incentive programs align with clinical objectives. The company also wants to make sure its incentive programs are actually rewarding quality performance and not just giving providers a list of boxes to check or hoops to jump through. Regence also finds it’s important to streamline its metrics to limit the number of items it asks providers to track. Providers are asked to perform many tasks for all their payers, so plans’ best chance to engage their providers fully comes in making sure requested tasks can actually be integrated into their day-to-day workflow. Regence has pay-for-performance incentive programs in partnership with its providers that are based on a number of quality indicators, including coding and documentation accuracy, quality scores for HEDIS and PDE measures, and financial performance indicators. Education and ongoing feedback are also key components to Regence’s provider engagement strategy. The insurer deploys a cross-functional team, called a SWAT team, that is armed with performance data and meets with providers to offer education and invite collaboration and dialogue to work together to improve member health. The SWAT team includes members from Regence’s pharmacy team, as well as revenue management, provider network, stars, quality and sales. The team aims to meet monthly with providers to identify opportunities for performance improvement. Because conversations in these meetings are performance-driven, Regence is dependent on complete, accurate and timely data. The insurer emphasizes data exchange to help both the PCPs and the health plan get a comprehensive, holistic picture of patients’ health status and their performance. The Challenges of Pharmacy Star Ratings • 11 The Clinical Data Strategy Regence recognizes that there are multiple sources of data it can leverage for a variety of activities that can drive quality outcomes and improve the member experience, including claims data, risk scores, consumer data, lab values, biostatistics, health risk assessments and pharmacy claims. All these elements help paint a picture of members and their health status, as well as their preferences and potential risk factors. Having visibility into these types of data is critical and fundamental to Regence’s population health models and initiatives. Once Regence has the data, it uses clinical data to identify members with gaps in care, and then stratifies these members based on risk factors. The insurer develops interventions designed to close those gaps and uses the data to develop predictive models to help project member behavior. Predictive analytic capabilities have evolved. Historically, the 80/20 rule meant that 20% of the members drive 80% of the claims. Trying to get a better picture of who those 20% are and trying to predict what they’re going to do next, Regence looks at diagnoses and utilization. The insurer now has many more factors that it can use to predict behavior. Among the sources data scientists review is prescription drug history, including the number of prescriptions and adherence levels. The insurer also can look at the number of different provider visits and ER utilization. Social factors also come into play, like whether members have a transportation issue, are homebound, have a social support system or are living in rural vs. urban situations. There are many factors that Regence can pull into its data analytics to enhance its models. Regence uses predictive analytics to identify who is most likely to adhere to a medication, who might be more likely to be readmitted to a facility, who might take a fall and whether that fall could result in an injury, and who is more likely to be compliant with preventive screenings. The insurer can also use consumer preferences to tailor interventions and outreach to better meet the needs of its members. Electronic medical record (EMR) alerts and admissions information help the company identify opportunities to impact some event-driven measures, including readmissions, osteoporosis if a female presents with a fracture, or COPD or asthma episodes. Regence incorporates a large amount of data into its analytics — but the quality of outputs is dependent on how complete the inputs are. Providers use risk stratification models in their practices, but they often tell Regence that they don’t have complete data on their end, either. As a health plan, Regence has disability data that the providers lack in the form of claims data or pharmacy claims data, for example. Providers in turn have data that insurers don’t — information found in the EMR, lab values and blood pressure readings, for example. 12 • AIS’s Management Insight Series Ideally, plans and providers can have an efficient flow of data exchange. Insurers’ clinical decision-making is based on complete and accurate data at the time of the patient encounter with the provider, so the insurer has patient/provider data in the same place at the same time. Through its clinical data strategy, Regence aims to capture a comprehensive view of the health status of all members so it can engage them in a holistic manner based on their preferences. This approach invites members to actively participate in their health care in a way that works for them. Optimizing Member Experience Regence’s fourth — and perhaps most important — lever is optimizing the member experience as members navigate through their health journey. Coordinated outreach optimizes member contacts: Regence wants members to have a positive experience when they interact with the insurer as a health plan, and wants members to be actively engaged in their health. To achieve that, Regence must be coordinated internally to be thoughtful and intentional in the way it engages members. Regence strives to provide concierge-level service for members where its outreach is coordinated and delivered in a manner that addresses the member holistically. The insurer also aligns its member and provider messaging so initiatives are focused and targeted. Having complete Medicare Part C and Part D gap information available at the time of the outreach enables the insurer’s team to have a single conversation with the member. That single conversation can address anything applicable to that member around gaps in care, PCP selection, chronic disease management, medication adherence or safer alternatives to high-risk medication — and all issues can be addressed in one phone call. Regence also utilizes personas to make sure its outreach is member-centric and relevant. The personas have become part of its program development, because when it develops outreach initiatives, it helps to personalize the impact of those campaigns. Meet Polly One of Regence’s personas is Polly, age 70, who is hypertensive, but is on top of her health and current with her preventive screenings. She has a good rapport with her PCP and checks in at least once a year for her annual wellness visit. She’s flying under the radar until she has an event — it’s icy outside, and she slips, has a fracture and is admitted to a facility. Before she goes home, she is given a high-risk medication. Since she has been admitted, she has been unable to go to the pharmacy to pick up her hypertension medication. Suddenly, Polly is triggering all sorts of stars alerts on Regence’s end. She has the osteoporosis management flag and is now at risk of a readmission. She has alerts for her high-risk medication and her medication adherence. The Challenges of Pharmacy Star Ratings • 13 When Regence looks at its personas, it thinks of how to communicate with Polly to make sure that she gets the care she needs in the manner that she prefers. She is a great example of a member who has had a handful of care needs that traditionally may have been managed by multiple people within Regence and across the continuum of care. But with its member experience strategy and stars improvement strategy, utilizing its four levers — disease management, clinical data, provider engagement and the member experience strategy — Regence can be intentional and mindful in how it coordinates Polly’s transitional care management, through the combined expertise of its pharmacy team, disease management nurses and care managers and her PCP. While the insurer may have a lot of activity and personnel working on her transition plan, from her vantage point, ideally, her road to recovery is seamless and fluid. The Challenges of Pharmacy Star Ratings • 15 Q&A Session Question: What is the makeup of your concierge team? Kristine Walhof: We have 80 member service specialists in total that are fully trained on how to engage with our Medicare Advantage members. We work very closely with them. We also work with our care management team to make sure there is some alignment between what outreach is happening, and when and who is being outreached. So, there is a coordinated effort to ensure our members aren’t being over touched. A little Regence goes a long way. So if we can only reach out once and be very purposeful and intentional with that, we get better results. Question: How do you coordinate that one-time call with the needs of member availability? Walhof: That is one of our challenges. We are still evolving our processes. But it really is dependent on getting all of our data pulled together and having a good understanding of what that member needs, and utilizing our internal systems to make sure that we do have the gap data and member preference data front and center before those calls are made. Question: Vivien touched on some of the upcoming changes to the star ratings program. Statins used in diabetics was one of those identified for 2018. How is that different from the retired diabetes treatment measure, and what challenge does that present? Vivien Chan: For the statin measure, we are at a point where we have a better coordinated approach, whether it’s a quality incentive program or engagement with providers. I think that would be good leverage for us now compared to previous years where we didn’t have as strong of a structure in that area. What we’re trying to look at now is partnering with our care management team, who is already doing some of the care gap closure work, and trying to understand where the members overlap between their identification and targeting process and our targeting process, and where are the different touch points possible where we can insert this measure. We’re having those conversations now pretty early on, and mapping out organizationally where the touch points are and how we can approach people with these touch points without duplicating effort to drive this gap closure. Question: On the 2018 display measures, how much of opioid overutilization are you incorporating into this integrated approach? Chan: That’s a very good question. Opioids are a hot topic these days, whether it’s in the Medicare population or just in the general commercial world. What we have in place currently is a lot of the things that other plans have been doing: OMS [the Medicare Part D Overutilization Monitoring System] and the overutilization monitoring that CMS has mandated. We are looking into how to better target these people. We have some internal pilots that were started in the commercial population. We’re trying to learn from that, because anytime when we are looking at opioid overutilization, it really isn’t just about the Part D measures; it really is taking into consideration the care 16 • AIS’s Management Insight Series of a member. Are they using physical medicine? Are they using other types of pain control? Are there fraud, waste and abuse issues that are coming up? So, I think it’s really looking at the holistic approach of how can we get more in tune with the right measure. Right at this moment we are mainly doing the Medicare mandated programs. But through the pilots that we have in place for commercial, which is more of a cross-functional, multidisciplinary review of the top category of patients that are identified, we are hoping to learn something from that pilot and see if there are any ways to apply some of the learnings in our Part D world. Question: We’ve got one more question for you about these measures for the statin use and diabetes measure. What are your thoughts around overcoming the October 2015 FDA warning that statin use causes type 2 diabetes? Chan: Isn’t that a perfect storm? I think we are going to have to dive into that clinical evidence a little bit more to understand that, and to see if there are any subgroups of patients that we need to be paying attention to in terms of applying our strategies differently. We’re pretty early on in looking at that data. We’ll have more coming up. Question: This is a question for Kristine. Going back to the concierge team, do you use clinicians, non-clinicians? What is the makeup of that team? Walhof: We are still evolving how that team is going to function. It’s really more of a concept at this point. We want our member services team and our clinical team, when they are reaching out to the member, to be able to incorporate all of the relevant information for that member. We have clinicians reaching out. We have member services folks reaching out. We have non-clinician members of our clinical team reaching out as well. Our goal is to make sure that whenever anyone is picking up the phone, they have an understanding already of what other contacts were made so that they are better able to address the whole needs of that member. We’re still evolving with the concept, trying to make it a little bit more broad based. But the concept is one that has changed the way we interact with our members. Question: Could you clarify about the timing of the efforts you undertook to integrate all these different areas of expertise? Walhof: It started early 2014 throughout 2015, and they continue to evolve. I think that’s the beauty of the star ratings program. You can’t just sit still. You have to continue to look at your processes, who is part of your program, and who is helping to drive improvement so that we can continue to meet the needs of our members and continue to see improvement. Question: This question is for Vivien. Can you identify the predictive analytics firm that you partnered with? Chan: Not at this point. I think it’s a proprietary issue that we have in agreement with the firm. So, unfortunately not at this point. The Challenges of Pharmacy Star Ratings • 17 Question: How are the member-focused strategies and the provider-focused strategies that you mentioned triggered? Are there alerts sent from regions that these people need to be intervened with? Chan: A lot of these are triggered real time. There are ones that are triggered real time by our pharmacy claims monitoring. There are also certain ones with the targeted mailings that I talked about. Those are retrospectively pulled from the pharmacy claims data, and we apply analytics to it before we send out the targeted mailing. So, there are two types. One is real time, like the IVRs [interactive voice response]. The primary medication nonadherence pieces and provider faxes are real time. We have certain logics that are built in to make sure that we’re not over-communicating within a certain timeframe. Let’s say if a member has multiple adherence issues, we’re not going to keep bombarding them with that. There are certain parameters we have built around that to cap the number of communications that will be going out. Then the retrospectively reviewed data that are pertaining to some of the targeted mailings, those are done after the claims have accumulated over a certain time period. Question: I have another clarification for Kristine. You mentioned up to 60% to 70% of plans ratings are driven by a provider. Is that Part D plans in general? Walhof: This would be your overall star ratings, so looking at your HEDIS metrics and looking at Part D. It’s combined. And it’s an estimate. You look at who is really the largest influencer of getting gaps closed. But a lot of that is being driven by our providers. That’s one of our critical strategies to make sure the providers in our networks are really working to be five star level providers. The Challenges of Pharmacy Star Ratings • 19 Appendix A: AIS Coverage of Medicare Advantage Star Ratings Results for 2016 Plans More MA Plans Get Four-Star Ratings for ’16, but Scores of Humana, WellCare Dip There were lots of winners and some losers in the 2016 star quality ratings unveiled by CMS Oct. 8, 2015, and several in both categories were affected substantially, industry consultants tell AIS, by this being the first year the agency didn’t have fixed fourstar thresholds that plans knew before getting rated. Overall, there were considerably more Medicare Advantage prescription drug plans — 49% versus 40% of contracts in 2015 — and more big plans getting four stars and above, and the percentage of MA-PD enrollees in those plans leaped to 71% from 60%. But there also was what one observer called abundant “churn” in the top performers, and differing directions in performance of MA-PDs versus stand-alone Prescription Drug Plans (PDPs). The winners in the new rankings included some of the “usual suspects” — five Kaiser Permanente plans and other vertically integrated insurers, especially in New England and the upper Midwest — that got the top five-star rating. However, seven of the 17 plans (including MA-only and PDPs) receiving the top rating are new to the list, including several for-profit plans. There were 12 MA-PD five-star plans, up from 11 in 2015. Most publicly held insurers made gains in their enrollment-weighted star ratings for 2016, according to securities analyst Thomas Carroll of Stifel. Aetna Inc. had the highest average rating among them at 4.3, up from 3.9 in 2015, and industry giant UnitedHealth Group, while trailing many others with a 3.9 average rating, was up from 3.5 in 2015. The losers in the new ratings, Carroll’s analysis shows, include Humana Inc., which had strong gains in 2015 but had its enrollment-weighted average drop to 4.0 in 2016 from 4.2 and lost its only five-star rating, for a CarePlus Health Plans, Inc. unit in Florida. Also in the minus column was WellCare Health Plans, Inc., which dipped to a 3.1 average from 3.2 and has two of the six plans that will get CMS’s Low Performer Icon (LPI) for 2016 because of consistently low star scores. In addition, there are some losers by plan type and location, and analysts offered varying reasons and implications for those. PDPs, for instance, dropped from 2015 in the number of contracts earning 3.5 stars and above and rose in contracts with ratings of 2.5 and two stars. 20 • AIS’s Management Insight Series Perhaps an indirect loser in the new ratings is MA Special Needs Plans (SNPs) — but for an unusual reason. CMS included a column in the 2016 ratings that showed whether each of the 12 five-star MA-PD contracts had SNPs as part of their contracts, and seven of them did. Analysts queried by AIS concurred that this could hurt SNPs in their pleas for relief in the star ratings, which they contend unfairly penalize plans serving highly disadvantaged populations. However, Stephen Wood, a principal in Clear View Solutions, LLC, says that the SNP-related data in CMS’s table can be misleading. He cites as an example a millionmember Kaiser five-star contract that has only 60,000 SNP members, all of them Medicare-Medicaid dual eligibles. “So the SNP didn’t really contribute to the five stars” and thus should not enable CMS to argue that SNPs don’t need stars help, Wood tells AIS. The MA-PD contracts getting the top rating are Cigna HealthCare of Arizona, Inc., Essence Healthcare, Inc. in Illinois and Missouri; Group Health Plan, Inc. in Minnesota and Wisconsin; Gundersen Health Plan in Iowa and Wisconsin; Kaiser units in California, Colorado, Hawaii, the Mid-Atlantic states and the Northwest; Martin’s Point Generations, LLC in Maine and New Hampshire; Sierra Health and Life Insurance Co., a United unit, in nine states (representing Erickson Advantage plans for Erickson retirement communities); and Tufts Associated Health Maintenance Organization in Massachusetts. MA-only contracts that got five stars are Medical Associates Health Plan, Inc. in Iowa and Illinois; Medical Associates Clinic Health Plan in Wisconsin; and Dean Health Plan, Inc. in Wisconsin. The six contracts getting the LPI for 2016 are Cuatro LLC — known as Access Medicare — in New York; GHS Managed Health Care Plans, Inc. (owned by Health Care Service Corp.) in Oklahoma; Touchstone Health HMO, Inc. in New York; United’s Sierra unit in Utah; WellCare of Louisiana, Inc.; and WellCare’s Windsor Health Plan, 2013 to 2016 Overall Star-Rating Distribution for MA-PD Contracts 2013 2014 Overall Rating Number of Contracts Weighted Number by of % Enrollment Contracts 5 stars 11 2.46 9.42 4.5 stars 54 12.08 4 stars % 3.25 10.23 19.59 65 17.62 25.02 86 21.77 30.32 102 27.64 35.71 30.49 136 34.43 26.78 112 30.35 19.55 16.63 73 18.48 10.98 66 17.89 8.60 2.37 12 3.25 0.90 0.08 0 0.00 0.00 9.56 15.81 64 14.85 62 13.87 12.56 3.5 stars 131 29.31 3 stars 2 stars Total Contracts Average Star Rating* Number of Contracts 2016 Weighted by Enrollment 2.55 2.5 stars 9.88 20.55 61 15.44 87 20.19 21.68 36.48 143 33.18 127 28.41 20.25 109 25.29 60 13.42 5.28 16 3.71 1.09 26 6.58 0.21 1 0.23 0.01 2 0.51 0.45 447 431 3.71 11 Weighted Number by of % Enrollment Contracts 2.78 2 11 2015 Weighted by % Enrollment 395 3.86 * The average star rating is weighted by enrollment. SOURCE: CMS Fact Sheet on 2016 star ratings released Oct. 8, 2015. 12 369 3.92 4.03 The Challenges of Pharmacy Star Ratings • 21 Inc. in Arkansas, Mississippi, South Carolina and Tennessee. CMS added a footnote indicating that the Cuatro, Sierra and Windsor contracts are “eligible for termination at the end of 2016.” Looking at the overall implications of the new stars data, John Gorman, executive chairman of Gorman Health Group, LLC, asserts that not-for-profit plans continue to “stomp for-profits” in achieving high scores. About 70% of the not-for-profit contracts got four stars or above, he tells AIS, while just 39% of the for-profits did. SNP contracts improved at about the same rate as MA HMOs, according to Gorman, achieving an average star rating of 3.61, up from 3.47 in 2015. HMOs and PPOs gained to 3.87 from 3.79, he says, adding that the SNP gains are “a very big deal” and may “debunk” in CMS’s mind the arguments for special treatment. Gorman cautions that the overall star-ratings climate will change significantly for 2017, noting that only 369 MA-PDs got rated for 2016 but 188 others will be in 2016 because of several factors, including a CMS change in the minimum size qualifying for star ratings from 1,000 members to 500. Looking at the MA results as a whole, stars specialist Jane Scott, who recently became senior vice president of professional services for Health Integrated, says that there was “not much improvement” from 2015 to 2016. “Measures moved; the industry didn’t,” she tells AIS, referring to such aspects as restored, new, dropped and modified star measures as well as the first year without four-star cut-points that let plans know in advance what they’ll have to achieve to get that rating. She points to such reductions as from 4.2 for 2015 to 3.2 for 2016 in the average score for colorectal cancer screening and from 4.2 to 3.3 in diabetes care – kidney disease monitoring as among indicators that the trend was mixed. There also was a drop from 3.7 to 3.1 on diabetes care – eye exam and from 4.6 to 3.3 in improving or maintaining physical health. Some of these reductions, Scott suggests, result from lack of sufficient member engagement since the services are available but need a “member response” to get done. On the positive side of the MA star-ratings ledger were gains in several measures that plans have scored poorly on in the past. Improving or maintaining mental health, for example, went from 2.5 to 3.3, monitoring physical activity from 2.2 to 2.9 and osteoporosis management from 2.1 to 2.5. Even diabetes care — blood sugar controlled went from 3.3 to 3.9. Some of the big changes for plans this time were on Part D measures, and there were different trends among MA-PD plans versus PDPs. “Medication adherence for diabetes medications,” for instance, improved from 3.5 to 3.9 for MA-PDs but fell from 3.0 to 2.7 for PDPs. Most of the moves, though, were in same direction, including appeals auto-forwarded (up from 3.6 to 4.5 for MA-PDs and from 2.5 to 4.1 for PDPs) and appeals upheld (down from 3.7 to 3.3 for MA-PDs and from 3.9 to 3.1 for PDPs). Scott points 22 • AIS’s Management Insight Series out that auto-forwarding is a process measure, while appeals upheld relates to the substance of the appeals. And the downward movement in the latter category, suggests Wood, “has everything to do with drug prices,” including substantial increases in copayments to help plans offset big price hikes and expensive new drugs. Several of the analysts saw some of the measure scores involved indicating that plans must push their pharmacy benefit managers (PBMs) harder to aid them. The average stars score for price accuracy on the CMS Medicare Plan Finder, for example, plummeted from 4.6 to 3.5 for the MA-PDs, and the MPF is a “PBM function,” Wood remarks. While PDPs did well on that measure, remaining at 4.7 stars for 2016, the decline in PDP ratings overall (to an average of 3.4 stars from 3.75 in 2015) was the biggest surprise in the new ratings, says Michael Lutz, a director in the health reform practice at consulting firm Avalere Health LLC. Part of the explanation for this, he tells AIS, lies in changes on how some of the measures are calculated from year to year, and another factor was a new measure, comprehensive medication review completion rates in medication therapy management programs, which got an average score of just 2.3 for both MA-PDs and PDPs. Along similar lines, he notes, CMS retired for 2016 a diabetes-related measure on which plans had done well, although it also restored breast-cancer screening and callcenter foreign-language availability measures — not used in 2015 for methodologychange reasons — on which plans did well. For the PDPs, a bigger factor than new or restored measures was complaints about the drug plan, for which the average score fell from 4.3 in 2015 to 3.5 for 2016. Lutz speculates that the causes of this could include PDPs instituting more copayment tiers to deal with rising drug prices. And the impact of any one sharp star-measure score decline is bigger for PDPs than for MA-PDs, he adds, since there are far fewer rating measures for the PDPs. Still that couldn’t explain the entire downward trend for PDPs, especially since average scores on all three medication-adherence measures were down from 2015 to 2016 for PDPs, while all three were ip for MA-PDs. The Challenges of Pharmacy Star Ratings • 23 Counter to Recent Stars Trend, Local PPOs Fared Better Than HMOs Local Medicare Advantage PPOs narrowly beat MA HMOs in terms of average MA Star Ratings Quality Performance By Plan Type, 2012-2016 CMS star quality ratings for 2016, acEnrollment-weighted average star rating cording to an analysis of the new ratings 4.5 prepared by consulting firm McKinsey & Co. in October 2015. The PPOs, McKinsey 4.0 said, have an enrollment-weighted star rating of 4.16 for 2016, the first time in the 3.5 four years tracked that PPOs have scored HMO/HMO-POS better on this basis than have HMOs. MA 3.0 Local PPO private-fee-for-service plans and regional PFFS Regional PPO PPOs continued to trail the two leaders. 2.5 2013 2014 2016 2012 2015 Other findings from the analysis include Note: PFFS=private fee for service; POS=point of service. that the change in CMS’s “cutpoint” SOURCE: McKinsey & Co. analysis of CMS’s 2016 Part C methodology starting in 2016 to not have & D Medicare star rating data; October 2015. View the full report at http://healthcare.mckinsey.com/assessingpredetermined thresholds for earning 2016-medicare-advantage-star-ratings. a four-star rating on individual scoring measures had only a negligible net impact on differences in MA plans’ overall stars performance between 2015 and 2016. 24 • AIS’s Management Insight Series Plans With 4.5 Stars or More Contract Number Organization Type Parent Organization Contract Name 2016 Overall Rating H0354 Local CCP CIGNA CIGNA HEALTHCARE OF ARIZONA, INC. 5 H0524 H0630 H1230 H2150 Local CCP Kaiser Foundation Health Plan, Inc. KAISER FOUNDATION HP, INC. 5 Local CCP Kaiser Foundation Health Plan, Inc. KAISER FOUNDATION HP OF CO 5 Local CCP Kaiser Foundation Health Plan, Inc. KAISER FOUNDATION HP, INC. 5 1876 Cost Kaiser Foundation Health Plan, Inc. KAISER FNDN HP OF THE MID-ATLANTIC STS 5 5 H2256 Local CCP Tufts Associated HMO, Inc. TUFTS ASSOCIATED HEALTH MAINTENANCE ORGANIZATION H2462 H2610 H5262 H5591 1876 Cost HealthPartners, Inc. GROUP HEALTH PLAN, INC. (MN) 5 Local CCP Essence Group Holdings Corporation ESSENCE HEALTHCARE, INC. 5 Local CCP Gundersen Lutheran Health System Inc. GUNDERSEN HEALTH PLAN 5 Local CCP Martin's Point Health Care, Inc. MARTIN'S POINT GENERATIONS, LLC 5 5 5 H5652 Local CCP UnitedHealth Group, Inc. SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. H9003 H0332 H0609 H0710 H1019 H1170 H1365 H1463 H1468 Local CCP Kaiser Foundation Health Plan, Inc. KAISER FOUNDATION HP OF THE N W Local CCP Kelsey-Seybold Medical Group, PLLC KS PLAN ADMINISTRATORS, LLC 4.5 Local CCP UnitedHealth Group, Inc. PACIFICARE OF COLORADO, INC 4.5 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE INSURANCE COMPANY 4.5 Local CCP Humana Inc. CAREPLUS HEALTH PLANS, INC. 4.5 Local CCP Kaiser Foundation Health Plan, Inc. KAISER FOUNDATION HP OF GA, INC. 4.5 Local CCP Martin's Point Health Care, Inc. MARTIN'S POINT GENERATIONS, LLC 4.5 Local CCP The Carle Foundation HEALTH ALLIANCE CONNECT, INC. 4.5 Local CCP Humana Inc. HUMANA BENEFIT PLAN OF ILLINOIS, INC. 4.5 4.5 H1537 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE INSURANCE COMPANY OF NEW YORK H1608 Local CCP Aetna Inc. COVENTRY HEALTH AND LIFE INSURANCE COMPANY 4.5 H1609 Local CCP Aetna Inc. AETNA HEALTH, INC. 4.5 4.5 H1951 Local CCP Humana Inc. HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. H2001 Local CCP UnitedHealth Group, Inc. SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. 4.5 H2226 H2228 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE INSURANCE COMPANY 4.5 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE INSURANCE COMPANY 4.5 H2230 Local CCP Blue Cross and Blue Shield of Massachusetts, Inc. BCBS OF MASSACHUSETTS HMO BLUE, INC. 4.5 H2261 Local CCP Blue Cross and Blue Shield of Massachusetts, Inc. BCBS OF MASSACHUSETTS HMO BLUE, INC. 4.5 H2320 H2354 H2419 H2422 H2459 Local CCP Spectrum Health System PRIORITY HEALTH 4.5 Local CCP HealthPlus of Michigan HEALTHPLUS OF MICHIGAN 4.5 Local CCP South Country Health Alliance SOUTH COUNTRY HEALTH ALLIANCE 4.5 Local CCP HealthPartners, Inc. HEALTHPARTNERS, INC. 4.5 Local CCP UCare Minnesota UCARE MINNESOTA 4.5 H2802 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE OF THE MIDLANDS, INC. 4.5 H3132 H3204 H3305 Local CCP AIDS Healthcare Foundation AHF MCO OF FLORIDA, INC. 4.5 Local CCP Presbyterian Healthcare Services PRESBYTERIAN HEALTH PLAN 4.5 Local CCP MVP Health Care, Inc. MVP HEALTH PLAN, INC. 4.5 H3344 Local CCP Independent Health Association, Inc. INDEPENDENT HEALTH BENEFITS CORPORATION 4.5 The Challenges of Pharmacy Star Ratings • 25 Plans With 4.5 Stars or More (continued) Contract Number Organization Type Parent Organization Contract Name 2016 Overall Rating H3362 Local CCP Independent Health Association, Inc. INDEPENDENT HEALTH ASSOCIATION, INC. 4.5 H3388 Local CCP Capital District Physicians' Health Plan, Inc. CAPITAL DISTRICT PHYSICIANS' HEALTH PLAN, INC. 4.5 H3447 Local CCP Anthem Inc. HEALTHKEEPERS, INC. 4.5 4.5 H3471 Local CCP The Carle Foundation HEALTH ALLIANCE NORTHWEST HEALTH PLAN, INC. H3597 H3664 H3668 Local CCP Aetna Inc. AETNA HEALTH, INC. (ME) 4.5 Local CCP Aultman Health Foundation AULTCARE HEALTH INSURING CORPORATION 4.5 Local CCP Trinity Health MOUNT CARMEL HEALTH PLAN, INC. 4.5 4.5 H3817 Local CCP Cambia Health Solutions, Inc. REGENCE BLUECROSS BLUESHIELD OF OREGON H3907 H3916 H3952 H3954 H3957 H4454 H4461 H4604 H4605 H5010 Local CCP UPMC Health System UPMC HEALTH PLAN, INC. 4.5 Local CCP Highmark Health HIGHMARK SENIOR HEALTH COMPANY 4.5 Local CCP Independence Health Group, Inc. KEYSTONE HEALTH PLAN EAST, INC. 4.5 Local CCP Geisinger Health System GEISINGER HEALTH PLAN 4.5 Local CCP Highmark Health KEYSTONE HEALTH PLAN WEST, INC. 4.5 Local CCP CIGNA HEALTHSPRING OF TENNESSEE, INC. 4.5 Local CCP Humana Inc. CARITEN HEALTH PLAN INC. 4.5 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE OF UTAH, INC. 4.5 Local CCP Cambia Health Solutions, Inc. REGENCE BLUECROSS BLUESHIELD OF UTAH 4.5 Local CCP Cambia Health Solutions, Inc. ASURIS NORTHWEST HEALTH 4.5 H5042 Local CCP Capital District Physicians' Health Plan, Inc. CDPHP UNIVERSAL BENEFITS, INC. 4.5 H5050 Local CCP Group Health Cooperative GROUP HEALTH COOPERATIVE 4.5 4.5 4.5 H5211 Local CCP Marshfield Clinic Health System, Inc. SECURITY HEALTH PLAN OF WISCONSIN, INC. H5215 Local CCP Ministry Health Care, Inc. NETWORK HEALTH INSURANCE CORPORATION H5216 H5253 H5410 H5425 H5431 H5521 H5526 H5883 H5938 Local CCP Humana Inc. HUMANA INSURANCE COMPANY 4.5 Local CCP UnitedHealth Group, Inc. UNITEDHEALTHCARE OF WISCONSIN, INC. 4.5 Local CCP CIGNA HEALTHSPRING OF FLORIDA 4.5 Local CCP SCAN Health Plan SCAN HEALTH PLAN 4.5 Local CCP HealthSun Health Plans, Inc HEALTHSUN HEALTH PLANS, INC. 4.5 Local CCP Aetna Inc. AETNA LIFE INSURANCE COMPANY 4.5 Local CCP HealthNow New York Inc. HEALTHNOW NEW YORK INC. 4.5 Local CCP Blue Cross Blue Shield of Michigan BLUE CARE NETWORK OF MICHIGAN 4.5 Local CCP Guidewell Mutual Holding Corporation CAPITAL HEALTH PLAN 4.5 4.5 H6622 Local CCP Humana Inc. HUMANA WI HEALTH ORGANIZATION INSURANCE CORP H6815 H7301 Local CCP Health Net, Inc. HEALTH NET HEALTH PLAN OF OREGON 4.5 Local CCP Aetna Inc. COVENTRY HEALTH CARE OF ILLINOIS, INC. 4.5 H7728 Local CCP Anthem Inc. ANTHEM HEALTH PLANS OF NEW HAMPSHIRE, INC. 4.5 Fallon Community Health Plan FALLON COMMUNITY HEALTH PLAN 4.5 Providence Health & Services PROVIDENCE HEALTH PLAN 4.5 MVP Health Care, Inc. MVP HEALTH PLAN, INC. 4.5 Local CCP H9001 Local CCP H9047 Local CCP H9615 SOURCE: CMS. 26 • AIS’s Management Insight Series Plans With 2.5 Stars or Fewer Contract Number Organization Type Parent Organization Contract Name 2016 Overall Rating H0571 Local CCP Chinese Hospital Association CHINESE COMMUNITY HEALTH PLAN 2.5 H1112 H1903 H3054 H3327 H4227 H4866 H5698 H5985 Local CCP WellCare Health Plans, Inc. WELLCARE OF GEORGIA, INC. 2.5 Local CCP WellCare Health Plans, Inc. WELLCARE OF LOUISIANA, INC. 2.5 Local CCP Constellation Health, LLC. CONSTELLATION HEALTH, LLC. 2.5 Local CCP Touchstone Health Partnership, Inc. TOUCHSTONE HEALTH HMO, INC. 2.5 Local CCP Independence Health Group, Inc. VISTA HEALTH PLAN, INC. 2.5 Local CCP Cuatro LLC. CUATRO LLC 2.5 Local CCP WellCare Health Plans, Inc. WINDSOR HEALTH PLAN, INC. 2.5 Local CCP Tenet Healthcare Corporation PHOENIX HEALTH PLANS, INC. 2.5 2.5 H6972 Local CCP CIGNA HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. H9285 Local CCP America's 1st Choice NY Holdings, LLC ATLANTIS HEALTH PLAN, INC. 2.5 WellCare Health Plans, Inc. WELLCARE HEALTH INSURANCE COMPANY OF KENTUCKY, INC 2.5 H9730 Local CCP SOURCE: CMS. Percentage of Plans In Each Star Rating Level 3.5 Stars 33.18% 3 Stars 25.29% 2.5 Stars 3.71% 4 Stars 20.19% SOURCE: CMS 5 Stars 2.55% 4.5 Stars 14.85% The Challenges of Pharmacy Star Ratings • 27 2016 Part D Star Ratings for Plans Serving Dual Eligibles CMS Contract ID Organization Name Plan Name Program Type Part D Star Rating H5991 AFFINITY HEALTH PLAN, INC. Affinity Medicare Solutions (HMO SNP) D-SNP 4.5 H5991 H5969 H9122 H3240 H7200 AFFINITY HEALTH PLAN, INC. Affinity Medicare Ultimate (HMO SNP) D-SNP 4.5 ALOHACARE AlohaCare Advantage Plus (HMO SNP) D-SNP 3.5 ALPHACARE OF NEW YORK, INC. AlphaCare Total (HMO SNP) D-SNP 2.5 AMERIGROUP NEW JERSEY, INC. Amerivantage Specialty + Rx (HMO SNP) D-SNP 4.0 AMERIGROUP TENNESSEE, INC. Amerivantage Specialty + Rx (HMO SNP) D-SNP 3.5 H1849 ANTHEM HEALTH PLANS OF KENTUCKY, INC. Anthem Dual Advantage (HMO SNP) D-SNP 4.0 H5854 ANTHEM HEALTH PLANS, INC. Anthem Dual Advantage (HMO SNP) D-SNP 4.5 H5619 ARCADIAN HEALTH PLAN, INC. Humana Gold Plus SNP-DE H5619-003 (HMO SNP) D-SNP 4.0 H0321 H0321 H0321 H3814 H3814 H5995 ARIZONA PHYSICIANS IPA, INC. UnitedHealthcare Dual Complete (HMO SNP) AZ D-SNP 3.0 ARIZONA PHYSICIANS IPA, INC. UnitedHealthcare Dual Complete (HMO SNP) D-SNP 3.0 ARIZONA PHYSICIANS IPA, INC. UnitedHealthcare Dual Complete ONE (HMO SNP) AZ D-SNP 3.0 ATRIO HEALTH PLANS ATRIO Special Needs Plan (HMO SNP) D-SNP 3.5 ATRIO HEALTH PLANS ATRIO Special Needs Plan (Rogue) (HMO SNP) D-SNP 3.5 ATRIO HEALTH PLANS ATRIO Special Needs Plan (Willamette) (HMO SNP) D-SNP 3.5 H5422 BLUE CROSS BLUE SHIELD HEALTHCARE PLAN OF GEORGIA BCBSHP Dual Advantage (HMO SNP) D-SNP 3.5 H0564 BLUE CROSS OF CALIFORNIA Anthem Dual Advantage (HMO SNP) D-SNP 4.0 5.0 H1350 BLUE CROSS OF IDAHO HEALTH SERVICE, INC. True Blue Special Needs Plan (HMO SNP) ID LTC+Managed Care H2425 BLUE PLUS SecureBlue (HMO SNP) CMS Duals Demo-MN 4.0 H2108 BRAVO HEALTH MID-ATLANTIC, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 3.0 H3949 BRAVO HEALTH PENNSYLVANIA, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 3.5 H5590 BRIDGEWAY HEALTH SOLUTIONS Bridgeway Health Solutions Advantage (HMO SNP) AZ D-SNP 4.0 H0908 BUCKEYE COMMUNITY HEALTH PLAN, INC. Buckeye Health Plan Advantage (HMO SNP) D-SNP 3.5 R6801 CARE IMPROVEMENT PLUS OF TEXAS INSURANCE COMPANY Care Improvement Plus Dual Advantage (Regional PPO SNP) D-SNP 3.5 R3444 CARE IMPROVEMENT PLUS SOUTH Care Improvement Plus Dual Advantage (Regional CENTRAL INSURANCE CO. PPO SNP) D-SNP 3.5 R9896 CARE IMPROVEMENT PLUS SOUTH Care Improvement Plus Dual Advantage (Regional CENTRAL INSURANCE CO. PPO SNP) D-SNP 3.5 H5209 CARE WISCONSIN HEALTH PLAN, INC. Partnership (HMO SNP) D-SNP 4.0 H5928 H0544 H1019 CARE1ST HEALTH PLAN Care1st TotalDual Plan (HMO SNP) D-SNP 3.5 CAREMORE HEALTH PLAN CareMore Connect (HMO SNP) D-SNP 5.0 CAREPLUS HEALTH PLANS, INC. CareNeeds (HMO SNP) D-SNP 5.0 H4461 CARITEN HEALTH PLAN INC. Humana Gold Plus SNP-DE H4461-022 (HMO SNP) D-SNP 5.0 H5649 CENTRAL HEALTH PLAN OF CALIFORNIA, INC. Central Health Medi-Medi Plan (HMO SNP) D-SNP 4.0 H0571 CHINESE COMMUNITY HEALTH PLAN CCHP Senior Select Program (HMO SNP) D-SNP 4.0 H0439 CIGNA HEALTHCARE OF GEORGIA, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 3.5 continued 28 • AIS’s Management Insight Series 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID Organization Name Plan Name Program Type Part D Star Rating H9725 CIGNA HEALTHCARE OF NORTH CAROLINA, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 4.0 H2225 COMMONWEALTH CARE ALLIANCE, Senior Care Options Program (HMO SNP) INC. D-SNP 5.0 H3071 COMMUNITY CARE ALLIANCE OF ILLINOIS, NFP Community Care Alliance of Illinois, NFP (HMO SNP) D-SNP 3.0 H2034 COMMUNITY CARE HEALTH PLAN, INC. Community Care's Partnership Program (HMO SNP) D-SNP 3.5 H2034 COMMUNITY CARE HEALTH PLAN, INC. Community Care's Partnership Program Disabled (HMO SNP) D-SNP 3.5 H5207 COMMUNITY CARE HEALTH PLAN, INC. Community Care's Partnership Program (HMO SNP) D-SNP 3.5 H3655 COMMUNITY INSURANCE COMPANY Anthem Dual Advantage (HMO SNP) D-SNP 4.5 H9525 COMPCARE HEALTH SERVICES INSURANCE CORPORATION Anthem Dual Advantage (HMO SNP) D-SNP 4.0 H3054 H4866 CONSTELLATION HEALTH, LLC. Genesis - Constellation Health (HMO SNP) D-SNP 3.0 CUATRO LLC Access Medicare Pearl (HMO SNP) D-SNP 3.0 H5608 DENVER HEALTH MEDICAL PLAN, INC. Denver Health Medicare Choice (HMO SNP) D-SNP 4.0 H5087 H5087 H3347 H3347 EASY CHOICE HEALTH PLAN INC. Easy Choice Access Plan (HMO SNP) D-SNP 3.5 EASY CHOICE HEALTH PLAN INC. Easy Choice Freedom Plan (HMO SNP) D-SNP 3.5 ELDERPLAN, INC. Elderplan For Medicaid Beneficiaries (HMO SNP) D-SNP 3.0 ELDERPLAN, INC. Elderplan Plus Long Term Care (HMO SNP) D-SNP 3.0 H3370 EMPIRE HEALTHCHOICE HMO, INC. Empire Dual Advantage (HMO SNP) D-SNP 4.0 H9001 FALLON COMMUNITY HEALTH PLAN NaviCare (HMO SNP) D-SNP 4.0 H3818 H5427 H5427 FAMILYCARE HEALTH PLANS, INC. FamilyCare Community (HMO SNP) D-SNP 4.0 FREEDOM HEALTH, INC. Freedom Medi-Medi Full (HMO SNP) D-SNP 4.5 FREEDOM HEALTH, INC. Freedom Medi-Medi Partial (HMO SNP) D-SNP 4.5 H9190 GATEWAY HEALTH PLAN OF OHIO, INC. Gateway Health Medicare Assured Diamond (HMO SNP) D-SNP 3.0 H9190 GATEWAY HEALTH PLAN OF OHIO, INC. Gateway Health Medicare Assured Ruby (HMO SNP) D-SNP 3.0 H5932 GATEWAY HEALTH PLAN, INC. Gateway Health Medicare Assured Diamond (HMO SNP) D-SNP 4.0 H5932 GATEWAY HEALTH PLAN, INC. Gateway Health Medicare Assured Ruby (HMO SNP) D-SNP 4.0 H3954 H5528 H6864 H5685 GEISINGER HEALTH PLAN Geisinger Gold Secure Rx (HMO SNP) D-SNP 4.0 GROUP HEALTH INCORPORATED EmblemHealth Dual Eligible (PPO SNP) D-SNP 3.5 GUILDNET, INC. GuildNet Gold (HMO-POS SNP) D-SNP 3.5 HAP MIDWEST HEALTH PLAN, INC. HAP Midwest Health Plan (HMO SNP) D-SNP 3.5 H1416 HARMONY HEALTH PLAN OF ILLINOIS, INC. WellCare Access (HMO SNP) D-SNP 3.0 H3822 HEALTH CARE SERVICE CORPORATION Blue Cross Medicare Advantage Dual Care (HMO SNP) D-SNP 3.5 H5587 HEALTH CHOICE ARIZONA, INC. Health Choice Generations (HMO SNP) AZ D-SNP 3.0 H3330 HEALTH INSURANCE PLAN OF GREATER NEW YORK EmblemHealth Dual Eligible (HMO SNP) D-SNP 3.5 H3330 HEALTH INSURANCE PLAN OF GREATER NEW YORK EmblemHealth Dual Eligible Group (HMO SNP) D-SNP 3.5 H0351 HEALTH NET OF ARIZONA, INC. Health Net Amber (HMO SNP) AZ D-SNP 4.5 The Challenges of Pharmacy Star Ratings • 29 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID Part D Star Rating Organization Name Plan Name Program Type H0562 H9207 HEALTH NET OF CALIFORNIA,INC. Health Net Seniority Plus Amber I (HMO SNP) D-SNP 4.5 HEALTH PARTNERS PLANS, INC. Health Partners Medicare Special (HMO SNP) D-SNP 3.0 H5859 HEALTH PLAN OF CAREOREGON, INC. CareOregon Advantage Plus (HMO-POS SNP) D-SNP 3.5 H3959 HEALTHAMERICA PENNSYLVANIA, INC. Advantra Cares (HMO SNP) D-SNP 4.0 H2422 HEALTHPARTNERS, INC. HealthPartners Minnesota Senior Health Options (HMO SNP) CMS Duals Demo-MN 4.5 H2165 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 4.5 H4513 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 4.0 H4528 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 3.5 H0150 H5410 HEALTHSPRING OF ALABAMA, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 4.0 HEALTHSPRING OF FLORIDA Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 5.0 H1415 HEALTHSPRING OF TENNESSEE, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 3.5 H4407 HEALTHSPRING OF TENNESSEE, INC. Cigna-HealthSpring TotalCare SMS (HMO SNP) D-SNP 4.5 H4454 HEALTHSPRING OF TENNESSEE, INC. Cigna-HealthSpring TotalCare (HMO SNP) D-SNP 5.0 H4454 HEALTHSPRING OF TENNESSEE, INC. Cigna-HealthSpring TotalCare AR (HMO SNP) D-SNP 5.0 H3672 HOMETOWN HEALTH PLAN SecureCare SNP (HMO SNP) D-SNP 4.0 H4141 HUMANA EMPLOYERS HEALTH PLAN OF GEORGIA, INC. Humana Gold Plus SNP-DE H4141-003 (HMO SNP) D-SNP 3.0 H4141 HUMANA EMPLOYERS HEALTH PLAN OF GEORGIA, INC. Humana Gold Plus SNP-DE H4141-010 (HMO SNP) D-SNP 3.0 H1951 HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. Humana Gold Plus SNP-DE H1951-032 (HMO SNP) D-SNP 4.5 H1951 HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. Humana Gold Plus SNP-DE H1951-033 (HMO SNP) D-SNP 4.5 H1951 HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. Humana Gold Plus SNP-DE H1951-034 (HMO SNP) D-SNP 4.5 H1951 HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. Humana Gold Plus SNP-DE H1951-041 (HMO SNP) D-SNP 4.5 H3533 HUMANA HEALTH COMPANY OF NEW YORK, INC. Humana Gold Plus SNP-DE H3533-002 (HMO SNP) D-SNP 3.0 H3533 HUMANA HEALTH COMPANY OF NEW YORK, INC. Humana Gold Plus SNP-DE H3533-004 (HMO SNP) D-SNP 3.0 H8953 HUMANA HEALTH PLAN OF OHIO, INC. Humana Gold Plus SNP-DE H8953-007 (HMO SNP) D-SNP 4.0 H8953 HUMANA HEALTH PLAN OF OHIO, INC. Humana Gold Plus SNP-DE H8953-014 (HMO SNP) D-SNP 4.0 H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-042 (HMO SNP) D-SNP 4.0 H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-053 (HMO SNP) D-SNP 4.0 H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-054 (HMO SNP) D-SNP 4.0 H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-064 (HMO SNP) D-SNP 4.0 H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-070 (HMO SNP) D-SNP 4.0 continued 30 • AIS’s Management Insight Series 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID Organization Name Plan Name Program Type Part D Star Rating H2012 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2012-095 (HMO SNP) D-SNP 4.0 H2649 HUMANA HEALTH PLAN, INC. Humana Gold Plus SNP-DE H2649-028 (HMO SNP) D-SNP 4.0 H4007 HUMANA HEALTH PLANS OF PUERTO RICO, INC. Humana Gold Plus SNP-DE H4007-016 (HMO SNP) D-SNP 3.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Community HMO SNP DE (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-077A (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-102 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-103A (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-104A (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-167 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-168 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-209 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-210 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-213 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-214 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-222 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-226 (HMO SNP) D-SNP 4.5 H1036 HUMANA MEDICAL PLAN, INC. Humana Gold Plus SNP-DE H1036-231 (HMO SNP) D-SNP 4.5 H2237 INDEPENDENT CARE HEALTH PLAN, INC. iCare Family Care Partnership (HMO SNP) D-SNP 4.0 H2237 INDEPENDENT CARE HEALTH PLAN, INC. iCare Medicare Plan (HMO SNP) D-SNP 4.0 H2417 ITASCA MEDICAL CARE IMCare Classic (HMO SNP) CMS Duals Demo-MN 4.0 H0630 KAISER FOUNDATION HP OF CO Senior Advantage Medicare Medicaid Plan (HMO SNP) D-SNP 5.0 H1170 KAISER FOUNDATION HP OF GA, INC. Senior Advantage Medicare Medicaid Plan (HMO SNP) D-SNP 4.0 H0524 KAISER FOUNDATION HP, INC. Senior Advantage Medicare Medi-Cal Plan North (HMO SNP) D-SNP 5.0 H0524 KAISER FOUNDATION HP, INC. Senior Advantage Medicare Medi-Cal Plan South (HMO SNP) D-SNP 5.0 H3337 LIBERTY HEALTH ADVANTAGE, INC. Liberty Health Advantage Dual Power (HMO SNP) D-SNP 3.0 H8189 MANAGED HEALTH SERVICES, WISCONSIN Managed Health Services Advantage (HMO SNP) D-SNP 3.5 H3359 H3359 MANAGED HEALTH, INC. Healthfirst CompleteCare (HMO SNP) D-SNP 4.0 MANAGED HEALTH, INC. Healthfirst Life Improvement Plan (HMO SNP) D-SNP 4.0 H6623 MARICOPA COUNTY SPECIAL HEALTH CARE DISTRICT Maricopa Care Advantage (HMO SNP) AZ D-SNP 3.5 H5577 MCS ADVANTAGE, INC. MCS Classicare Platino Ideal (HMO SNP) D-SNP 3.0 The Challenges of Pharmacy Star Ratings • 31 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID H5577 H5577 Part D Star Rating Organization Name Plan Name Program Type MCS ADVANTAGE, INC. MCS Classicare Platino Maximo (HMO SNP) D-SNP 3.0 MCS ADVANTAGE, INC. MCS Classicare Platino Superior (HMO SNP) D-SNP 3.0 4.0 H2458 MEDICA HEALTH PLANS Medica DUAL Solution (HMO SNP) CMS Duals Demo-MN H5420 MEDICA HEALTHCARE PLANS, INC. Medica HealthCare Plans MedicareMax Plus (HMO-POS SNP) D-SNP 4.0 H5779 MERIDIAN HEALTH PLAN OF ILLINOIS, INC. Meridian Advantage Plan of Illinois (HMO SNP) D-SNP 3.0 H5475 MERIDIAN HEALTH PLAN OF MICHIGAN, INC. Meridian Advantage Plan of Michigan (HMO SNP) D-SNP 3.0 H0423 METROPLUS HEALTH PLAN, INC. MetroPlus Advantage Plan (HMO SNP) D-SNP 3.5 H4003 MMM HEALTHCARE, LLC. Medicare y Mucho Mas - DIAMANTE CHOICE (HMO SNP) D-SNP 3.5 H4003 MMM HEALTHCARE, LLC. Medicare y Mucho Mas - DIAMANTE EXCEL (HMO SNP) D-SNP 3.5 H4003 MMM HEALTHCARE, LLC. Medicare y Mucho Mas - DIAMANTE EXTRA (HMO SNP) D-SNP 3.5 H5810 MOLINA HEALTHCARE OF CALIFORNIA Molina Medicare Options Plus (HMO SNP) D-SNP 3.5 H8130 MOLINA HEALTHCARE OF FLORIDA, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 3.5 H5926 MOLINA HEALTHCARE OF MICHIGAN, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 4.0 H9082 MOLINA HEALTHCARE OF NEW MEXICO, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 4.0 H0490 MOLINA HEALTHCARE OF OHIO, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 3.0 H7678 MOLINA HEALTHCARE OF TEXAS, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 4.0 H5628 MOLINA HEALTHCARE OF UTAH, INC. Healthy Advantage (HMO SNP) D-SNP 4.5 H5628 MOLINA HEALTHCARE OF UTAH, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 4.5 H5823 MOLINA HEALTHCARE OF WASHINGTON, INC. Molina Medicare Options Plus (HMO SNP) D-SNP 4.0 H2879 MOLINA HEALTHCARE OF WISCONSIN Molina Medicare Options Plus (HMO SNP) D-SNP 4.5 H5215 NETWORK HEALTH INSURANCE CORPORATION NetworkCares (PPO SNP) D-SNP 4.5 H5430 ONECARE BY CARE1ST HEALTH PLAN ARIZONA INC. Care1st+ (HMO SNP) AZ D-SNP 3.5 H5594 H5594 OPTIMUM HEALTHCARE, INC. Optimum Emerald Full (HMO SNP) D-SNP 4.5 OPTIMUM HEALTHCARE, INC. Optimum Emerald Partial (HMO SNP) D-SNP 4.5 H5433 ORANGE COUNTY HEALTH AUTHORITY OneCare (HMO SNP) D-SNP 3.5 H3113 H1961 H5985 OXFORD HEALTH PLANS (NJ), INC. UnitedHealthcare Dual Complete ONE (HMO SNP) D-SNP 5.0 PEOPLES HEALTH, INC. Peoples Health Secure Health (HMO SNP) D-SNP 4.0 PHOENIX HEALTH PLANS, INC. Phoenix Advantage Plus (HMO SNP) AZ D-SNP 3.0 H4527 PHYSICIANS HEALTH CHOICE OF TEXAS LLC UnitedHealthcare Dual Complete Focus (HMO SNP) D-SNP 4.0 H4004 H1045 PMC MEDICARE CHOICE, LLC Premier Preferred (HMO SNP) D-SNP 3.5 PREFERRED CARE PARTNERS INC. Preferred Medicare Assist (HMO-POS SNP) D-SNP 3.5 H2416 PRIMEWEST CTRL COUNTY-BASED PrimeWest Senior Health Complete (HMO SNP) PURCHASING INITIATIVE CMS Duals Demo-MN 4.0 continued 32 • AIS’s Management Insight Series 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID Organization Name Plan Name Program Type Part D Star Rating H2773 QUALITY HEALTH PLANS OF NEW YORK, INC. Advantage Value One NY - Dual (HMO SNP) D-SNP 4.0 H3811 SAMARITAN HEALTH PLANS, INC. Samaritan Advantage Special Needs Plan (HMO SNP) D-SNP 4.0 H5428 H5425 H5425 H2224 H2224 H4525 H5471 SAN MATEO HEALTH COMMISSION CareAdvantage (HMO SNP) D-SNP 4.0 SCAN HEALTH PLAN SCAN Connections (HMO SNP) D-SNP 4.5 SCAN HEALTH PLAN SCAN Connections at Home (HMO SNP) D-SNP 4.5 SENIOR WHOLE HEALTH, LLC Senior Whole Health (HMO SNP) D-SNP 4.0 SENIOR WHOLE HEALTH, LLC Senior Whole Health NHC (HMO SNP) D-SNP 4.0 SHA, L.L.C FirstCare Advantage Select (HMO SNP) D-SNP 3.5 SIMPLY HEALTHCARE PLANS, INC. Simply Complete (HMO SNP) D-SNP 3.5 H2419 SOUTH COUNTRY HEALTH ALLIANCE SeniorCare Complete (HMO SNP) CMS Duals Demo-MN 5.0 H5703 SOUTH COUNTRY HEALTH ALLIANCE AbilityCare (HMO SNP) D-SNP 3.5 H5580 SOUTHWEST CATHOLIC HEALTH NETWORK CORPORATION Mercy Care Advantage (HMO SNP) AZ D-SNP 3.5 H5190 SUNSHINE STATE HEALTH PLAN, INC. Sunshine Health Advantage (HMO SNP) D-SNP 3.5 H5294 SUPERIOR HEALTH PLAN, INC. Superior HealthPlan Advantage (HMO SNP) D-SNP 3.5 H3328 THE NEW YORK STATE CATHOLIC HEALTH PLAN, INC. Fidelis Dual Advantage (HMO SNP) D-SNP 4.5 H3328 THE NEW YORK STATE CATHOLIC HEALTH PLAN, INC. Fidelis Dual Advantage Flex (HMO SNP) D-SNP 4.5 H3328 THE NEW YORK STATE CATHOLIC HEALTH PLAN, INC. Fidelis Medicaid Advantage Plus (HMO SNP) D-SNP 4.5 H2174 TRILLIUM COMMUNITY HEALTH PLAN Trillium Advantage Dual (HMO SNP) D-SNP 4.0 H2256 TUFTS ASSOCIATED HEALTH MAINTENANCE ORGANIZATION Tufts Health Plan Senior Care Options (HMO SNP) D-SNP 4.5 H2456 UCARE MINNESOTA UCare's Minnesota Senior Health Options (HMO SNP) CMS Duals Demo-MN 3.5 H4590 UNITEDHEALTHCARE BENEFITS OF UnitedHealthcare Dual Complete (HMO SNP) TEXAS, INC. D-SNP 4.0 H4514 UNITEDHEALTHCARE COMMUNITY PLAN OF TEXAS, LLC UnitedHealthcare Dual Complete (HMO SNP) D-SNP 3.5 H0624 UNITEDHEALTHCARE INSURANCE COMPANY UnitedHealthcare Dual Complete (HMO SNP) D-SNP 3.5 H2226 UNITEDHEALTHCARE INSURANCE COMPANY UnitedHealthcare Senior Care Options (HMO SNP) D-SNP 5.0 H5008 UNITEDHEALTHCARE INSURANCE COMPANY UnitedHealthcare Dual Complete (HMO SNP) D-SNP 3.5 R3175 UNITEDHEALTHCARE INSURANCE COMPANY UnitedHealthcare Dual Complete RP (Regional PPO SNP) D-SNP 4.0 R5287 UNITEDHEALTHCARE INSURANCE COMPANY UnitedHealthcare Dual Complete RP (Regional PPO SNP) D-SNP 3.5 H0151 UNITEDHEALTHCARE OF ALABAMA, INC. UnitedHealthcare Dual Complete (HMO SNP) D-SNP 3.5 H3387 UNITEDHEALTHCARE OF NEW YORK, INC. UnitedHealthcare Dual Complete (HMO SNP) D-SNP 4.0 H5253 UNITEDHEALTHCARE OF WISCONSIN, INC. UnitedHealthcare Dual Complete LP (HMO SNP) D-SNP 4.5 H0251 UNITEDHEALTHCARE PLAN OF THE UnitedHealthcare Dual Complete (HMO SNP) RIVER VALLEY, INC. D-SNP 4.0 The Challenges of Pharmacy Star Ratings • 33 2016 Part D Star Ratings for Plans Serving Dual Eligibles (continued) CMS Contract ID Part D Star Rating Organization Name Plan Name Program Type H0838 H4279 UNIVERSAL CARE, INC. Dual Coverage (HMO SNP) D-SNP 3.5 UPMC FOR YOU, INC UPMC for You Advantage (HMO SNP) D-SNP 4.0 H2161 UPPER PENINSULA HEALTH PLAN, LLC Upper Peninsula Health Plan Plus (HMO SNP) D-SNP 3.5 H4227 H4227 H0154 H5549 H5549 H5549 H3259 VISTA HEALTH PLAN, INC. AmeriHealth VIP Care (HMO SNP) D-SNP 3.5 VISTA HEALTH PLAN, INC. Keystone VIP Choice (HMO SNP) D-SNP 3.5 VIVA HEALTH, INC. VIVA Medicare Extra Value (HMO SNP) D-SNP 4.0 VNS CHOICE VNSNY CHOICE Medicare Maximum (HMO SNP) D-SNP 4.0 VNS CHOICE VNSNY CHOICE Medicare Preferred (HMO SNP) D-SNP 4.0 VNS CHOICE VNSNY CHOICE Total (HMO SNP) D-SNP 4.0 VOLUNTEER STATE HEALTH PLAN BlueCare Plus (HMO SNP) D-SNP 3.5 H9730 WELLCARE HEALTH INSURANCE COMPANY OF KENTUCKY, INC Wellcare Access (HMO SNP) D-SNP 2.5 H2491 WELLCARE HEALTH INSURANCE OF ARIZONA, INC. 'Ohana Liberty (HMO-POS SNP) D-SNP 3.0 H0712 H1032 H1032 H1032 H1112 H1903 H3361 H3361 H1264 H5698 H5698 WELLCARE OF CONNECTICUT, INC. WellCare Access (HMO SNP) D-SNP 3.0 WELLCARE OF FLORIDA, INC. WellCare Access (HMO SNP) D-SNP 3.5 WELLCARE OF FLORIDA, INC. WellCare Liberty (HMO SNP) D-SNP 3.5 WELLCARE OF FLORIDA, INC. WellCare Select (HMO SNP) D-SNP 3.5 WELLCARE OF GEORGIA, INC. WellCare Access (HMO SNP) D-SNP 2.5 WELLCARE OF LOUISIANA, INC. WellCare Access (HMO SNP) D-SNP 2.5 WELLCARE OF NEW YORK, INC. WellCare Access (HMO SNP) D-SNP 3.0 WELLCARE OF NEW YORK, INC. WellCare Liberty (HMO SNP) D-SNP 3.0 WELLCARE OF TEXAS, INC. WellCare Access (HMO SNP) D-SNP 3.5 WINDSOR HEALTH PLAN, INC. WellCare Access (HMO SNP) D-SNP 3.0 WINDSOR HEALTH PLAN, INC. WellCare Comp Access (HMO SNP) D-SNP 3.0 SOURCE: DUAL Medicare-Medicaid Dual Eligibles Database, an online subscription database and newsfeed from AIS. Visit http:// aishealthdata.com/dual for more information and free interactive demo. 34 • AIS’s Management Insight Series CMS Regulation and Guidance CMS Suspension of Automatic Star-Rating Cuts for Sanctioned Plans Stuns Observers An unusual March 8, 2016, memo from CMS is causing both astonishment and thankfulness in portions of the Medicare Advantage industry because of its potential impact on the star quality rating system and its effect on plan revenues and member benefits. Other industry segments, along with a few securities analysts and consumer advocates, wondered why CMS was taking an action that so directly helped one large insurer, Cigna Corp., which otherwise would have had its high star ratings — and the upcoming bonuses tied to them — sharply cut back as of March 31, 2016. The March 8, 2016, memo, not publicized by CMS, suspended the automatic and major reductions in star ratings that Cigna and another much smaller MA plan would incur if they are still under the agency’s so-called “intermediate sanctions” as of the end of March 2016. “The suspension is a significant positive for Cigna,” wrote Barclays securities analyst Joshua Raskin in a March 8, 2016, research note. “Candidly, this is the first time that we can remember CMS making such a significant policy change for a specific carrier” without doing it via a proposed or final rule. “We are discouraged by CMS’ announcement, which diminishes accountability and transparency related to plan sanctions,” Stacy Sanders, federal policy director at the Medicare Rights Center, told Kaiser Health News. If CMS was concerned that the requirement mandating a reduction to 2.5 stars for a plan undergoing the agency’s intermediate sanctions was too severe for a high-star-rated MA organization (MAO), she said, it could have elected to drop star ratings by just one star as it does for low-rated plans rather than suspend star reduction for such sanctioned plans entirely. CMS in a Jan. 21, 2016, letter from its Medicare Parts C and D Oversight and Enforcement Group imposed an indefinite suspension on Cigna’s marketing and enrollment for new members in both its MA and stand-alone Prescription Drug Plan (PDP) products. The agency cited a host of severe deficiencies — including in coverage determinations and handling of appeals and grievances — and said that Cigna has a “longstanding history of noncompliance.” CMS May Have Feared Market Destabilization Observers queried by AIS did not attribute the policy change to an agency desire to show largesse to Cigna, which stands to gain between $213 million and $350 million from the shift based on estimates by securities firms. But some of the observers do suggest that the size of Cigna, which had more than 550,000 MA enrollees as of CMS’s February 2016 monthly data, might have led the agency to fear that slashing the insur- The Challenges of Pharmacy Star Ratings • 35 er’s ratings and thus indirectly its benefits for 2017 might have destabilized substantial portions of the MA market during the Annual Election Period (AEP) in fall 2015. Specifically, Cigna/HealthSpring had more than 350,000 members as of the February 2016 data in five MA plans rated four stars and above for 2016, including 52,485 in a five-star plan in Arizona that was one of only 12 nationwide to earn this top honor plus nearly 150,000 in 4.5-star plans. And Cigna’s situation takes on greater significance because it has agreed to be acquired by Anthem, Inc. in a still-pending deal. CMS for its part did not cite the potential impact on the AEP as a reason for the policy change. The four-paragraph memo from Jennifer Shapiro, acting director of the Medicare Drug Benefit and C & D Data Group, sent out only via the electronic Health Plan Management System (HPMS) and not publicized, instead emphasized the changing star-rating achievement background. “When CMS announced this policy for the 2012 Star Ratings,” Shapiro wrote, “relatively few contracts achieved ratings of 4 stars or above and fewer than 30% of Medicare Advantage plan enrollees were in plans offered under these highly rated contracts. Today, 49% of MA contracts, representing 71% of MA plan enrollees, have achieved ratings of 4 stars or above, compared to an estimated 17% in 2009.” In response to the draft 2017 Call Letter issued Feb. 19, 2016, the memo said, CMS received “multiple comments” suggesting that it revise the mandatory-stars-reduction policy for “sanctioned contracts.” Shapiro added that “commenters raised several concerns, including one noting that high-rated contracts can be subjected to a more severe penalty than low-rated contracts” since the higher ones could lose multiple stars while low-rated contracts could lose only one star. MAOs have made this argument before, but their pleas have become more urgent in recent years. Among the commenters seeking a change was not only the America’s Health Insurance Plans trade group but also CAPG, which represents the risk-bearing provider groups that have become big players in MA. After having considered the comments “and the growth in the number of highly rated contracts, CMS agrees that we should reassess the impact of intermediate sanctions on the calculation of star ratings,” the memo said. So “effective immediately and on a prospective basis, CMS is suspending the automatic sanctions-based reduction in Star Ratings,” according to Shapiro. She wrote that CMS will propose a revised approach in the draft 2018 Call Letter, which comes out in February 2017. Action Was Needed Before March 31 The reason for announcing the change through an HPMS memo, Shapiro continued, was so that the policy change could be applied prior to the March 31, 2016, deadline for making adjustments to a contract’s star ratings based on its sanction status. March 31 is the date CMS’s Medicare Plan Finder and Quality Bonus Payment (QBP) determinations are adjusted for such changes, so the memo before that date allows 36 • AIS’s Management Insight Series CMS to give this information to MA sponsors before those data updates without changing its “established timelines,” Shapiro wrote. This means that CMS will use the original 2016 star ratings, unless changed via CMS’s appeals process, to determine QBP eligibility for 2017. The whole timetable is very significant since there appears to be next to no chance that Cigna would be able to convince CMS by March 31, 2016, to lift the intermediate sanctions. Aetna Inc., which was hit with those MA and PDP sanctions in April 2010, did not have them lifted until June 2011, and intermediate sanctions for Health Net, Inc. lasted from November 2010 to August 2011. Consultants, including two former CMS MA officials, had mixed reactions to the CMS memo. Avalere Health Vice President Tom Kornfield, who joined the company from CMS in 2015, says the impact of the major lowering of Cigna’s high star ratings that would have occurred were it not for the memo could have had a significant impact on beneficiaries, including loss of benefits and even possible loss of some of its MA plans. But Kornfield also tells AIS “I’m not certain I understand why CMS didn’t propose or mention this” issue in the draft Call Letter. “That would have been an appropriate vehicle” for such a policy change, while doing it via an HPMS memo seems “out of left field,” he asserts. Asked if there might have been political factors involved in the decision since nearly half of the next AEP falls during the final weeks of campaigning in a big election year when an adverse impact on MA premiums and benefits could be an issue, Kornfield replies that he doesn’t know. He points out it’s possible Cigna has been making rapid progress on remedying the deficiencies and that the policy change could be one way of recognizing that. The only other MA plan that appears now to be in MA intermediate sanctions is Ultimate Health Plans, Inc. in Florida, but Cigna is far larger. The issue CMS is addressing in the memo seems to be a legitimate one, suggests Michael Adelberg, a former top CMS MA official who now is senior director at FaegreBD Consulting. “As more beneficiaries come into 4- and 5-star-rated contracts,” he tells AIS, “the impact of dropping those contracts to 2.5 grows. This is not just an issue for the MAO, but also for the beneficiaries when the MAO submits bids based on the payment reductions. Large numbers of beneficiaries with reduced benefits creates difficulties for everyone.” Asked about making the change via an HPMS memo, Adelberg replies, “This was not a ho-hum HPMS memo. Most HPMS memos offer new technical or administrative guidance. This one was about policy.” Is there the danger of a mixed message on the importance of CMS sanctions when the agency does something like this? “With or without a star-ratings reduction,” responds Adelberg, “the potential of missing an open-enrollment season forces appropriate attention from the sanctioned MAO.” The Challenges of Pharmacy Star Ratings • 37 Star Ratings for Duals Inovalon Study Shows Duals’ Characteristics Yield Poor Health Outcomes Health plans with large percentages of dual-eligible beneficiaries may not be at fault for the overall poor health outcomes reflected in their Medicare Advantage star ratings, suggests a March 31, 2015, study from Inovalon, Inc. The study, which analyzed claims data from more than 2.2 million MA plan members, may be the first large-scale analysis to find a distinction between a plan’s quality of care and the health outcomes related to its members’ dual status. Analysts found that non-duals experienced better health outcomes regardless of the proportion of dual eligible members served by the same plan, even if that plan had lower-than-average star ratings. Researchers determined that factors outside of a health plan’s control, such as socioeconomic status, chronic health issues and the surrounding community accounted for more than 70% of the performance gap between dual and non-dual members. Hospital readmissions, a measure that gets the top three-point weight in star ratings, were attributed to living in a poor neighborhood in over 18% of cases, according to Inovalon. When compared to non-dual members, duals live in poverty at much higher rates. Health problems that are more likely to plague duals, such as dementia, schizophrenia and chronic obstructive pulmonary disease, are also common causes of readmission. Duals were also found to be more likely to live in communities with limited access to physicians, another significant cause of high readmission rates and lower adherence to care. Because of the complexity of these factors, the Inovalon researchers recommended that CMS adjust its Star Ratings measures to account for the high-risk characteristics observed in the duals population. Establishing new guidelines would help pave the way for accuracy in evaluating quality of care for duals, and plans could move forward without being penalized for issues that are beyond their control, the report said. Personal Engagement, Communication Technology Could Boost Star Ratings and Health Outcomes for Duals Plans Creating a superior experience for seniors is a crucial part of running a successful, top-rated Medicare Advantage (MA) plan, but there is no one-size-fits-all approach to member engagement, especially for plans that serve medically frail populations such as Medicare-Medicaid dual-eligibles. “[Seniors] have different needs and the way we need to communicate with them is different,” Jim Fields, a government programs and healthcare reform specialist at 38 • AIS’s Management Insight Series Oliver Wyman said at a June 11, 2015, Welltok webinar on keeping seniors engaged with their health outcomes. Fields said MA plans should focus on authentic and empathetic experiences between members and providers to foster completion of the Agency for Healthcare Research and Quality’s Consumer Assessment of Healthcare Providers Systems (CAHPS) surveys and ensure seniors take an active role in their healthcare. These steps can be much more difficult, however, for plans serving duals, largely because of longstanding socioeconomic and health literacy issues. “[Duals] may be less independent, so the channels you use may need to shift,” Fields said. “You often have to help them take the step to get them into the provider and follow through on care guidelines. It’s a more intensive, more hand-holding activated model.” Since duals are less likely to have reliable access to transportation, Fields asserted in-home vendors and certain technology could “without a doubt” improve the population’s adherence to care and opinion of the healthcare system. “As you get into the Medicaid space, [members are] more likely to have a cell phone than they are to have a permanent address,” he said. And though some duals may be too frail to use a phone the same way other seniors can, in-home technology plans can install to track medication schedules and relay that information back to providers could prove invaluable for duals. In addition to adherence to care, consumer engagement through Internet technology can also increase a senior’s likelihood to complete a CAHPS survey, a key part of measuring a plan’s star rating. Plans with dual membership are known to struggle with star ratings improvement, so influencing these habits could be game-changing. “Even a snapshot on their screen that says ‘Congratulations on getting your flu shot this year,’ has a direct effect on how people respond to the CAHPS survey,” Fields said. “Seniors are increasingly receptive to the Internet as a channel for engagement.” Another method plans can use to improve health outcomes and boost star ratings is creating incentives for seniors to adhere to best care practices, such as keeping regular appointments with providers and receiving recommended inoculations. Incentives, which can range from gift cards to plan savings, may prove enticing for duals. “Monetary rewards are the most effective,” Fields said. “Plan savings is often harder for folks to see, as opposed to a tangible monetary thing.” Diabetes Management Could Help Boost Stars for Struggling Duals Plans An increased focus on chronic disease management programs may be one key to helping struggling Medicare Advantage (MA) plans achieve higher star ratings, experts discussed at AIS’s Oct. 22, 2015, virtual conference, 2016 Medicare Star Ratings: Translating Trends Into Strategies. This is particularly true for Dual-Eligible Special The Challenges of Pharmacy Star Ratings • 39 Needs Plans (D-SNPs), which consistently score ratings too low to be eligible for CMS bonuses; further penalizing plans that in some cases are already struggling to stay afloat. Robert Brett, vice president of Medicare at CareSource, mentioned several areas that have been most challenging for duals plans, including osteoporosis management and cancer screenings. Focusing on any of these areas could improve performance, but Brett recommended plans focus on four diabetes measures as the most cost-effective way to achieve a ratings boost: u C13 Diabetes Care – Eye Exam; u C14 Diabetes Care – Kidney Disease Monitoring; u C15 Diabetes Care – Blood Sugar Controlled; u D12 Medication Adherence for Diabetes Medication. “Diabetes is a key area to focus on,” he said, adding that these four measures make up 10% of a plan’s overall star rating. “It’s something where you can get some bang from your buck when talking about interventions.” Brett outlined several strategies plans can implement to improve diabetes management, including: u Improving monitoring of targeted members and performing trend analysis, u Improving member outreach, u Facilitating close interaction between members and their primary care physicians. Member outreach has proven critical to better health outcomes, but this would be a huge obstacle for duals plans to overcome, according to Stephen Wood, co-founder of Clear View Solutions, LLC, who also spoke at the conference. “In many plans, over half of the phone numbers that they have for their duals members are bad,” he said. “How are you going to go out and have folks get reminded about having their breast cancer screening if you can’t get a hold of them?” Brett mentioned several solutions for finding hard-to-reach members, such as sharing data with pharmacy benefit managers, which often have more accurate contact information on file, and working with physicians to obtain that data from members directly whenever possible. While these interventions may prove helpful, the overall stars outlook is still gloomy for duals plans. Although four D-SNPs achieved overall 5-star ratings for 2016, Wood said that can be misleading, and needs to be examined more carefully. “If you peeled the onion and looked at [SNP enrollment] as part of the overall plan, it was a tiny fraction of the plan’s enrollment,” he said. With 90% of plans scoring between 3 and 3.5 stars in 2015 failing to improve performance in 2016, Wood said he is not particularly optimistic that plans scoring even lower than that will ever be able to recover. 40 • AIS’s Management Insight Series Stars Memo Offers Two Complex Options to Aid Plans Serving Low-SES Members There seems to be no question about what CMS regarded as the most important issue it needed to deal with in its annual star ratings enhancements memo, which the agency disseminated Nov. 12, 2015. CMS spent more than seven pages outlining two complex approaches it is considering for adjusting 2017 star ratings to take into account the impact of socioeconomic status (SES) and disability. While industry executives and consultants tell AIS they appreciate the hard work and desire to help that CMS is exhibiting in the proposals, they have doubts about how effective either option would be. “I’m afraid there’s no silver bullet here,” says, for instance, Christie Teigland, Ph.D., senior director of statistical research at health data analytics firm Inovalon, Inc., which did a major study of the issue using data and other input from six MA sponsors and two trade groups in 2014. There were some other major subjects in the new stars memo, which came from Amy Larrick, acting director of CMS’s Medicare Drug Benefit and C & D Data Group. Those subjects included a warning that the agency would be scrutinizing plans’ medication therapy management (MTM) programs for possible evidence of biased data. And CMS said it would remove two measures from use in scoring star ratings for 2017 and is making methodology changes on six other measures. But the agency is not adding any star scoring measures for 2017, and even a new adjustment it reports considering for Puerto Rico contracts in some ways involves the same kind of SES-related issues at the apparent forefront of its stars agenda for 2017 and beyond. In general, Melissa Smith, senior consultant at Gorman Health Group, LLC, tells AIS, she is “not terribly surprised” by anything in the stars memo. She calls the absence of new stars measures for 2017 “welcome news” for health plans. They also will be heartened by CMS’s temporary removal, outlined in the memo, of the measure on improving bladder control, which MA sponsors generally have not done well on, she says. The other temporarily withdrawn measure, on high risk medication (HRM), is in a different situation since 330 plans were rated at four stars or above on it, Smith notes. And one of the reasons CMS gave for this planned withdrawal is that “there is a significant association between dual eligible/low income status and HRM use….We recommend that the measure developers further review this measure to better understand the associations.” In the meantime, CMS apparently wants to finalize its own interim approach to the SES/disability issue in stars. Its major new step toward that end, outlined after the agency cautioned that “CMS’ work is not complete,” is presenting two alternative approaches to adjusting for such disadvantaged status. The first one, which CMS called the Categorical Adjustment Index, is a factor added to or subtracted from a contract’s overall and/or summary (i.e., MA or Part D) The Challenges of Pharmacy Star Ratings • 41 star rating to adjust “for the average within-contract disparity” based on the contract’s percentages of low-income subsidy (LIS)/dual eligible (DE) and/or disabled beneficiaries. “This approach is equivalent to case-mix adjustment or patient-mix adjustment,” CMS explained before detailing the specific ways it could be derived. The second option, which CMS called Indirect Standardization, “would be applied to a subset of the individual star ratings measure scores.” The adjustment would focus, according to the memo, on the “within-contract LIS/DE and/or disability status difference while allowing for the existence of true differences in quality by contract.” The method would calculate an “expected measure score” using the percentages of these disadvantaged groups per measure “multiplied by the adjusted mean national performance for each subgroup.” Then, CMS added, “each contract would be judged against its expected performance,” and this observed-to-expected ratio would be multiplied by the adjusted national mean performance for all Medicare beneficiaries to get an “adjusted measure score.” The problem on both approaches, according to Teigland, is that there is “lots of missing information we’d need to evaluate.” Among details she says are needed to assess the options are how many and which star measures the methods would be applied to. Teigland calls the first option a “much more blunt approach” than the second, which she reports “leaning toward” preferring. But if the first option used a broader set of risk-adjustment measures — such as adjustments for age, gender, income and prevalence of chronic conditions — that “would help,” Teigland says. Whichever option CMS selects (and it raised the third possibility of a hybrid of the two), “my gut is it will not have much impact” on plans serving highly disadvantaged populations, Teigland tells AIS. She fears this could be a step back toward the onlysmall-changes-needed position CMS took Sept. 10, 2015, in unveiling its new research on the impact of duals/LIS status on stars performance, before both the agency’s acting administrator and Center for Medicare director indicated about a month later that meaningful aid is on the way. But Teigland also is quick to acknowledge that CMS seems to be really trying to give short-term relief to plans that need it. Inovalon is formulating questions to ask Larrick about the approaches, Teigland says, and she praises the responsiveness of Larrick’s group in the past. However, regardless of how quick it is in answering, Inovalon would not be able to do all the needed modeling before the Dec. 10, 2015, comment deadline, Teigland says. So she’s hoping indications that CMS is willing to keep making changes until the advance 2017 MA payment notice and Call Letter is released in mid-February 2016 were correct. John Baackes, CEO of L.A. Care Health Plan, which serves many duals and other disadvantaged populations, looks at the big picture on the proposed adjustments and calls it “good news.” While there weren’t sufficient useful data before for CMS to make pay adjustments for these populations, this is changing, he suggests. And he tells AIS that the kinds of options CMS is considering in the memo could have a “material” beneficial impact on L.A. Care’s finances. 42 • AIS’s Management Insight Series In looking at other portions of the stars memo, Smith sees as significant the degree to which CMS intends in the coming years to introduce more measures related to medication. Among the new measures the memo lists as under consideration for 2018 and beyond, for example, are ones related to “medication reconciliation post discharge,” statin therapy for patients with cardiovascular disease and statin use in persons with diabetes. CMS said all of those will be added to the display page (i.e., not used for actual stars calculations) for 2017 and that it intends to use them for computing star ratings in 2018. Moreover, the agency said it will add three measures related to use of opioids from multiple providers or at high dosage in persons with cancer to the display page in 2018, although it does not yet recommend adding them for actual star ratings because of several methodology concerns. It has plans for a new measure on antipsychotic use in persons with dementia for similar reasons. The agency also is concerned about how plans are acting on existing medication measures related to MTM. “We intend to review and apply any relevant MTM program audit findings that could demonstrate sponsors’ MTM data were biased, outside of the Data Validation results,” the memo said. CMS explained that it suspects such sponsor activities as “attempts to restrict eligibility from their approved MTM programs, encouraging beneficiary opt-out of MTM programs within the first 60 days, or CMRs [i.e., Comprehensive Medication Reviews] that do not meet CMS’ definition per guidance. CMS may perform additional audits or reviews to ensure the validity of data for specific contracts. Without rigorous validation of star ratings data, there is risk that CMS would reward contracts with falsely high ratings.” Smith says what her consulting firm is hearing is that some plans are “working aggressively to manage the denominator” in MTM calculations such as by removing member opt-outs from it to get higher fractions for purposes of boosting star-measure scores. She advises plans to review their “work flows and decision trees” on MTM, especially regarding opt-out criteria, to ensure they are meeting beneficiary needs and the goals of CMS. The agency took a full page in the 27-page memo to discuss an “additional adjustment for contracts in Puerto Rico,” which does not have an LIS, to identify beneficiaries there whose incomes would qualify for an LIS designation in the mainland. This would be used in conjunction with whichever option CMS selects on the SES/disabled issue to try to help beneficiaries and plans there. The memo also points to “unique challenges that Puerto Rican contracts face in improving medication adherence” because beneficiaries there have high out-of-pocket costs. CMS considered reducing weights of the three Part D medication adherence measures in Puerto Rico but rejected that when its modeling indicated the highest ratings would be unchanged, and asked for other proposals “to account for the barriers” there. The Challenges of Pharmacy Star Ratings • 43 Trade Groups Cite Doubts About CMS’s Two Options for Adjusting Stars for SES Trade associations representing Medicare Advantage plans serving disadvantaged populations expressed numerous reservations about the two options CMS in November 2015 requested comments on for offering interim relief on star quality ratings for such plans. In their comment letters filed by the Dec. 10, 2015, CMS deadline, the associations all thanked CMS for seeking ways to help these financially ailing plan sponsors in the short run, but said they had numerous questions remaining even after a Dec. 3, 2015, one-hour CMS conference call on the two options. And they generally voiced doubts about how effective either option would be in granting meaningful relief. At least one association alluded to another potentially troublesome issue that came up in the agency’s responses on the conference call. Regardless of which option CMS chooses, affected plans may not be able to preview their star ratings for the next year until significantly later in the year than they do now. And if CMS adopts the Indirect Standardization (IS) option, it needs to be based on the current year’s data, adding more uncertainties to the process for insurers. The reservations expressed in the comment letters could lead CMS not to adopt either option. That would parallel action it took in 2015 when it dropped a proposal to reduce the weighting of certain stars measures that plans serving populations with low socioeconomic status (SES) do poorly on after the vast majority of the comments it received were critical of that approach. In their comments, stakeholders praised aspects of both the IS option and the Categorical Adjustment Index (CAI) alternative, which was favored by the Medicare Payment Advisory Commission (MedPAC), among others. CAI uses a factor added to or subtracted from a contract’s overall and/or summary star rating to adjust withincontract disparities based on an MA contract’s percentages of low-income subsidy (LIS)/Medicare-Medicaid dual eligible (DE) and/or disabled beneficiaries. IS instead would adjust a subset of individual star ratings scores to account for within-contract LIS/DE and/or disability status differences not related to plan quality and would use an “expected measure score” as its key standard. MedPAC, in its Dec. 2, 2015, letter from Chairman Francis (Jay) Crosson, M.D., said it has “two concerns” regarding the IS approach. One of them, according to the letter, is that several of the measures for which CMS and MedPAC “found population-based differences are measures that are reported based on medical record sampling (generally 411 records)” or on which some contracts report based on sampling while others report based on “the universe of enrollees to whom the measure applies.” The question MedPAC has, explained Crosson, is whether such a sample “yields a sufficient number of records for a subpopulation within a contract to be able to determine a valid measure result for the subpopulation.” 44 • AIS’s Management Insight Series He said the second concern is that “if all enrollees within a subpopulation are used to determine an all-contract expected rate, then undue weight would be given to contracts that report based on the universe of enrollees to whom the measure applied.” An MA contract with far more diabetic enrollees than another plan has would yield a disproportionate impact on the key calculation of the expected rate, Crosson observed. By contrast, he asserted, the CAI “is administratively less complicated but still addresses the concerns plans have raised.” So “we would thus urge CMS to implement this approach as an interim measure” given CMS’s desire to offer interim relief, he added. ACAP Seeks Examples on Duals Plans The Association for Community Affiliated Plans, on the other hand, did not take a position on the two options, contending that more information from CMS is needed before ACAP can determine how well the two approaches “adjust for SES.” Specifically, wrote ACAP CEO Meg Murray in the comments letter, it would like more detail and numeric examples to show this, “particularly for contracts with majority or 100% dual [i.e., Medicare-Medicaid dual eligibles] enrollment.” Murray said ACAP also would like estimates of how star ratings would change for contracts under each approach and the “strengths and weaknesses” of each. And the letter asked for clarification on whether plans would need a minimum number of LIS/ dual or disabled enrollees in order to receive an adjustment via either approach. Of ACAP’s 61 “safety net” plan members, 15 are in the ongoing CMS-backed demonstration for duals. On the CAI, Murray wrote, ACAP asks to clarify whether institutionalized beneficiaries would be included in the adjustment. And on the IS option CMS should “clarify to what extent, if at all, the subset of adjusted measures would change year-to-year” and whether insurers “should expect both upside and downside adjustment” since in a few stars measures plans with high proportions of disabled duals outperform other plans, she said. Consulting firm Avalere Health LLC prepared an analysis of the two options as part of a stars-changes webinar Dec. 8, 2015. In this webinar, Christie Teigland, Ph.D., an Avalere CAI and Indirect Standardization Differ In Approach Across Key Parameters Method 1: Categorical Method 2: Indirect Adjustment Index (CAI) Standardization Star Measures Adjusted Subset of 16 HEDIS/ HOS/PDE measures Subset of 16 HEDIS/ HOS/PDE measures Data Measurement Year Current year or prior year Current year Adjustment Level Adjusts based on mean difference in Star Rating (adjusted rating vs. observed rating) for groups of contracts Adjusts measure scores at contract level Method Linear regression to calculate impact of dual eligible/disabled on measure score Logistic regression to calculate impact of dual eligible/disabled on measure score Adjustment Groups of contracts with similar levels of duals/ disabled Contract Note: HEDIS=Healthcare Effectiveness Data and Information Set; HOS=Health Outcomes Survey; PDE=Prescription Drug Event. SOURCE: © 2015 Avalere Health/An Inovalon Company; used in Avalere webinar “Medicare Advantage Star Ratings: Sifting Through Proposed Changes and What They Mean,” Dec. 8, 2015. The Challenges of Pharmacy Star Ratings • 45 vice president, said that more information is needed even after CMS’s Dec. 3, 2015, call on the options, especially on how the average national rates the agency would use in the options are calculated. But she added that for many of the stars measures, any adjustments would be “pretty small” under either option. This is especially true since the 16 star measures that could be adjusted account for only 27% of potential stars scores, according to Teigland. All of those points figured prominently in the comments about the proposals submitted by two other associations, America’s Health Insurance Plans (AHIP) and the SNP Alliance, which represents MA Special Needs Plans. The lengthy SNP Alliance comments centered on “four principal concerns” it had about the proposed options. The first one, the alliance wrote, is that “both options may only minimally account for underlying disparities in star performance for plans specializing in care of poor Medicare beneficiaries.” It based that conclusion on research showing that adjusting for just dual and/or LIS status “does not capture the full amount of within-contract differences in stars performance.” Moreover, according to the alliance, “The number of measures targeted for inclusion in both models represents a subset of measures that research suggests are impacted by SES. We strongly urge CMS to apply the interim adjustment to all measures in the star rating system related to patient care” while excluding measures such as customer service and appeals rates “directly related to plan activity.” Alliance Fears Interim Policy Could Stay The alliance also asserted that “neither option incorporates many of the factors shown in the research to be of greatest importance in accounting for social determinants of health” such as income, education, occupation and social supports. The third concern was that CMS has not furnished sufficient information for plans to understand and fully evaluate the feasibility and impact of the options despite the details given in the agency’s Request for Comments (RFC) and call. Finally, and perhaps underlying its other three reservations, the alliance said it “is concerned that the interim proposal could easily become long-term policy without a more aggressive approach to address social determinants of health.” Its recommendations, based on these reservations, included that “CMS should risk adjust quality measures in the star rating system for beneficiaries’ socio-demographic characteristics beginning in 2018.” For interim relief, it suggested that CMS maximize the number of measures selected for adjustment, with “at a minimum” using all 12 measures showing “a negative performance gap for dual eligible members, as well as consideration of neighborhood poverty and physician shortage factors.” AHIP’s comments, sent by Mark Hamelburg, senior vice president, federal programs, begin by calling the proposals “an important step forward.” But AHIP too says it is being hampered in commenting by the lack of “detailed analyses of the potential 46 • AIS’s Management Insight Series impact of each approach.” While CMS has said it would be furnishing more details on both approaches when it releases the draft 2017 MA pay notice and Call Letter in February 2016, Hamelburg noted, “we strongly urge the agency to release this information as far in advance of that time as possible,” along with estimates of the “total impact of each approach on the program as a whole.” The comments go on to raise concerns about the time factors, pointing out CMS said in the Dec. 3, 2015, call that the options would require additional data-processing steps, resulting in a “compressed timeframe” for MA and Part D plans’ preview of their star ratings. But AHIP’s strongest comments related to the potential impact on “high performing plans.” Citing comments in both the RFC and Dec. 3,2 015, call that the proposed adjustments could reduce star ratings for some plans, Hamelburg wrote that “we strongly disagree with this approach. No contract should be penalized if the adjusted Part C or D Summary Rating or Overall Rating calculated under the adopted approach would be lower than the unadjusted rating.” The Social Security Act “does not require changes to the star ratings system to be budget neutral,” he added. Those comments may have stemmed from responses made in the CMS call by Elizabeth Goldstein, Ph.D., director of consumer assessment and plan performance in the agency’s Medicare Drug Benefit and C&D Data Group. She said that the adjustments could be either positive or negative and that CMS in any interim relief would be able to make adjustments only for those factors it has sufficient data on. This raises a question articulated by Avalere Vice President and former CMS official Tom Kornfield during that firm’s webinar. “The question is whether it’s better to do nothing than to do this,” he asserted, and to wait for the permanent relief CMS may be able to offer in 2018 after it gets more data. Avalere’s sense of the options outlined in the RFC, he added, is “this doesn’t really solve the problem.” The Challenges of Pharmacy Star Ratings • 47 Impact on Performance Contract Enrollment for Plans Whose Star Ratings Increased in 2016 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment BLUE CROSS AND BLUE SHIELD OF ALABAMA 3.5 4 49,576 53,395 7.70% HEALTHSPRING OF ALABAMA, INC. 3.5 4 52,553 59,025 12.32% Contract Number Contract Name H0104 H0150 H0151 H0154 H0251 UNITEDHEALTHCARE PLAN OF THE RIVER VALLEY, INC. H0354 UNITEDHEALTHCARE OF ALABAMA, INC. % Change 3 3.5 33,725 35,179 4.31% 3.5 4 45,419 43,315 -4.63% 3 3.5 41,024 43,540 6.13% CIGNA HEALTHCARE OF ARIZONA, INC. 4.5 5 43,215 52,816 22.22% H0602 ROCKY MOUNTAIN HEALTH MAINTENANCE ORGANIZATION 3.5 4 24,343 24,211 -0.54% H0609 H0624 H0710 H0755 H0908 H0913 H1111 H1304 H1517 H1659 H1849 PACIFICARE OF COLORADO, INC 4 4.5 100,105 240,691 140.44% UNITEDHEALTHCARE INSURANCE COMPANY 3 3.5 3,647 4,850 32.99% 439.41% VIVA HEALTH, INC. 4 4.5 2,593 13,987 3.5 4 37,587 42,537 13.17% 2 3 1,025 885 -13.66% 2.5 3 1,044 912 -12.64% 3 3.5 8,490 9,624 13.36% 3.5 4 7,914 5,526 -30.17% 3 3.5 3,358 2,705 -19.45% 2.5 3 4,570 5,151 12.71% ANTHEM HEALTH PLANS OF KENTUCKY, INC. 3 3.5 6,085 7,143 17.39% H1951 HUMANA HEALTH BENEFIT PLAN OF LOUISIANA, INC. 4 4.5 124,570 134,912 8.30% H2001 SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. 4 4.5 611,215 726,013 18.78% H2029 HUMANA INSURANCE OF PUERTO RICO, INC. 3 3.5 1,833 1,141 -37.75% H2165 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. 3 3.5 6,310 6,985 10.70% H2224 H2226 SENIOR WHOLE HEALTH, LLC 3.5 4 10,146 10,642 4.89% UNITEDHEALTHCARE INSURANCE COMPANY 3.5 4.5 12,775 14,302 11.95% H2256 TUFTS ASSOCIATED HEALTH MAINTENANCE ORGANIZATION 4.5 5 104,955 105,966 0.96% H2354 HEALTHPLUS OF MICHIGAN 2.5 4.5 20,411 17,514 -14.19% H2416 PRIMEWEST CTRL COUNTY-BASED PURCHASING INITIATIVE 3.5 4 1,952 1,937 -0.77% H2419 H2425 H2462 H2610 UNITEDHEALTHCARE INSURANCE COMPANY OXFORD HEALTH PLANS (CT), INC. BUCKEYE COMMUNITY HEALTH PLAN, INC. WELLCARE HEALTH PLANS OF NEW JERSEY, INC. UNITEDHEALTHCARE OF GEORGIA, INC. REGENCE BLUE SHIELD OF IDAHO ANTHEM INSURANCE COMPANIES, INC. PIEDMONT COMMUNITY HEALTHCARE, INC. 4 4.5 1,554 1,581 1.74% BLUE PLUS SOUTH COUNTRY HEALTH ALLIANCE 3.5 4 8,049 7,836 -2.65% GROUP HEALTH PLAN, INC. (MN) 4.5 5 48,841 50,104 2.59% ESSENCE HEALTHCARE, INC. 4.5 5 51,896 58,906 13.51% H2775 AMERICAN PROGRESSIVE LIFE & HLTH INS COMPANY OF NY 3.5 4 10,913 12,789 17.19% H2802 UNITEDHEALTHCARE OF THE MIDLANDS, INC. 3.5 4.5 9,980 29,324 193.83% H2905 SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. 2.5 3 3,513 3,415 -2.79% H2906 H2931 H3107 HOMETOWN HEALTH PLAN, INC. 3.5 4 3,932 3,309 -15.84% HEALTH PLAN OF NEVADA, INC. 3.5 4 50,917 50,765 -0.30% 3 3.5 75,720 77,144 1.88% continued OXFORD HEALTH PLANS (NJ), INC. 48 • AIS’s Management Insight Series Contract Enrollment for Plans Whose Star Ratings Increased in 2016 (continued) Contract Number Contract Name H3154 H3204 H3240 H3307 HORIZON HEALTHCARE OF NEW JERSEY, INC. 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change 3 3.5 25,891 23,684 -8.52% 3.5 4.5 33,047 36,629 10.84% AMERIGROUP NEW JERSEY, INC. 3 3.5 6,897 7,606 10.28% OXFORD HEALTH PLANS (NY), INC. 3 3.5 72,471 81,186 12.03% H3328 THE NEW YORK STATE CATHOLIC HEALTH PLAN, INC. 3.5 4 35,800 54,272 51.60% H3337 H3344 H3379 H3421 H3447 H3597 H3668 H3749 H3814 H3817 H3818 H3907 H3957 H4003 H4004 H4005 H4011 H4279 H4407 H4454 H4461 H4525 H4527 H4590 H4604 H4605 H5008 H5009 H5010 H5106 H5294 H5425 LIBERTY HEALTH ADVANTAGE, INC. PRESBYTERIAN HEALTH PLAN 2.5 3 4,985 5,972 19.80% INDEPENDENT HEALTH BENEFITS CORPORATION 4 4.5 3,806 3,903 2.55% UNITEDHEALTHCARE OF NEW YORK, INC. 3 3.5 27,443 29,357 6.97% AMERICA'S 1ST CHOICE HEALTH PLANS, INC. 3 3.5 3,759 2,960 -21.26% 42.16% HEALTHKEEPERS, INC. 3.5 4.5 2,538 3,608 AETNA HEALTH, INC. (ME) 4 4.5 7,186 6,507 -9.45% MOUNT CARMEL HEALTH PLAN, INC. 4 4.5 49,121 54,464 10.88% UNITEDHEALTHCARE OF OKLAHOMA, INC. 3 3.5 28,003 26,895 -3.96% ATRIO HEALTH PLANS 3 3.5 2,122 2,206 3.96% REGENCE BLUECROSS BLUESHIELD OF OREGON 3.5 4.5 64,292 53,688 -16.49% FAMILYCARE HEALTH PLANS, INC. 3 3.5 3,388 4,370 28.98% UPMC HEALTH PLAN, INC. 4 4.5 114,746 124,359 8.38% KEYSTONE HEALTH PLAN WEST, INC. 4 4.5 117,257 118,680 1.21% MMM HEALTHCARE, LLC. 3.5 4 170,721 168,620 -1.23% PMC MEDICARE CHOICE, LLC 3.5 4 30,091 23,234 -22.79% -7.53% TRIPLE-S ADVANTAGE, INC. 3 3.5 17,778 16,440 MMM HEALTHCARE, LLC. 2.5 3 10,530 * NA UPMC FOR YOU, INC 3.5 4 18,922 21,040 11.19% HEALTHSPRING OF TENNESSEE, INC. 3.5 4 8,317 11,416 37.26% HEALTHSPRING OF TENNESSEE, INC. 4 4.5 91,134 96,142 5.50% CARITEN HEALTH PLAN INC. 4 4.5 106,407 108,656 2.11% SHA, L.L.C 2.5 3.5 3,254 3,145 -3.35% PHYSICIANS HEALTH CHOICE OF TEXAS LLC 3.5 4 27,426 29,213 6.52% UNITEDHEALTHCARE BENEFITS OF TEXAS, INC. 3.5 4 200,330 209,620 4.64% UNITEDHEALTHCARE OF UTAH, INC. 3.5 4.5 46,349 53,594 15.63% REGENCE BLUECROSS BLUESHIELD OF UTAH 3.5 4.5 12,370 10,917 -11.75% 3 3.5 9,059 12,803 41.33% UNITEDHEALTHCARE INSURANCE COMPANY REGENCE BLUESHIELD 3.5 4 28,908 21,404 -25.96% ASURIS NORTHWEST HEALTH 3.5 4.5 1,389 1,349 -2.88% HIGHMARK SENIOR SOLUTIONS COMPANY 3.5 4 7,072 7,052 -0.28% SUPERIOR HEALTH PLAN, INC. 2.5 3 2,275 779 -65.76% SCAN HEALTH PLAN 4 4.5 157,341 167,513 6.46% H5430 ONECARE BY CARE1ST HEALTH PLAN ARIZONA INC. 3 3.5 1,695 1,838 8.44% H5431 H5439 H5521 H5528 H5533 HEALTHSUN HEALTH PLANS, INC. 4 4.5 27,529 32,429 17.80% HEALTH NET LIFE INSURANCE COMPANY 3 3.5 1,261 669 -46.95% AETNA LIFE INSURANCE COMPANY 4 4.5 582,102 680,813 16.96% GROUP HEALTH INCORPORATED 2.5 3 7,729 7,102 -8.11% UPMC HEALTH NETWORK, INC. 3.5 4 9,451 8,839 -6.48% The Challenges of Pharmacy Star Ratings • 49 Contract Enrollment for Plans Whose Star Ratings Increased in 2016 (continued) 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change VANTAGE HEALTH PLAN, INC. 3.5 4 12,951 15,948 23.14% H5652 SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. 4.5 5 4,613 4,763 3.25% H5746 AMERIGROUP COMMUNITY CARE OF NEW MEXICO, INC. 2.5 3.5 2,388 1,954 -18.17% H5817 H5823 H5926 H5932 H5969 H5995 AMERIGROUP TEXAS, INC. 3 4 34,165 29,875 -12.56% MOLINA HEALTHCARE OF WASHINGTON, INC. 3 3.5 8,853 9,753 10.17% MOLINA HEALTHCARE OF MICHIGAN, INC. 3 3.5 12,771 10,068 -21.17% Contract Number Contract Name H5576 GATEWAY HEALTH PLAN, INC. 2.5 3.5 48,761 50,792 4.17% ALOHACARE 2.5 3.5 458 742 62.01% 3 3.5 3,129 3,087 -1.34% 2.5 3 25,419 25,709 1.14% 36.74% ATRIO HEALTH PLANS H6528 CARE IMPROVEMENT PLUS SOUTH CENTRAL INSURANCE CO. H6815 H7187 H7678 H7917 H8189 H8748 H9104 HEALTH NET HEALTH PLAN OF OREGON 4 4.5 23,306 31,869 UNITEDHEALTHCARE INSURANCE COMPANY 3 3.5 10,114 10,949 8.26% MOLINA HEALTHCARE OF TEXAS, INC. 3 3.5 5,969 3,786 -36.57% SCAN HEALTH PLAN R6801 CARE IMPROVEMENT PLUS OF TEXAS INSURANCE COMPANY R7444 UNITEDHEALTHCARE INSURANCE COMPANY BLUECROSS BLUESHIELD OF TENNESSEE, INC. 3.5 4 89,791 93,045 3.62% MANAGED HEALTH SERVICES, WISCONSIN 2.5 3.5 1,040 988 -5.00% 3 3.5 2,480 3,396 36.94% 3.5 4 1,415 219 -84.52% 2.5 3 71,426 78,411 9.78% 3.5 4 20,443 20,847 1.98% UNITEDHEALTHCARE INSURANCE COMPANY NOTE: Does not include new or terminated plans. SOURCE: Calculated by AIS from CMS data for AIS's Medicare and Medicaid Market Data: 2016. Contract Enrollment for Plans Whose Star Ratings Decreased in 2016 Contract Number H0423 H0523 H0544 H0545 H0838 H1019 H1035 H1036 H1045 H1099 H1112 H1350 H1415 Contract Name 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change METROPLUS HEALTH PLAN, INC. 3.5 3 8,571 8,500 -0.83% AETNA HEALTH OF CALIFORNIA INC. 3.5 3 16,778 16,078 -4.17% CAREMORE HEALTH PLAN 4.5 4 50,997 51,971 1.91% INTER VALLEY HEALTH PLAN, INC. 4 3.5 22,250 23,089 3.77% UNIVERSAL CARE, INC. 3.5 3 7,156 10,151 41.85% CAREPLUS HEALTH PLANS, INC. 5 4.5 105,623 110,197 4.33% FLORIDA HEALTH CARE PLAN, INC. 4.5 4 13,205 14,256 7.96% HUMANA MEDICAL PLAN, INC. 4.5 4 424,214 453,064 6.80% PREFERRED CARE PARTNERS, INC. 4 3.5 54,138 118,863 119.56% HEALTH FIRST HEALTH PLANS 4 3.5 29,695 34,429 15.94% WELLCARE OF GEORGIA, INC. 3 2.5 32,202 36,396 13.02% BLUE CROSS OF IDAHO CARE PLUS, INC. 4.5 4 22,325 19,777 -11.41% HEALTHSPRING OF TENNESSEE, INC. 3.5 3 17,742 19,718 11.14% continued 50 • AIS’s Management Insight Series Contract Enrollment for Plans Whose Star Ratings Decreased in 2016 (continued) Contract Name 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change H1417 H1418 H1716 H1994 H2225 H2450 H2486 H2654 H2663 H2667 H3152 H3156 H3251 H3342 H3384 H3660 H3672 H3813 H3815 H3822 H3864 H3909 H3959 H4523 H4875 H5050 H5302 H5414 H5522 H5594 H5698 H6328 H6360 H6609 H6743 H7149 H7220 HEALTH ALLIANCE CONNECT, INC. 4.5 4 8,584 9,977 16.23% HUMANA INSURANCE COMPANY 4 3.5 12,221 11,086 -9.29% HUMANA INSURANCE COMPANY 4 3.5 13,293 11,303 -14.97% SELECTHEALTH, INC. 4.5 3.5 31,489 37,030 17.60% COMMONWEALTH CARE ALLIANCE, INC. 4.5 4 6,321 6,924 9.54% MEDICA INSURANCE COMPANY 4.5 4 169,498 186,103 9.80% HUMANA MEDICAL PLAN OF UTAH, INC. 4 3.5 5,665 5,309 -6.28% UNITEDHEALTHCARE OF THE MIDWEST, INC. 4 3.5 47,003 78,061 66.08% COVENTRY HEALTH CARE OF MISSOURI, INC 4.5 4 56,787 49,936 -12.06% COVENTRY HEALTH CARE OF MISSOURI, INC 4.5 4 19,534 22,467 15.01% AETNA HEALTH, INC. (NJ) 4 3.5 37,432 39,800 6.33% AMERIHEALTH HMO, INC. 3.5 3 18,643 19,120 2.56% HEALTH CARE SERVICE CORPORATION 4 3 18,712 16,681 -10.85% EMPIRE HEALTHCHOICE ASSURANCE, INC. 3.5 3 44,327 6,850 -84.55% HEALTHNOW NEW YORK INC. 4.5 4 23,907 23,396 -2.14% SUMMACARE INC. 4 3.5 27,126 24,273 -10.52% HEALTH PLAN OF THE UPPER OHIO VALLEY 4 3.5 5,218 13,111 151.26% MODA HEALTH PLAN, INC. 4 3.5 12,192 15,456 26.77% ALIGNMENT HEALTH PLAN 3.5 3 19,038 19,329 1.53% HEALTH CARE SERVICE CORPORATION 4 3 23,137 28,117 21.52% PACIFICSOURCE COMMUNITY HEALTH PLANS 4 3.5 24,168 27,342 13.13% QCC INSURANCE COMPANY 4.5 4 8,800 9,407 6.90% HEALTHAMERICA PENNSYLVANIA, INC. 4.5 4 69,591 63,262 -9.09% AETNA HEALTH, INC. (TX) 3.5 3 13,099 23,533 79.65% PRIORITY HEALTH 4.5 4 17,224 24,278 40.95% GROUP HEALTH COOPERATIVE 5 4.5 82,410 89,561 8.68% AETNA HEALTH, INC. (GEORGIA) 3.5 3 1,357 256 -81.13% AETNA HEALTH, INC. (FL) 4 3.5 18,299 66,803 265.06% HEALTHASSURANCE PENNSYLVANIA, INC. 4.5 4 72,300 69,659 -3.65% OPTIMUM HEALTHCARE, INC. 4.5 4 38,941 43,950 12.86% WINDSOR HEALTH PLAN, INC. 3 2.5 38,471 42,312 9.98% CARE N' CARE INSURANCE COMPANY, INC. 4 3.5 9,846 10,550 7.15% HEALTHSPAN INTEGRATED CARE 5 4 13,656 11,445 -16.19% HUMANA INSURANCE COMPANY 4.5 4 785,863 665,282 -15.34% ATRIO HEALTH PLANS 3.5 3 6,980 8,681 24.37% COVENTRY HEALTH CARE OF NEBRASKA, INC. 4 3.5 6,442 8,053 25.01% INDIANA UNIVERSITY HEALTH PLANS, INC. 4.5 4 11,862 16,361 37.93% H7787 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. 3.5 3 5,035 6,664 32.35% H8578 H8953 H9047 H9302 H9859 HEALTH NEW ENGLAND, INC. 4.5 4 8,577 9,143 6.60% HUMANA HEALTH PLAN OF OHIO, INC. 4 3.5 40,289 47,470 17.82% PROVIDENCE HEALTH PLAN 5 4.5 46,865 49,783 6.23% SOUNDPATH HEALTH 3.5 3 20,772 26,405 27.12% MVP HEALTH PLAN, INC. 4.5 4 11,515 13,042 13.26% Contract Number The Challenges of Pharmacy Star Ratings • 51 Contract Enrollment for Plans Whose Star Ratings Decreased in 2016 (continued) 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change R3332 BLUE CROSS AND BLUE SHIELD OF FLORIDA, INC. 3.5 3 49,517 46,811 -5.46% R5826 HUMANA INSURANCE COMPANY 4 3.5 499,277 523,542 4.86% Contract Number Contract Name NOTE: Does not include new or terminated plans. SOURCE: Calculated by AIS from CMS data for AIS's Medicare and Medicaid Market Data: 2016. Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change 4 4 10,065 14,866 47.70% 3.5 3.5 6,865 6,978 1.65% MEDISUN, INC. 4 4 45,791 60,953 33.11% ARIZONA PHYSICIANS IPA, INC. 3 3 35,940 38,788 7.92% KS PLAN ADMINISTRATORS, LLC 4.5 4.5 29,432 30,304 2.96% HEALTH NET OF ARIZONA, INC. Contract Number Contract Name H0028 CHA HMO, INC. H0294 CARE IMPROVEMENT PLUS WISCONSIN INSURANCE COMPANY H0302 H0321 H0332 H0351 H0504 H0524 H0543 H0562 H0564 H0571 H0630 H0712 H1016 H1026 H1032 H1109 H1170 H1230 H1264 H1286 H1302 H1365 H1406 H1416 H1463 H1468 H1510 H1607 3.5 3.5 38,633 26,138 -32.34% CALIFORNIA PHYSICIANS' SERVICE 4 4 101,927 94,458 -7.33% KAISER FOUNDATION HP, INC. 5 5 1,010,165 1,057,312 4.67% UHC OF CALIFORNIA 4 4 328,182 335,288 2.17% HEALTH NET OF CALIFORNIA,INC. 4 4 164,571 142,976 -13.12% BLUE CROSS OF CALIFORNIA 3.5 3.5 12,558 20,371 62.22% CHINESE COMMUNITY HEALTH PLAN 2.5 2.5 8,542 8,553 0.13% KAISER FOUNDATION HP OF CO 5 5 95,800 101,441 5.89% WELLCARE OF CONNECTICUT, INC. 3 3 12,516 7,240 -42.15% -1.16% AVMED, INC. 4 4 31,054 30,694 HEALTH OPTIONS, INC. 3.5 3.5 60,439 61,232 1.31% WELLCARE OF FLORIDA, INC. 3.5 3.5 106,324 92,088 -13.39% AETNA HEALTH INC.(GEORGIA) 3.5 3.5 3,547 3,855 8.68% KAISER FOUNDATION HP OF GA, INC. 4.5 4.5 24,362 26,026 6.83% KAISER FOUNDATION HP, INC. 5 5 30,764 32,038 4.14% WELLCARE OF TEXAS, INC. 3 3 31,597 33,442 5.84% UNITEDHEALTHCARE INSURANCE COMPANY 3.5 3.5 13,608 14,966 9.98% BLUE CROSS OF IDAHO CARE PLUS, INC. 3.5 3.5 10,400 7,322 -29.60% MARTIN'S POINT GENERATIONS, LLC 4.5 4.5 2,499 2,305 -7.76% 4 4 48,592 54,776 12.73% HUMANA HEALTH PLAN, INC. HARMONY HEALTH PLAN OF ILLINOIS, INC. 3 3 14,982 14,735 -1.65% HEALTH ALLIANCE CONNECT, INC. 4.5 4.5 7,604 8,538 12.28% HUMANA BENEFIT PLAN OF ILLINOIS, INC. 4.5 4.5 6,266 7,069 12.82% 4 4 41,890 45,804 9.34% 3.5 3.5 8,457 1,719 -79.67% continued HUMANA INSURANCE COMPANY ANTHEM INSURANCE COMPANIES, INC. 52 • AIS’s Management Insight Series Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016 (continued) 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change H1608 COVENTRY HEALTH AND LIFE INSURANCE COMPANY 4.5 4.5 30,181 131,953 337.21% H1609 H1903 H1944 H1961 H2012 H2108 H2150 H2174 H2228 H2230 H2237 H2261 H2312 H2320 AETNA HEALTH, INC. 4.5 4.5 3,746 4,399 17.43% WELLCARE OF LOUISIANA, INC. 2.5 2.5 9,264 10,098 9.00% UNITEDHEALTHCARE OF NEW ENGLAND, INC. 4 4 7,314 32,040 338.06% PEOPLES HEALTH, INC. 4 4 54,106 54,690 1.08% HUMANA HEALTH PLAN, INC. 4 4 194,553 201,824 3.74% BRAVO HEALTH MID-ATLANTIC, INC. 3 3 20,907 21,517 2.92% KAISER FNDN HP OF THE MID-ATLANTIC STS 5 5 61,772 66,254 7.26% TRILLIUM COMMUNITY HEALTH PLAN 3 3 3,892 3,385 -13.03% UNITEDHEALTHCARE INSURANCE COMPANY 4.5 4.5 1,729 184,462 10568.71% BCBS OF MASSACHUSETTS HMO BLUE, INC. 4.5 4.5 28,097 32,536 15.80% INDEPENDENT CARE HEALTH PLAN, INC. 3.5 3.5 5,932 6,332 6.74% BCBS OF MASSACHUSETTS HMO BLUE, INC. 4.5 4.5 10,757 10,488 -2.50% 4 4 43,289 43,844 1.28% PRIORITY HEALTH 4.5 4.5 80,629 85,770 6.38% H2322 ALLIANCE HEALTH AND LIFE INSURANCE COMPANY 3.5 3.5 3,258 3,328 2.15% H2406 SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. 4 4 947 1,926 103.38% H2422 H2456 H2458 H2459 HEALTHPARTNERS, INC. 4.5 4.5 3,180 3,153 -0.85% UCARE MINNESOTA 3.5 3.5 10,280 10,697 4.06% 4 4 10,039 10,385 3.45% 4.5 4.5 85,475 81,542 -4.60% H2491 WELLCARE HEALTH INSURANCE OF ARIZONA, INC. 3 3 10,742 8,111 -24.49% H2593 H2649 H2672 H2701 CAREMORE HEALTH PLAN OF ARIZONA, INC. 3.5 3.5 16,093 11,777 -26.82% 4 4 134,166 220,746 64.53% COVENTRY HEALTH CARE OF KANSAS, INC. 3.5 3.5 17,780 17,925 0.82% NEW WEST HEALTH SERVICES 3.5 3.5 17,960 14,370 -19.99% H2816 AMERICAN PROGRESSIVE LIFE & HLTH INS COMPANY OF NY 4 4 27,236 32,413 19.01% H2836 H2944 H2949 H2960 H3206 H3305 H3312 ANTHEM HEALTH PLANS, INC. 3.5 3.5 1,293 * NA HUMANA INSURANCE COMPANY 3.5 3.5 19,217 13,344 -30.56% HUMANA HEALTH PLAN, INC. 4 4 41,789 44,077 5.48% HOMETOWN HEALTH PLAN, INC. 4 4 13,652 13,935 2.07% PRESBYTERIAN INSURANCE COMPANY, INC. 3.5 3.5 8,630 8,092 -6.23% MVP HEALTH PLAN, INC. 4.5 4.5 35,992 30,526 -15.19% AETNA HEALTH, INC. (NY) 3.5 3.5 16,606 15,725 -5.31% H3330 HEALTH INSURANCE PLAN OF GREATER NEW YORK 3 3 116,037 113,830 -1.90% H3335 H3347 H3351 H3359 H3361 EXCELLUS HEALTH PLAN, INC. 4 4 45,938 46,223 0.62% ELDERPLAN, INC. 3 3 14,405 12,018 -16.57% EXCELLUS HEALTH PLAN, INC. 4 4 71,650 69,173 -3.46% MANAGED HEALTH, INC. 4 4 127,910 134,043 4.79% WELLCARE OF NEW YORK, INC. 3 3 45,514 41,481 -8.86% Contract Number Contract Name HEALTH ALLIANCE PLAN OF MICHIGAN MEDICA HEALTH PLANS UCARE MINNESOTA HUMANA HEALTH PLAN, INC. The Challenges of Pharmacy Star Ratings • 53 Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016 (continued) Contract Number Contract Name 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change 0.00% H3362 H3370 H3387 INDEPENDENT HEALTH ASSOCIATION, INC. 4.5 4.5 89,792 89,791 EMPIRE HEALTHCHOICE HMO, INC. 3.5 3.5 63,458 67,499 6.37% UNITEDHEALTHCARE OF NEW YORK, INC. 3.5 3.5 13,499 26,328 95.04% H3388 CAPITAL DISTRICT PHYSICIANS' HEALTH PLAN, INC. 4.5 4.5 39,609 37,072 -6.41% H3404 BLUE CROSS AND BLUE SHIELD OF NORTH CAROLINA 3.5 3.5 50,227 37,655 -25.03% H3449 BLUE CROSS AND BLUE SHIELD OF NORTH CAROLINA 3.5 3.5 76,525 69,818 -8.76% H3528 H3653 H3655 H3664 H3706 H3755 H3805 H3810 H3811 H3832 H3916 H3923 CONNECTICARE, INC. 4 4 60,674 56,240 -7.31% PARAMOUNT CARE, INC. 4 4 15,074 16,177 7.32% COMMUNITY INSURANCE COMPANY 3.5 3.5 125,615 134,124 6.77% AULTCARE HEALTH INSURING CORPORATION 4.5 4.5 20,108 20,989 4.38% GENERATIONS HEALTHCARE, INC. 3.5 3.5 5,697 6,677 17.20% COMMUNITY CARE HMO, INC 4 4 29,153 29,128 -0.09% UNITEDHEALTHCARE OF OREGON, INC. 4 4 24,596 92,680 276.81% H3924 GEISINGER INDEMNITY INSURANCE COMPANY H3931 H3949 H3952 H3954 H3962 AETNA HEALTH, INC, (PA) H4007 HUMANA HEALTH PLANS OF PUERTO RICO, INC. H4036 ALLCARE HEALTH PLAN, INC. 4 4 2,129 1,981 -6.95% 3.5 3.5 4,979 5,217 4.78% 4 4 37,076 37,147 0.19% HIGHMARK SENIOR HEALTH COMPANY 4.5 4.5 175,104 191,920 9.60% CAPITAL ADVANTAGE INSURANCE COMPANY 3.5 3.5 14,618 14,221 -2.72% 4 4 7,062 15,064 113.31% SAMARITAN HEALTH PLANS, INC. HAWAII MEDICAL SERVICE ASSOCIATION 4 4 28,453 53,521 88.10% BRAVO HEALTH PENNSYLVANIA, INC. 3.5 3.5 57,247 53,205 -7.06% KEYSTONE HEALTH PLAN EAST, INC. 4.5 4.5 86,438 90,584 4.80% GEISINGER HEALTH PLAN 4.5 4.5 68,189 71,070 4.23% 4 4 10,872 10,831 -0.38% 3.5 3.5 49,400 37,808 -23.47% ANTHEM INSURANCE COMPANIES, INC. 4 4 1,457 26,338 1707.69% H4141 HUMANA EMPLOYERS HEALTH PLAN OF GEORGIA, INC. 3.5 3.5 36,483 43,665 19.69% H4152 BLUE CROSS & BLUE SHIELD OF RHODE ISLAND 4 4 51,317 52,628 2.55% H4213 H4346 H4506 USABLE MUTUAL INSURANCE COMPANY KEYSTONE HEALTH PLAN CENTRAL, INC. 3 3 17,380 16,581 -4.60% 3.5 3.5 6,057 6,697 10.57% SELECTCARE OF TEXAS, INC. 4 4 62,874 65,531 4.23% H4513 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. 4 4 87,615 99,250 13.28% H4514 UNITEDHEALTHCARE COMMUNITY PLAN OF TEXAS, LLC 3 3 29,080 29,045 -0.12% H4528 HEALTHSPRING LIFE & HEALTH INSURANCE COMPANY, INC. 3 3 8,769 6,650 -24.16% H4564 H4754 H4866 H4909 SCOTT AND WHITE HEALTH PLAN CAREMORE HEALTH PLAN OF NEVADA 4 4 27,879 28,993 4.00% PACIFICSOURCE COMMUNITY HEALTH PLANS 3.5 3.5 9,190 10,014 8.97% CUATRO LLC 2.5 2.5 3,458 4,333 25.30% 4 4 2,472 864 -65.05% ANTHEM HEALTH PLANS OF VIRGINIA, INC. continued 54 • AIS’s Management Insight Series Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016 (continued) Contract Number Contract Name 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change -51.55% H5042 H5087 H5209 H5211 H5215 H5216 H5253 H5262 H5410 CDPHP UNIVERSAL BENEFITS, INC. 4.5 4.5 7,345 3,559 EASY CHOICE HEALTH PLAN INC. 3.5 3.5 33,568 32,117 -4.32% 4 4 1,002 1,252 24.95% SECURITY HEALTH PLAN OF WISCONSIN, INC. 4.5 4.5 43,053 43,518 1.08% NETWORK HEALTH INSURANCE CORPORATION 4.5 4.5 63,003 62,406 -0.95% HUMANA INSURANCE COMPANY 4.5 4.5 33,771 31,622 -6.36% UNITEDHEALTHCARE OF WISCONSIN, INC. 4.5 4.5 93,969 326,452 247.40% H5415 HUMANA HEALTH INSURANCE COMPANY OF FLORIDA, INC. H5420 MEDICA HEALTHCARE PLANS, INC. H5422 BLUE CROSS BLUE SHIELD HEALTHCARE PLAN OF GEORGIA H5427 H5428 H5433 FREEDOM HEALTH, INC. CARE WISCONSIN HEALTH PLAN, INC. GUNDERSEN HEALTH PLAN 5 5 14,237 14,455 1.53% HEALTHSPRING OF FLORIDA 4.5 4.5 50,650 52,088 2.84% 4 4 14,932 11,126 -25.49% 3.5 3.5 35,276 37,580 6.53% 3 3 1,489 1,850 24.24% 4 4 62,920 69,386 10.28% SAN MATEO HEALTH COMMISSION 3.5 3.5 817 783 -4.16% ORANGE COUNTY HEALTH AUTHORITY 3.5 3.5 12,818 1,294 -89.90% H5434 BLUE CROSS AND BLUE SHIELD OF FLORIDA, INC. 3.5 3.5 11,002 10,487 -4.68% H5435 H5471 H5520 H5525 H5526 H5530 H5549 H5577 UNITEDHEALTHCARE INSURANCE COMPANY H5580 SOUTHWEST CATHOLIC HEALTH NETWORK CORPORATION H5587 H5590 H5591 H5608 H5609 H5619 H5628 H5649 H5656 H5774 H5793 H5810 H5826 H5859 H5883 H5928 HEALTH CHOICE ARIZONA, INC. 3.5 3.5 47,299 39,504 -16.48% SIMPLY HEALTHCARE PLANS, INC. 4 4 20,989 31,382 49.52% HEALTH NET LIFE INSURANCE COMPANY 4 4 39,115 37,902 -3.10% HUMANA BENEFIT PLAN OF ILLINOIS, INC. 4 4 55,130 73,844 33.95% HEALTHNOW NEW YORK INC. 4.5 4.5 18,793 17,933 -4.58% ANTHEM HEALTH PLANS OF KENTUCKY, INC. 3.5 3.5 4,841 3,795 -21.61% VNS CHOICE MCS ADVANTAGE, INC. BRIDGEWAY HEALTH SOLUTIONS MARTIN'S POINT GENERATIONS, LLC 3 3 19,030 21,522 13.10% 3.5 3.5 172,112 197,570 14.79% 3 3 17,544 18,442 5.12% 3 3 9,470 9,675 2.16% 3.5 3.5 1,485 1,575 6.06% 14.09% 5 5 30,536 34,837 DENVER HEALTH MEDICAL PLAN, INC. 3.5 3.5 4,293 4,399 2.47% GEMCARE HEALTH PLAN INC. 3.5 3.5 8,161 7,508 -8.00% 4 4 7,218 75,847 950.80% MOLINA HEALTHCARE OF UTAH, INC. 3.5 3.5 8,649 8,498 -1.75% CENTRAL HEALTH PLAN OF CALIFORNIA, INC. 3.5 3.5 27,243 32,258 18.41% SELECTCARE HEALTH PLANS, INC. 3 3 3,946 3,139 -20.45% TRIPLE S ADVANTAGE, INC. 3 3 98,686 104,844 6.24% AETNA HEALTH, INC. (CT) 4 4 21,703 24,440 12.61% MOLINA HEALTHCARE OF CALIFORNIA 3 3 3,909 4,248 8.67% COMMUNITY HEALTH PLAN OF WASHINGTON 3 3 17,386 7,446 -57.17% HEALTH PLAN OF CAREOREGON, INC. 3.5 3.5 11,298 12,708 12.48% BLUE CARE NETWORK OF MICHIGAN 4.5 4.5 68,138 75,964 11.49% CARE1ST HEALTH PLAN 3.5 3.5 51,433 64,856 26.10% ARCADIAN HEALTH PLAN, INC. The Challenges of Pharmacy Star Ratings • 55 Contract Enrollment for Plans Whose Star Ratings Stayed the Same in 2016 (continued) 2015 Overall Rating 2016 Overall Rating 2015 Contract Enrollment 2016 Contract Enrollment % Change CAPITAL HEALTH PLAN 4.5 4.5 17,121 18,486 7.97% H5970 HUMANA INSURANCE COMPANY OF NEW YORK 3.5 3.5 9,048 11,822 30.66% H5985 H5991 H6306 PHOENIX HEALTH PLANS, INC. 2.5 2.5 12,933 13,761 6.40% AFFINITY HEALTH PLAN, INC. 3.5 3.5 6,511 9,939 52.65% 4 4 4,020 5,197 29.28% H6622 HUMANA WI HEALTH ORGANIZATION INSURANCE CORP 4.5 4.5 37,872 45,213 19.38% H6801 H7006 H7200 H7301 GHS MANAGED HEALTH CARE PLANS, INC. 3 3 4,279 4,635 8.32% ATRIO HEALTH PLANS 4 4 2,988 3,459 15.76% AMERIGROUP TENNESSEE, INC. 3 3 6,466 7,783 20.37% COVENTRY HEALTH CARE OF ILLINOIS, INC. 4.5 4.5 15,455 17,724 14.68% H7728 ANTHEM HEALTH PLANS OF NEW HAMPSHIRE, INC. 4.5 4.5 1,505 1,713 13.82% H8145 HUMANA INSURANCE COMPANY 4 4 142,189 131,366 -7.61% H8552 ANTHEM BLUE CROSS LIFE AND HEALTH INS COMPANY 3.5 3.5 12,684 7,693 -39.35% H8604 H8649 H9001 H9003 H9082 THP INSURANCE COMPANY 3.5 3.5 693 1,055 52.24% AETNA HEALTH OF UTAH, INC. 3.5 3.5 9,301 7,119 -23.46% FALLON COMMUNITY HEALTH PLAN 4.5 4.5 17,619 18,241 3.53% 5 5 77,157 81,778 5.99% 3.5 3.5 1,564 2,805 79.35% H9572 BCBS OF MICHIGAN MUTUAL INSURANCE COMPANY 4 4 326,358 345,110 5.75% H9615 MVP HEALTH PLAN, INC. 4.5 4.5 19,459 22,858 17.47% H9730 WELLCARE HEALTH INSURANCE COMPANY OF KENTUCKY, INC 2.5 2.5 6,420 7,731 20.42% H9947 BLUE CROSS AND BLUE SHIELD OF GEORGIA, INC. 3 3 2,721 2,401 -11.76% R3175 UNITEDHEALTHCARE INSURANCE COMPANY 3.5 3.5 3,417 3,470 1.55% R3444 CARE IMPROVEMENT PLUS SOUTH CENTRAL INSURANCE CO. 3 3 56,288 64,956 15.40% R5287 UNITEDHEALTHCARE INSURANCE COMPANY 3.5 3.5 183,167 202,103 10.34% R5342 UNITEDHEALTHCARE INSURANCE COMPANY OF NEW YORK 3.5 3.5 85,946 103,903 20.89% R5941 ANTHEM INSURANCE COMPANIES, INC. 3 3 103,212 100,147 -2.97% R9896 CARE IMPROVEMENT PLUS SOUTH CENTRAL INSURANCE CO. 3 3 163,706 168,125 2.70% Contract Number Contract Name H5938 FIRSTCAROLINACARE INSURANCE COMPANY KAISER FOUNDATION HP OF THE N W MOLINA HEALTHCARE OF NEW MEXICO, INC. NOTE: Does not include new or terminated plans. SOURCE: Calculated by AIS from CMS data for AIS's Medicare and Medicaid Market Data: 2016. The Challenges of Pharmacy Star Ratings • 57 Appendix B: CMS Fact Sheet — 2016 Star Ratings Fact Sheet - 2016 Star Ratings One of the Centers for Medicare & Medicaid Services’ (CMS) most important strategic goals is to improve the quality of care and general health status for Medicare beneficiaries. CMS publishes the Part C and D Star Ratings each year to: measure quality in Medicare Advantage (MA) and Prescription Drug Plans (PDPs or Part D plans), assist beneficiaries in finding the best plan for them, and determine MA Quality Bonus Payments. Moreover, the ratings support the efforts of CMS to improve the level of accountability for the care provided by physicians, hospitals, and other providers. CMS continues to see increases in the number of Medicare beneficiaries in high-performing Medicare Advantage (MA) plans. Star Ratings are driving improvements in Medicare quality. The information included in this Fact Sheet is evidence of such improvement and is based on the 2016 Star Ratings published on Medicare Plan Finder (MPF) on October 8, 2015. Background Medicare Advantage with prescription drug coverage (MA-PD) contracts are rated on up to 44 unique quality and performance measures; MA-only contracts (without prescription drug coverage) are rated on up to 32 measures; and stand-alone PDP contracts are rated on up to 15 measures. Each year, CMS conducts a comprehensive review of the measures that make up the Star Ratings, considering the reliability of the measures, clinical recommendations, feedback received from stakeholders, and data issues. All measures transitioned from the Star Ratings are included in the display measure available from this page http://go.cms.gov/partcanddstarratings. Changes to existing measures are summarized in Attachment A. The Star Ratings measures span five broad categories: Outcomes Intermediate Outcomes Patient Experience Access Process For the 2016 Star Ratings, outcomes and intermediate outcomes continue to be weighted three times as much as process measures, and patient experience and access measures are weighted 1.5 times as much as process measures. CMS assigns a weight of 1 to all new measures. The Part C and D quality improvement measures receive a weight of 5 to further reward contracts for the strides they made to improve the care provided to Medicare enrollees. CMS continues to lower the overall Star Rating for contracts with serious compliance issues, defined as the imposition of enrollment or marketing sanctions. Highlights of Contract Performance in 2016 Star Ratings1 Changes in Ratings from 2015 The last row in Table 1 details the trend in the average overall Star Ratings weighted by enrollment for MA-PDs for the period of 2013 to 2016. Approximately 49 percent of MA-PDs (179 contracts) that will be active in 2016 earned four stars or higher for their 2016 overall rating. This is nearly a 9 percentage point increase from 40 percent of active contracts earning four stars or higher for their 2015 overall rating. Weighted by enrollment, close to 71 percent of MA-PD enrollees are in contracts with four or more stars. This is nearly an 11 percentage point increase from 60 percent of enrollees in contracts with four or more stars in 2015. 1 Tables contained in this document may not have sums of percentages of 100.00 due to rounding. 1 58 • AIS’s Management Insight Series Table 1: 2013 - 2016 Overall Star Rating Distribution for MA-PD Contracts 2013 Overall Rating 5 stars 4.5 stars 4 stars Number of Contracts % 2014 Weighted by Enrollment 11 54 2.46 12.08 9.42 15.81 Number of Contracts % 2015 Weighted by Enrollment 11 64 2.55 14.85 9.56 20.55 Number of Contracts 2016 Weighted by Enrollment Number of Contracts 11 2.78 61 15.44 9.88 19.59 12 65 3.25 17.62 10.23 25.02 % % Weighted By Enrollment 62 13.87 12.56 87 20.19 21.68 86 21.77 30.32 102 27.64 35.71 3.5 stars 3 stars 131 127 29.31 28.41 36.48 20.25 143 109 33.18 25.29 30.49 16.63 136 34.43 73 18.48 26.78 10.98 112 66 30.35 17.89 19.55 8.60 2.5 stars 60 13.42 5.28 16 3.71 1.09 26 6.58 2.37 12 3.25 0.90 2 447 0.45 0.21 1 431 0.23 0.01 2 395 0.51 0.08 0 369 0.00 0.00 2 stars Total Number of Contracts Average Star Rating* 3.71 3.86 3.92 4.03 * The average Star Rating is weighted by enrollment. The last row in Table 2 details the trend in the average Part D Ratings weighted by enrollment for PDPs for the period of 2013 to 2016 (Table 2). Approximately 41 percent of PDPs (24 contracts) that will be active in 2016 received four or more stars for their 2016 Part D rating. Weighted by enrollment, close to 32 percent of PDP enrollees are in contracts with four or more stars. There were more significant changes in the PDP scores this year due to one measure being retired (Diabetes Treatment), and 3 measures were included that were not used in the prior year. Given the smaller number of measures for PDPs, these changes have a more significant impact. Table 2: 2012 - 2016 Part D Rating Distribution for PDPs 2013 Part D Rating 5 stars 4.5 stars Number of Contracts % 4 5 5.71 7.14 4 stars 17 3.5 stars 3 stars 17 17 2.5 stars 2 stars 1.5 stars Total Number of Contracts 2014 Weighted by Enrollment Number of Contracts % 1.85 3.52 5 6 6.94 8.33 24.29 12.2 16 24.29 24.29 23.35 55.08 18 17 9 12.86 3.23 1 0 1.43 0.00 0.77 0.00 70 Average Star Rating* 3.30 * The average Star Rating is weighted by enrollment. 2015 Weighted by Enrollment Number of Contracts % 2016 Weighted Number by of Enrollment Contracts % Weighted by Enrollment 0.13 3.34 3 4.92 11 18.03 1.50 7.28 2 10 3.39 16.95 0.13 1.63 22.22 5.29 17 27.87 43.94 12 20.34 29.95 25.00 23.61 52.39 14.16 18 29.51 7 11.48 40.40 0.61 12 14 20.34 23.73 21.76 38.88 8 11.11 5.62 3 4.92 5.99 8 13.56 7.65 1 1 1.39 1.39 0.00 19.07 1 1 1.64 1.64 0.01 0.27 1 0 1.69 0.00 0.01 0 72 61 3.05 59 3.75 3.40 5-Star Contracts 17 contracts are highlighted on MPF with a high performing (gold star) icon; 12 are MA-PD contracts (Table 3), 3 are MA-only contracts (Table 4), and 2 are PDPs (Table 5). The seven new 5-star contracts for this year are: Cigna Healthcare of Arizona, Inc. (H0354) Tufts Associated Health Maintenance Organizations (H2256) Group Health Plan, Inc. (MN) (H2462) Essence Healthcare Inc. (H2610) Medical Associates Clinic Health Plan (H5256) Sierra Health and Life Insurance Company, Inc. (H5652) Tufts Insurance Company (S0655) 2 The Challenges of Pharmacy Star Ratings • 59 Table 3: MA-PD Contracts Receiving the 2016 High Performing Icon H0354 CIGNA HEALTHCARE OF ARIZONA, INC. Enrolled Non-EGHP Service Area* 10/2015 43,881 3 counties in AZ H0524 KAISER FOUNDATION HP, INC. 1,037,349 31 counties in CA Not applicable Yes Yes H0630 H1230 KAISER FOUNDATION HP OF CO KAISER FOUNDATION HP, INC. 98,584 17 counties in CO 31,396 3 counties in HI Not applicable Not applicable Yes Yes Yes Yes H2150 KAISER FNDN HP OF THE MID-ATLANTIC STS 63,681 D.C., 11 counties in MD, 9 counties in VA H2256 H2462 TUFTS ASSOCIATED HEALTH MAINTENANCE ORGANIZATION GROUP HEALTH PLAN, INC. (MN) H2610 H5262 H5591 H5652 SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. H9003 KAISER FOUNDATION HP OF THE N W Contract Contract Name EGHP Service Area* 13 counties in AZ 5 Star Last Year No SNP Yes Not applicable Yes No 104,812 10 counties in MA 49,484 87 counties in MN, 8 counties in WI Not applicable Not applicable No No Yes No ESSENCE HEALTHCARE, INC. 52,525 3 counties in IL, 10 counties in MO Not applicable No No GUNDERSEN HEALTH PLAN MARTIN'S POINT GENERATIONS, LLC 14,287 5 counties in IA, 11 counties in WI 32,611 16 counties in ME, 2 counties in NH Not applicable Most of the U.S. Yes Yes No Yes Not applicable No Yes 1 county in OR, 1 county in WA Yes No 1 county in CO, 1 county in KS, 2 counties in MA, 4,502 3 counties in MD, 1 county in MI, 3 counties in NJ, 2 counties in PA, 2 counties in TX, 1 county in VA 79,591 9 counties in OR, 4 counties in WA *An EGHP is a non-Employer Group and Employer Group Health Plan. Table 4: MA-only Contracts Receiving the 2016 High Performing Icon2 Contract Enrolled 10/2015 Contract Name Non-EGHP Service Area EGHP Service Area 5 Star Last Year H1651 H5256 MEDICAL ASSOCIATES HEALTH PLAN, INC. MEDICAL ASSOCIATES CLINIC HEALTH PLAN 10,398 6 counties in IA, 1 county in IL 3,176 4 counties in WI Not applicable Not applicable Yes No H5264 DEAN HEALTH PLAN, INC. 24,898 8 counties in WI Not applicable Yes Table 5: PDP Contracts Receiving the 2016 High Performing Icon Contract Contract Name Enrolled 10/2015 S0655 TUFTS INSURANCE COMPANY S5753 WISCONSIN PHYSICIANS SERVICE INSURANCE CORPORATION Non-EGHP Service Area 7,874 Not applicable 22,999 1 region - Wisconsin EGHP Service Area 5 Star Last Year 35 regions No 38 regions Yes Low Performers 6 contracts are identified on the MPF with the Low Performing Icon (LPI) for consistently low quality ratings. 3 contracts are receiving the LPI for low Part C ratings of 2.5 or fewer stars from 2014 through 2016. 3 contracts are receiving the LPI for low Part C or D ratings of 2.5 or fewer stars from 2014 through 2016. Below is the list of contracts receiving an LPI for 2016 (Table 6). Table 6: 2016 Contracts with a Low Performing Icon (LPI) Contract Contract Name Parent Organization H1903 H2905* WELLCARE OF LOUISIANA, INC. WellCare Health Plans, Inc. SIERRA HEALTH AND LIFE INSURANCE COMPANY, INC. UnitedHealth Group, Inc. H3327 H4866* H5698* H6801 Reason for LPI Enrolled 10/2015 Part C or D Part C 10,167 3,490 TOUCHSTONE HEALTH HMO, INC. Touchstone Health Partnership, Inc Part C or D 10,864 CUATRO LLC WINDSOR HEALTH PLAN, INC. Cuatro LLC. WellCare Health Plans, Inc. Part C Part C 4,519 40,606 GHS MANAGED HEALTH CARE PLANS, INC. Health Care Service Corporation Part C or D 4,422 *These contracts are eligible for termination at the end of 2016. 2 MA-only contracts cannot offer SNPs. 3 60 • AIS’s Management Insight Series Tax Status and Performance Organizations that are non-profit tend to receive higher ratings than those that are for-profit. For MAPDs, approximately 70% of the non-profit contracts received 4 or more stars compared to 39% of the for-profit MA-PDs. Similarly, for PDPs approximately 63% of non-profit PDPs received 4 or more stars compared to 24% of the for-profit PDPs. Non-profit organizations also performed better than forprofit organizations last year. Below is the ratings distribution by tax status for MA-PD (Table 7) and PDP (Table 8) contracts. Table 7: Distribution of Overall Star Ratings for For-profit and Non-profit MA-PDs 2016 Overall Rating Number of For-Profit % ForProfit Weighted By Number of Enrollment For-Profit Non-Profit % NonProfit Weighted By Enrollment Non-Profit 5 stars 4.5 stars 3 33 1.20 13.15 0.96 23.83 9 32 7.63 27.12 28.96 27.41 4 stars 3.5 stars 61 90 24.30 35.86 38.56 25.70 41 22 34.75 18.64 29.97 7.12 3 stars 52 20.72 9.61 14 11.86 6.55 12 251 4.78 1.34 0 118 0.00 0.00 2.5 stars Total Number of Contracts Table 8: Distribution of Part D Ratings for For-profit and Non-profit PDPs 2016 Part D Rating Number of For-Profit % ForProfit 5 stars 4.5 stars 1 4 4 stars 3.5 stars 3 stars 2.5 stars 2 stars Total Number of Contracts Weighted By Number of Enrollment For-Profit Non-Profit % NonProfit Weighted By Enrollment Non-Profit 2.94 11.76 0.03 0.40 1 5 4.17 20.83 2.38 13.53 3 8.82 28.69 9 37.50 53.96 7 12 20.59 35.29 20.96 40.54 5 2 20.83 8.33 19.9 1.82 7 20.59 9.37 1 4.17 8.22 0 34 0.00 0.00 1 24 4.17 0.18 Length of Time in Program and Performance On average, higher Star Ratings are associated with more experience in the MA program. We see a similar pattern for PDPs. The tables below show the distribution of ratings by the number of years in the program (MA-PDs are shown in Table 9 and PDPs in Table 10). Table 9: Distribution of Overall Star Ratings by Length of Time in Program for MA-PDs 2016 Overall Rating 5 stars % Less than 5 years % 5 years to less than 10 years % Greater than 10 years 0.00 0.81 5.45 4.5 stars 4 stars 13.64 25.00 8.94 21.14 23.76 32.18 3.5 stars 18.18 39.02 27.72 3 stars 2.5 stars 29.55 13.64 28.46 1.63 8.91 1.98 44 123 202 Total Number of Contracts Table 10: Distribution of Part D Ratings by Length of Time in Program for PDPs 2016 Part D Rating 5 stars 4.5 stars % Less than 5 years % 5 years to less than 10 years 0.00 0.00 6.00 20.00 4 stars 40.00 26.00 3.5 stars 3 stars 0.00 40.00 34.00 8.00 2.5 stars 2 stars Total Number of Contracts 0.00 6.00 20.00 5 0.00 50 4 The Challenges of Pharmacy Star Ratings • 61 Performance of Contracts Eligible to Receive Low Income Subsidy (LIS) Auto-assignees Most contracts with a Star Rating and eligible to receive LIS auto-assignees (LIS contracts) continue to earn a Star Rating of 3 or more (Table 11). Thirteen out of 15 LIS contracts (86.7%) earned a Star Rating of 3 or more in 2016 compared to 15 contracts (93.8%) in 2015, 16 (84.2%) in 2014, and 17 (89.5%) in 2013. Table 11: Distribution of Part D Ratings for PDPs Eligible to Receive LIS Auto-assignees Part D Rating 2013 Number of 2013 % of 2014 Number of 2014 % of 2015 Number of 2015 % of 2016 Number of LIS Contracts LIS Contracts LIS Contracts LIS Contracts LIS Contracts LIS Contracts LIS Contracts 2016 % of LIS Contracts 4.5 stars 0 0.00 0 0.00 1 6.25 0 0.00 4 stars 1 5.26 4 21.05 4 25.00 2 13.30 6 31.58 6 31.58 8 50.00 4 26.70 10 52.63 6 31.58 2 12.50 7 46.70 2.5 stars 2 10.53 3 15.79 1 6.25 2 13.30 2 stars 0 0.00 0 0.00 0 0.00 0 0.00 1.5 stars 0 0.00 0 0.00 0 0.00 0 0.00 3.5 stars 3 stars Total Number of Contracts 19 19 16 15 Geographic Variation The following eight maps illustrate the average Star Ratings weighted by enrollment per county for MA-PDs and PDPs across the U.S., including territories, between 2013 and 2016.3 These maps exclude the employer group health plans. Counties shaded in green indicate that the average Overall Star Rating weighted by enrollment in the county for MA-PDs or average Part D Rating for PDPs is four or more stars. Counties shaded in yellow indicate that the average rating weighted by enrollment for the county for MA-PDs or PDPs is three stars. Areas shaded in orange indicate that the average rating weighted by enrollment is less than three stars. Areas in gray indicate data are not available for those counties. Among the changes and updates are: 3 The availability of highly rated MA-PDs has increased since 2013. The MA-PD maps for 2016 compared to 2013 show significantly more light green (3.5 stars) and green (4 or more stars) compared to yellow (3 stars) and orange (2.5 stars) in 2013. In 2016 the average rating weighted by enrollment for PDPs across the county is at least 3.0 stars. Comparisons of Star Ratings across years do not reflect annual revisions made by CMS to the Star Rating’s methodology or measure set. 5 62 • AIS’s Management Insight Series 6 For a full-color version of maps on the following pages, see https://www. cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovGenIn/ Downloads/2016-Part-C-and-D-Medicare-Star-Ratings-Data-v04-07-2016-.zip. Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2015 Star Ratings - Enrollment Weighted Average MA-PD Overall Rating in Non-EGHP Counties 5 Stars 7 The Challenges of Pharmacy Star Ratings • 63 Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2014 Star Ratings - Enrollment Weighted Average MA-PD Overall Rating in Non-EGHP Counties 5 Stars 8 64 • AIS’s Management Insight Series Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2013 Star Ratings - Enrollment Weighted Average MA-PD Overall Rating in Non-EGHP Counties 5 Stars 9 The Challenges of Pharmacy Star Ratings • 65 10 66 • AIS’s Management Insight Series Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2015 Star Ratings - Enrollment Weighted Average PDP Part D Rating in Non-EGHP Counties 5 Stars 11 The Challenges of Pharmacy Star Ratings • 67 Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2014 Star Ratings - Enrollment Weighted Average PDP Part D Rating in Non-EGHP Counties 5 Stars 12 68 • AIS’s Management Insight Series Missing Data 1 Star 1.5 Stars 2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 2013 Star Ratings - Enrollment Weighted Average PDP Part D Rating in Non-EGHP Counties 5 Stars 13 The Challenges of Pharmacy Star Ratings • 69 70 • AIS’s Management Insight Series Average Star Rating for Each Measure Below we list the average Star Ratings for 2013, 2014, 2015, and 2016 Part C and D measures (Tables 12 and 13). In general, Star Ratings have gone up from 2013 to 2016 for most measures.4 Table 12: Average Star Rating by Part C Measure 2016 Measure Number Measure 2013 Average Star 2014 Average Star 2015 Average Star 2016 Average Star C01 C02 Breast Cancer Screening Colorectal Cancer Screening 3.0 3.5 3.3 n/a - not used in 2015 3.9 4.2 3.6 3.2 C03 Annual Flu Vaccine 3.2 3.4 3.3 3.3 C04 C05 Improving or Maintaining Physical Health Improving or Maintaining Mental Health 4.4 2.2 4.5 2.0 4.6 2.5 3.3 3.3 C06 Monitoring Physical Activity 2.1 2.4 2.2 2.9 C07 C08 Adult BMI Assessment Special Needs Plan (SNP) Care Management 3.7 n/a – new in 2015 3.8 n/a – new in 2015 3.8 2.7 4.1 2.5 C09 Care for Older Adults – Medication Review 3.0 3.6 3.9 4.3 C10 C11 Care for Older Adults – Functional Status Assessment Care for Older Adults – Pain Assessment 2.8 3.2 3.4 3.2 3.4 4.0 3.9 4.1 C12 Osteoporosis Management in Women who had a Fracture 1.4 1.9 2.1 2.5 C13 C14 Diabetes Care – Eye Exam Diabetes Care – Kidney Disease Monitoring 3.4 4.3 4.0 4.5 3.7 4.2 3.1 3.3 C15 Diabetes Care – Blood Sugar Controlled 3.1 3.3 3.3 3.9 C16 C17 Controlling Blood Pressure Rheumatoid Arthritis Management 3.5 3.3 3.5 3.7 3.7 3.5 3.4 3.2 C18 Reducing the Risk of Falling 3.3 3.4 3.3 2.7 C19 C20 Plan All-Cause Readmissions Getting Needed Care 3.0 3.5 3.5 3.6 3.0 3.4 3.3 3.5 C21 Getting Appointments and Care Quickly 3.4 3.5 3.5 3.4 C22 C23 Customer Service Rating of Health Care Quality 3.4 3.7 3.5 3.7 3.5 3.7 3.5 3.4 C24 Rating of Health Plan 3.3 3.4 3.4 3.3 C25 C26 Care Coordination Complaints about the Health Plan 3.4 3.0 3.4 3.0 3.4 4.2 3.4 3.9 C27 Members Choosing to Leave the Plan 3.5 3.7 4.3 4.2 C28 C29 Beneficiary Access and Performance Problems Health Plan Quality Improvement 3.5 3.1 3.4 n/a - not used in 2015 3.5 3.5 4.2 3.4 C30 Plan Makes Timely Decisions about Appeals 4.0 4.1 4.2 4.1 C31 C32 Reviewing Appeals Decisions Call Center – Foreign Language Interpreter and TTY Availability 3.3 4.2 3.3 3.7 4.4 n/a - not used in 2015 3.6 4.3 Changes in the average Star Rating do not always reflect changes in performance since for some measures there have been significant changes in industry performance and shifts in the distribution of scores. The pre-determined star thresholds were removed for the 2016 Star Ratings. Some measures may have greater shifts from 2015 to 2016 compared to other time periods due to the revision to the methodology used to determine the rating. 4 14 The Challenges of Pharmacy Star Ratings • 71 Table 13: Average Star Rating by Part D Measure for MA-PDs 2016 Measure Measure Number D01 Call Center – Foreign Language Interpreter and TTY Availability 2013 MA-PD Average Star 3.7 2014 MA-PD Average Star 2015 MA-PD 2016 MA-PD Average Star Average Star 3 n/a – not used in 2015 4.2 D02 Appeals Auto–Forward 3.4 3.4 3.6 4.5 D03 D04 Appeals Upheld Complaints about the Drug Plan 3.2 3.0 3.3 3 3.7 4.2 3.3 3.9 D05 Members Choosing to Leave the Plan 3.5 3.7 4.3 4.2 D06 D07 Beneficiary Access and Performance Problems Drug Plan Quality Improvement 3.5 3.4 3.3 n/a – not used in 2015 3.7 4.1 4.2 3.8 D08 Rating of Drug Plan 3.4 3.4 3.5 3.3 D09 D10 Getting Needed Prescription Drugs MPF Price Accuracy 3.5 3.8 3.5 3.9 3.4 4.6 3.4 3.5 D11 High Risk Medication 3.1 3.6 3.2 4.1 D12 D13 Medication Adherence for Diabetes Medications Medication Adherence for Hypertension (RAS antagonists) 3.1 3.0 3.7 3.7 3.5 3.1 3.9 4.1 D14 Medication Adherence for Cholesterol (Statins) 3.1 3.6 3.3 4.0 D15 MTM Program Completion Rate for CMR n/a – new in 2016 n/a – new in 2016 n/a – new in 2016 2.3 Table 14: Average Star Rating by Part D Measure for PDPs 2016 Measure Measure Number D01 Call Center – Foreign Language Interpreter and TTY Availability 2013 PDP Average Star 3.8 2014 PDP 2015 PDP 2016 PDP Average Star Average Star Average Star 3.7 n/a – not used in 2015 4.0 D02 D03 Appeals Auto–Forward Appeals Upheld 2.4 3.3 2.7 3.3 2.5 3.9 4.1 3.1 D04 Complaints about the Drug Plan 3.7 3.4 4.3 3.5 D05 D06 Members Choosing to Leave the Plan Beneficiary Access and Performance Problems 3.7 3.8 3.3 3.7 3.8 n/a – not used in 2015 3.6 3.9 D07 Drug Plan Quality Improvement 4.1 3.6 4.2 3.8 D08 D09 Rating of Drug Plan Getting Needed Prescription Drugs 3.6 3.7 3.7 4.1 3.9 3.8 3.2 3.6 D10 MPF Price Accuracy 4.2 4.1 4.7 4.7 D11 D12 High Risk Medication Medication Adherence for Diabetes Medications 3.1 3.3 2.8 3.1 2.7 3.0 3.1 2.7 D13 Medication Adherence for Hypertension (RAS antagonists) 3.2 3.6 3.8 3.6 D14 D15 Medication Adherence for Cholesterol (Statins) MTM Program Completion Rate for CMR 3.2 3.6 n/a – new in 2016 n/a – new in 2016 4.2 n/a – new in 2016 3.5 2.3 15 72 • AIS’s Management Insight Series Attachment A – 2016 Star Ratings Measure Changes Below are some additional changes to the 2016 Star Ratings in terms of the measures included. Specification Changes Part C measure: C12 – Osteoporosis Management in Women who had a Fracture – NCQA has added an upper age limit, extended the look back period for exclusions due to prior bone mineral testing, removed estrogens from this measure, and removed single-photon absorptiometry and dualphoton absorptiometry tests from the list of eligible bone-density tests. Part C measure: C16 – Controlling Blood Pressure – measure updated to include two different blood pressure thresholds based on age and diagnosis. Part C measure: C19 – Plan All-Cause Readmissions – now excludes planned readmissions from the measure and removes the current exclusion from the denominator for hospitalizations with a discharge date in the 30 days prior to the Index Admission Date. Part C measure: C30 – Plan Makes Timely Decisions about Appeals – removed dismissed appeals from the measure. Part C & D measures: C26 & D04 – Complaints about the Health/Drug Plan – modified the measurement period from 6 months of the current year to 12 months of the prior year. Part D measure: D03 – Appeals Upheld – modified the measurement period to coincide with the 12 month period of the Part D Appeals Auto-forward measure. Part D measures: D12 & D13 – both measures adjusted to account for beneficiaries with End-Stage Renal Disease (ESRD). Part D measures: D12, D13 & D14 – calculating the proportion of days now uses the date of death for a member instead of the last day of the month. Part D measures: D11 – D14 - Implemented PQA’s 2014 obsolete NDCs methodology. Part C & D CAHPS measures: Implemented CAHPS methodology modifications which permit low-reliability contracts to receive 5 stars or 1 star. Eliminated pre-determined 4-star thresholds Included data in HEDIS measures for contracts with 500-999 enrolled in July of the measurement year. Additions Part C measure: C01 - Breast Cancer Screening: with a weight of 1. Part D measure: D15 – Medication Therapy Management Program Completion Rate for Comprehensive Medication Reviews: with a weight of 1. Parts C & D measure: C32 & D01 - Call Center – Foreign Language Interpreter and TTY Availability: with a weight of 1.5. Parts C & D measures: C28 & D06 - Beneficiary Access and Performance Problems: with a weight of 1. 16 The Challenges of Pharmacy Star Ratings • 73 Transitioned Measures Transitioned measures are measures that have moved to the display page which can be found on the CMS website at this address: http://go.cms.gov/partcanddstarratings Part C measure: Improving Bladder Control Retired measures Part C measure: Cardiovascular Care: Cholesterol Screening Part C measure: Diabetes Care: Cholesterol Screening Part C measure: Diabetes Care: Cholesterol Controlled Part D measure: Diabetes Treatment 17 The Challenges of Pharmacy Star Ratings • 75 Appendix C: CMS Guidance and Documentation for 2017 Star Ratings DEPARTMENT OF HEALTH & HUMAN SERVICES Centers for Medicare & Medicaid Services 7500 Security Boulevard Baltimore, Maryland 21244-1850 CENTER FOR MEDICARE DATE: November 12, 2015 TO: All Medicare Advantage Organizations, Prescription Drug Plan Sponsors, and Other Interested Parties FROM: Amy K. Larrick, Acting Director, Medicare Drug Benefit and C & D Data Group SUBJECT: Request for Comments: Enhancements to the Star Ratings for 2017 and Beyond (Due by December 10, 2015 at 5pm ET) This document proposes methodology changes for the 2017 Star Ratings and display measures for Medicare Advantage (MA) and Prescription Drug Plans (PDPs). It also provides advanced notice of potential changes for the Star Ratings and display measures for 2018 and beyond. Based on this memo, MA Organizations, PDP sponsors, advocates, and other stakeholders have the opportunity to provide comments prior to the draft Call Letter, which is issued as part of the Advance Notice. The statutory deadlines in section 1853 result in a short comment period between the Advance and Final Rate Notices, which contain the Call Letter. We are issuing this Request for Comments (RFC) to provide an opportunity for stakeholders to provide input prior to the Call Letter process. CMS structured the current Star Ratings strategy to be consistent with the six priorities in the National Quality Strategy. The six priorities include: making care safer by reducing harm caused by the delivery of care; ensuring that each person and family are engaged as partners in their care; promoting effective communication and coordination of care; promoting the most effective prevention and treatment practices for the leading causes of mortality; working with communities to promote wide use of best practices to enable healthy living; and making quality care more affordable for individuals, families, employers, and governments by developing and spreading new health care delivery models. The measures span five broad categories, including: Outcome measures that focus on improvement to a beneficiary’s health as a result of care that is provided; Intermediate outcome measures that concentrate on ways to help beneficiaries move closer to achieving true outcomes; Patient experience measures that represent beneficiaries’ perspectives about the care they receive; 1 76 • AIS’s Management Insight Series Access measures that reflect processes or structures that may create barriers to receiving needed health care; and Process-of-care measures that capture a method by which health care is provided. The Star Ratings help inform beneficiaries about the performance of health and drug plans on the Medicare Plan Finder (MPF) website, as well as serve as the basis of Quality Bonus Payments (QBPs) for MA organizations. CMS continues to improve the Part C and Part D quality and performance measurement system and focuses it on beneficiary outcomes, beneficiary experience, population health, and health care efficiency. The goal is that the Star Ratings system will not only influence beneficiaries’ plan choices but also drive organizations and sponsors toward higher quality and more efficient care. Star Ratings is a year-round process for both CMS and sponsors. Below is an example for the next cycle of Star Ratings, beginning with this RFC: November 2015-February 2016: CMS provides guidance on methodology changes anticipated for the 2017 Star Ratings and display measures and advanced notice of potential changes for the Star Ratings and display measures for 2018 and beyond. CMS does this first through the RFC and then in the draft 2017 Call Letter. Sponsors and other stakeholders actively participate by submitting comments regarding the potential changes to the Star Ratings methodology or submitting additional comments to propose other changes not mentioned in the RFC. Comments to the RFC inform proposals in the draft Call Letter, and comments to the draft Call Letter inform the final Call Letter. February-April 2016: After consideration of all comments, CMS announces the methodology for the 2017 Star Ratings in the final 2017 Call Letter in the Rate Announcement issued. Ongoing: Sponsors review the underlying data used for the individual Star Ratings measures as they become available throughout the year, and notify CMS of any errors or questions in a timely manner. Summer 2016: CMS holds a Part C and D User call with sponsors to provide updates to the upcoming Star Ratings release, and conducts two plan preview periods via the Health Plan Management System (HPMS) to identify any necessary data corrections or revisions to our draft Technical Notes. During these preview periods, CMS expects sponsors to raise concerns about their raw measure data and Star Ratings. Changes to the methodologies for measure calculations or Star Rating calculations cannot be made during this time. November 2016: CMS offers MA organizations an appeals process for QBPs. The administrative review process is a two-step process that includes a request for reconsideration and a request for an informal hearing after CMS has sent the MA organization the reconsideration decision. This process may only be initiated on the basis of non-methodological challenges, such as a calculation error (miscalculation) or a data inaccuracy (incorrect data). Comments to this RFC should be submitted via the following link: https://cmsgov.wufoo.com/forms/enhancements-to-the-star-ratings-for-2017/ 2 The Challenges of Pharmacy Star Ratings • 77 The online form allows comments on up to 8 sections. If your organization’s comments exceed this maximum, you may submit the form more than once. Do not resubmit these comments to CMS via email. If you wish to submit additional supporting documents, you may send them via email to: [email protected]. NOTE: If you encounter the following error message while completing the survey, you must clear cookies and then resubmit the form: “There was a problem with your submission. Unable to create a new entry.” Comments submitted by Thursday, December 10 at 5pm ET will be considered as we finalize proposed changes for the 2017 Star Ratings for the draft 2017 Call Letter. CMS will post on its website all comments received to the RFC. Stakeholders will have another opportunity to comment on the 2017 Star Ratings methodology and proposed changes through the Advance Notice/draft Call Letter process. Questions related to this RFC may be sent to: [email protected]. Thank you for your participation. 3 78 • AIS’s Management Insight Series Enhancements to the 2017 Star Ratings and Beyond One of CMS’ most important strategic goals is to improve the quality of care and general health status of Medicare beneficiaries. For the 2017 Star Ratings, CMS continues to enhance the current methodology so it further aligns with our policy goals. Our priorities include enhancing the measures and methodology to reflect the true performance of organizations and sponsors, maintaining stability due to the link to payment, and providing advance notice of future changes. In this document, we describe the enhancements being considered for the 2017 Star Ratings and beyond. CMS is not considering adding any new measures for 2017 Star Ratings. Unless noted below, we anticipate the methodology remaining the same as the 2016 Star Ratings. For reference, the list of measures and methodology included in the 2016 Star Ratings is described in the Technical Notes available on the CMS webpage: http://go.cms.gov/partcanddstarratings. The cut points to determine star assignments for all measures and case-mix coefficients for the Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey and Health Outcomes Survey (HOS) will be updated for 2017 using the most current data available. As announced in previous years, we will review data quality across all measures, variation among organizations and sponsors, and measures’ accuracy and validity before making a final determination about inclusion of measures in the Star Ratings. A. Changes to Measures for 2017 CMS’ general policies regarding specification changes to Star Ratings measures: If a specification change to an existing measure is announced in advance of the measurement period, the measure remains in the Star Ratings; it will not be moved to the display page. If the change announced during the measurement period significantly expands the denominator or population covered by the measure, the measure is moved to the display page for at least one year. If the change announced during the measurement period does not significantly impact the numerator or denominator of the measure, the measure will continue to be included in the Star Ratings (e.g., when during the measurement period additional codes are added that would increase the number of numerator hits for a measure). The methodology for the following measures is being modified: 1. Improvement measures (Part C & D). The methodology for the improvement measures remains the same as in prior years. We have updated the measures included in the improvement measure to account for measures with at least two years of data. Please refer to the Appendix for updates to the measures to be used to calculate the 2017 improvement measures. We are also considering that if a contract’s CAHPS measure score moved to very low reliability with the exclusion of the enrollees with less than 6 months of continuous 4 The Challenges of Pharmacy Star Ratings • 79 enrollment for the 2015 survey administration, then the 2014 CAHPS measure score (used in 2015 Star Ratings) would be used instead as the baseline for the 2017 improvement calculation for that measure. 2. Reviewing Appeals Decisions/Appeals Upheld measures (Part C & D). Currently, these measures include cases that are reopened and decided by April 1 of the following contract year. In some instances, appeals filed in the 4th quarter of the year and then subsequently reopened may not be determined by the Independent Review Entity (IRE) by April 1. We propose for the 2017 Star Ratings to modify these measure specifications so that if a Reopening occurs and is decided prior to May 1, 2016, the Reopened decision would be used. Reopenings decided on or after May 1, 2016 would not be reflected in these data, and the original decision result would be used. 3. Contract Enrollment Data (Part C & D). Contract enrollment numbers are pulled from HPMS for the Part C and D “Complaints about the Health/Drug Plan” and the Part D “Appeals AutoForward” measures. Additionally, plan-level enrollment is pulled for the three Part C “Care for Older Adults” measures. For these measures, twelve months of enrollment files are pulled from HPMS, and the average enrollment from those months is used in the measure calculations. We propose going forward to adjust the twelve months from January to December to February through January of the relevant measurement period. HPMS enrollment numbers are determined between the end of the first week to the beginning of the second week of the prior month. For example, January enrollment numbers reflect enrollment as determined as of the first week of December. Thus, we are modifying the enrollment numbers to reflect an average of the HPMS enrollment from February through January of the measurement year. 4. Transition from ICD-9 to ICD-10 (Part C & D). The measure stewards, such as the Pharmacy Quality Alliance (PQA) and the National Committee for Quality Assurance (NCQA), are in the process of reviewing their measure specifications with diagnosis-related requirements to transition from ICD-9 to ICD-10. Once the PQA updates their measure specifications, we will provide more information. We will test and adopt the changes implemented by PQA as appropriate for the Part D Star Ratings and display measures. NCQA has incorporated the ICD-10 codes in the 2016 Healthcare Effectiveness Data and Information Set (HEDIS). During the transition period both ICD-9 and ICD-10 codes will be used due to the look-back periods for some measures. 5. Appeals Upheld measure (Part D). For the 2016 Star Rating Upheld measure, we excluded appeal cases for beneficiaries enrolled in hospice at any point during 2014. As noted in the 2016 Call Letter, this exclusion was only necessary for the 2016 measure as it is based on 2014 data that may have been affected by policy changes in 2014. This exclusion will not be continued for the 2017 Star Rating Appeals Upheld measure. 5 80 • AIS’s Management Insight Series 6. Medication Therapy Management (MTM) Program Completion Rate for Comprehensive Medication Reviews (CMR) measure (Part D). We will add a detailed file during each HPMS plan preview period to list each contract’s underlying denominator, numerator, and Data Validation score since exclusions are applied to the plan-reported MTM data. The CMR rate measure is an initial measure of the delivery of MTM services, and we continue to look forward to the development and endorsement of outcomes-based MTM measures as potential companion measures to the current MTM Star Rating. Lastly, we will be implementing additional data integrity checks (discussed later in this memo) to safeguard against inappropriate attempts to bias the data used for this measure. B. Removal of Measures from Star Ratings 1. Improving Bladder Control (Part C). This measure, collected through the Health Outcomes Survey (HOS), assesses the percentage of beneficiaries with a urine leakage problem who discussed their problem with their provider and received treatment for the problem. NCQA made three changes to this measure. First, NCQA changed the denominator of both indicators to include all adults with urinary incontinence, as opposed to limiting the denominator to those who consider urinary incontinence to be an issue. This will remove a potential bias toward only sampling patients who were treated unsuccessfully. Second, NCQA changed the treatment indicator to assess whether treatment was discussed, as opposed to it being received. This will change the measure focus from receiving potentially inappropriate treatments, which often have adverse side effects, to shared decision making between the patient and provider about the appropriateness of treatment. Third, NCQA added an outcome indicator to assess how much urinary incontinence impacts quality of life for beneficiaries. This outcome indicator will not be part of the Star Rating system until additional analyses have been completed. These changes required revising the underlying survey questions in HOS. The revised questions were first collected in 2015. As a result of these changes, this measure will not be reported in the 2017 Star Ratings. The revised measure will be reported on the 2017 display page since the survey was first fielded with the new questions in 2015. The 2016 display measure uses data from the old questions. 2. High Risk Medication (Part D). The High Risk Medication (HRM) measure calculates the percent of Medicare Part D beneficiaries 65 and older who received two or more prescription fills for the same HRM drug with a high risk of serious side effects in the elderly. The measure is endorsed by the PQA and National Quality Forum (NQF), and the HRM rate is calculated using the PQA specifications and medication list based on American Geriatrics Society (AGS) recommendations. The AGS recently released the 2015 update of the Beers Criteria. The HRM measure will be removed from the Star Ratings and moved to the display measures for 2017. This proposal is based on a number of factors. While the AGS states that the criteria may be used as both an educational tool and quality measure, the AGS further states that the intent is not to apply the criteria in a punitive manner. Specifically, the addition of a drug to the HRM list is not a contraindication to use, rather an encouragement to avoid use in 6 The Challenges of Pharmacy Star Ratings • 81 the senior population without consideration of risks and benefits based on individual patient considerations. This is a very difficult decisional balance to evaluate in a drug plan that does not have access to full clinical information. As the measure can be calculated only by using prescription drug event (PDE) data, medications cannot be included on the HRM List that have risks conditional on clinical factors that cannot be measured using PDE data alone. As a result, some “Avoid” medications are included in the measure, while others are not. This may create unintended consequences including the inappropriate encouragement of certain non-HRM medications, which may not be the best choice for an individual beneficiary’s clinical circumstance. Lastly, because it is under direct provider control and should not be affected by non-clinical beneficiary characteristics, the HRM measure was not included in CMS’ overall analysis to assess the impact of socio-economic status (SES) on the Star Ratings (discussed later in this memo). However, our initial analysis found that after controlling for contract effects and dual eligible or low income subsidy status, there is a significant association between dual eligible/low income status and HRM use. This association remains after further controlling for age, sex, and race/ethnicity. We recommend that the measure developers further review this measure to better understand the associations. Avoiding potentially inappropriate medications in older adults remains important for quality of care for Medicare beneficiaries. Therefore, the HRM measure will move to the 2017 display page and may be considered for the Star Ratings again in the future. We will continue to provide HRM measure reports to Part D sponsors on a monthly basis through the Patient Safety Analysis website, and we will continue to identify outliers. If measure updates are endorsed by the PQA with sufficient lead time ahead of the 2017 formulary and bid deadlines in May and June 2016, CMS may consider adoption for the 2019 display page (using 2017 data). We will provide additional information on updates if available in the draft 2017 Call Letter. C. Data Integrity It is essential that the data used for CMS’ Star Ratings are accurate and reliable. CMS’ policy is to reduce a contract’s measure rating to 1 star if it is determined that biased or erroneous data have been submitted. This would include cases where CMS finds mishandling of data, inappropriate processing, or implementation of incorrect practices by the organization/sponsor have resulted in biased or erroneous data. Examples would include, but are not limited to: a contract’s failure to adhere to HEDIS, HOS, or CAHPS reporting requirements; a contract’s failure to adhere to Plan Finder or PDE data requirements; a contract’s errors in processing coverage determinations/exceptions or organization determinations found through program audits or other reviews; compliance actions due to errors in operational areas that would directly impact the data reported or processed for specific measures; or a contract’s failure to pass Part C and D Reporting Requirements Data Validation related to organization/sponsor-reported data for specific measures. 7 82 • AIS’s Management Insight Series CMS has taken several steps in the past years to protect the integrity of the data; however, we continue to identify new vulnerabilities where inaccurate or biased data could exist. We also must safeguard against the Star Ratings program creating perverse incentives for sponsors. CMS program audits will soon include review of Part D sponsors’ MTM programs. We intend to review and apply any relevant MTM program audit findings that could demonstrate sponsors’ MTM data were biased, outside of the Data Validation results. CMS is concerned about sponsor activities that may not be detected by routine Data Validation standards, such as attempts to restrict eligibility from their approved MTM programs, encouraging beneficiary opt-out of MTM programs within the first 60-days, or CMRs that do not meet CMS’ definition per guidance. CMS may perform additional audits or reviews to ensure the validity of data for specific contracts. Without rigorous validation of Star Ratings data, there is risk that CMS will reward contracts with falsely high ratings. D. Impact of Socio-economic and Disability Status on Star Ratings A key goal of the MA and Part D programs is to achieve greater value and quality for all beneficiaries; therefore, an important corollary is that we do not distort quality signals in our measures, or mask true differences in quality of care. CMS continuously reviews the Star Ratings methodology to improve the process, incentivize plans, and provide information that is a true reflection of the performance and experience of the enrollees. The policies implemented must result in high quality of care and health outcomes for all of our beneficiaries, while acknowledging the unique challenges of serving traditionally underserved subsets of the population. A number of MA organizations and PDP sponsors believe that enrollment of a high percentage of dual eligible (DE) enrollees and/or enrollees who receive a low income subsidy (LIS) limits their plan’s ability to achieve high MA or Part D Star Ratings. CMS has responded to the concern of our stakeholders and has comprehensively gathered information to determine if the Star Ratings are sensitive to the socio-economic and disability status of a contract’s enrollees. If adjustments are made to address this issue, they must be data driven. For example, if a disparity is due to challenges in serving disabled beneficiaries, rather than in serving those with lower SES, then the adjustment should clearly focus on disability status of beneficiaries. Similarly, unless our methods are transparent and open to input from a breadth of sources, contracts will not be able to translate as easily our findings into actionable quality improvement steps. With support from our contractors, we have undertaken research to provide the scientific evidence as to whether MA or Part D sponsors that enroll a disproportionate number of vulnerable beneficiaries are systematically disadvantaged by the current Star Ratings. Last year, we issued a Request for Information to gather information directly from organizations to supplement the data that CMS collects, as we believe that plans and sponsors are uniquely positioned to provide both qualitative and quantitative information that is not available from other sources. In February and September 2015, we released details on our research and 8 The Challenges of Pharmacy Star Ratings • 83 findings to-date.1 We have also reviewed reports about the impact of SES on quality ratings, such as the report published by the National Quality Forum (NQF) posted at www.qualityforum.org/risk_adjustment_ses.aspx and both the Medicare Payment Advisory Commission’s (MedPAC) Report to the Congress: Medicare Payment Policy posted at http://www.medpac.gov/documents/reports/mar2015_entirereport_revised.pdf?sfvrsn=0 and their recent release on September 10th entitled Factors Affecting Variation in Medicare Advantage Plan Star Ratings posted at http://www.medpac.gov/documents/september-2015meeting-presentation-factors-affecting-variation-in-medicare-advantage-plan-starratings.pdf?sfvrsn=0. The IMPACT Act (P.L. 113-185) instructs ASPE (Office of the Assistant Secretary for Planning and Evaluation) to conduct a study before October 2016 that examines the effect of individuals’ SES on quality measures, resource use and other measures for individuals under the Medicare program. Because ASPE’s research agenda aligns closely with our goals, we have and will continue to work collaboratively with ASPE and other governmental agencies to broaden and expand the focus of the issue. We note that, as instructed by Congress in the IMPACT Act, ASPE is conducting additional research in this area and may make recommendations for additional changes in the future. We look forward to their continued input. Further, CMS has engaged measure developers, NCQA and the PQA, to examine measure specifications used in the Star Ratings program to determine if measure respecification is warranted. CMS’ work is not complete, and we will continue to work diligently to address this issue and others that may lie in the future with the goal that all MA and Part D beneficiaries receive the highest quality care possible. The Star Rating system was designed to foster continuous quality improvement in the MA and Part D programs. As such, the Star Ratings program - its measures and methodology - are reviewed on an ongoing basis. We are committed to providing beneficiaries information on Medicare Plan Finder that is a true reflection of the care and experience of the plans’ members and to incentivize plans based on this same information. As stated in the 2016 Final Call Letter, CMS believed additional research into the nature of the differential performance on a subset of measures was necessary before any interim or permanent changes in the Star Ratings measurements could be developed and implemented. The additional research conducted after the publication of the 2016 Final Call Letter allowed for further examination of LIS/DE differences (“effects”) and their magnitude. Due to the considerable overlap between LIS/DE beneficiaries and disabled beneficiaries, the research was expanded to consider the possible role of disability status. The research considered the association between the performance on Star Ratings measures and enrollment of LIS/DE/disabled beneficiaries, and the variability of differences in performance on each measure by contract to 1 The February release can be found at https://www.cms.gov/medicare/prescription-drugcoverage/prescriptiondrugcovgenin/performancedata.html The September release can be found at https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovGenIn/Downloads/Research-onthe-Impact-of-Socioeconomic-Status-on-Star-Ratingsv1-09082015.pdf 9 84 • AIS’s Management Insight Series gain a better understanding of LIS/DE differences revealed in the preliminary research.2 The methodology employed allowed for the delineation of within- and between-contract differences associated with LIS/DE and/or disability. Within-contract differences are differences that may exist between subgroups of enrollees in the same contract (e.g., if LIS/DE enrollees within a contract have a different mean or average performance on a measure than non-LIS/DE enrollees in the same contract). These differences may be favorable or unfavorable for LIS/DE/disabled beneficiaries and can be assessed separately from the overall level of performance for a contract. Between-contract differences in performance associated with LIS/DE/disability status (“between-contract disparities”) are the possible additional differences in performance between contracts associated with the contract’s proportion of LIS/DE/disabled enrollees that remain after considering within-contract disparities by LIS/DE/disability status. If LIS/DE/disabled beneficiaries are more or less likely than other beneficiaries to be enrolled in lower-quality contracts, then between-contract disparities may represent true differences between contracts in quality. Because of this possibility, betweencontract disparities may not be appropriate for adjustment due to the risk of masking true differences in quality. Adjusting for within-contract disparities is an approach aligned with the consensus reflected in the NQF report on sociodemographic adjustment, which states that, “…only the within-unit effects are adjusted for in a risk adjustment procedure because these are the ones that are related specifically to patient characteristics rather than differences across units” (National Quality Forum, 2014). Our research focused on measuring within-contract differences in performance for LIS/DE/disabled compared to non-LIS/DE/disabled beneficiaries. Our additional research findings are consistent with the preliminary results shared in the 2016 Final Call Letter. The research to date has provided scientific evidence that there exists a within-contract LIS/DE/disability effect for a subset of the Star Ratings measures. The size of the effect differs across measures and is not exclusively negative. CMS is firmly committed to building the foundation for a long-term solution that appropriately addresses the issue at hand and aligns with our policy goals. Any policy response must delineate the two distinct aspects of the LIS/DE and/or disability issue - quality and payment. The Star Rating Program focuses on accurately measuring the quality of care provided so any response must focus on enhancing the ability to measure actual quality differences among contracts. To address the LIS/DE/disability issue we must accurately address any sensitivity of the ratings to the SES of the beneficiaries enrolled in a contract at the basic building block of the rating system, the measure. CMS has encouraged the measure stewards to examine our findings and undertake an independent evaluation of the measures’ specifications to determine if measure re-specification is warranted. Concurrently, the payment response must focus on resource utilization and the predictive performance of the risk-adjustment models for 2 The research focused on a total of 16 clinical quality measures. A measure was excluded from analysis if the measure was already case-mix adjusted for SES (i.e., CAHPS and HOS measures), the focus of the measurement was not a beneficiary-level issue but rather a plan-level issue (e.g., appeals, call center, Part D price accuracy), the measure was scheduled to be retired or revised, or the measure was applicable to only Special Needs Plans (SNPs) (i.e., SNP Care Management, Care for Older Adults measures). 10 The Challenges of Pharmacy Star Ratings • 85 the unique cost patterns of beneficiaries in the community. CMS is also considering changes in the risk adjustment models for payment and issued a separate Request for Comments on October 28, 2015 to obtain feedback on potential revisions. We feel that these two approaches are complementary; holding contracts to a same quality standard is most appropriate when contracts are adequately resourced to provide the support their beneficiaries need to achieve good health outcomes. While the measure stewards are undertaking a comprehensive review of their measures used in the Star Ratings program, CMS is exploring two options for interim analytical adjustments to address the LIS/DE/disability effect: a Categorical Adjustment Index or Indirect Standardization. We believe each of the proposed methods, discussed in more detail below, align with the goals of: making adjustments that reflect the actual magnitude of the differences observed in the data; providing valid quality ratings to facilitate consumer choice; and providing incentives for MA and Part D quality improvement. In addition, we recognize the need for the options to be both transparent and feasible for the plans, as well as to maintain the integrity of the Star Ratings and the core of its methodology. Another issue we are examining is the manner to address the unique aspects of implementation of Medicare in Puerto Rico. Under statute, many of Department of Health and Human Services’ (HHS) programs, including Medicare and Medicaid, are implemented differently in Puerto Rico. In addition, Puerto Rico has a unique health care market with many low-income individuals in both Medicare and Medicaid and a complex legal history that affects the health care system in many ways. We are cognizant of the particular challenges in Puerto Rico and propose an additional analytical adjustment for contracts with a service area only in Puerto Rico to address the fact that the Part D low income subsidy (LIS) is not available there. The details of the possible interim analytical adjustments follow. While the interim policy responses to address the LIS/DE/disability effect are distinct analytical adjustments that are applied on the measure scores, they offer flexibility in their application. CMS is requesting feedback on each of the possible adjustments and the possible permutations of each option, or potential hybrid approaches. We welcome comments on the specific measures that should be adjusted3. The Categorical Adjustment Index The Categorical Adjustment Index is a factor that would be added or subtracted to a contract’s Overall and/or Summary Star Rating to adjust for the average within-contract disparity. Contracts would be categorized based on their percentages of LIS/DE and/or disabled 3 Our research focused on the following 16 measures: adult BMI assessment, rheumatoid arthritis management, breast cancer screening, controlling blood pressure, diabetes care – blood sugar controlled, diabetes care – eye exam, diabetes care – kidney disease monitoring, colorectal cancer screening, osteoporosis management in women who had a fracture, plan all-cause readmissions, annual flu vaccine, monitoring physical activity, reducing the risk of falling, medication adherence for diabetes medications, medication adherence for hypertension, and medication adherence for cholesterol. 11 86 • AIS’s Management Insight Series beneficiaries, and the Categorical Adjustment Index value would be the same for all contracts within each category. The Categorical Adjustment Index value is the star adjustment for contracts by category. The Categorical Adjustment Index values would be computed by comparing the mean Overall and/or Summary contract Star Rating derived from measure scores that are adjusted for LIS/DE and/or disability status to the mean Star Rating derived under the traditional methodology. The adjusted measure scores would be derived from regression models of beneficiary-level measure scores that adjust for the average within-contract difference in measure scores by LIS/DE and/or disability status for MA or PDP contracts without masking potential differences in quality across contracts. This approach is equivalent to case-mix adjustment or patient-mix adjustment in a patient-level linear regression model with contract intercepts and beneficiarylevel indicators of LIS/DE/disability status, similar to the approach currently used to adjust CAHPS patient experience measures. Measure scores are adjusted first and then the adjusted measure score is converted to a measure-level Star Rating using the measure thresholds for the given Star Ratings year. The Categorical Adjustment Index is derived via four steps: (1) contracts are divided into an initial set of categories based on some combination of LIS/DE/disability (this might be 10 categories corresponding to the 10 deciles of LIS/DE or the 16 combinations of LIS/DE quartile and disability quartile, or some other grouping); (2) the mean difference between the Adjusted Overall or Summary Star Rating and the Unadjusted Star Rating is computed within each of the initial categories; (3) the mean differences of the initial categories in step (2) are examined and combined into final adjustment groups such that initial categories with similar means form the groups; and (4) the mean star Adjustment is computed within each of the final adjustment groups - this is the Categorical Adjustment Index. The Star Rating measure’s specification is unchanged. The Categorical Adjustment Index is applied external to the specification and is applied at the Overall and/or Summary Star Rating. Each contract within a given final LIS/DE/disability group receives the same adjustment to its Overall and/or Summary Star Ratings. The index would be determined using the current year’s ratings. For the 2017 Star Ratings (measurement year 2015), the Categorical Adjustment Index values would be based on the observed values for the 2017 Star Ratings year using data from contracts that meet current reporting requirements. The table below depicts an example of the overall method and one possible scenario for groupings employed to determine the Categorical Adjustment Index. In preliminary work, these eight possible final categories of LIS/DE/disability were derived by combining 16 initial combinations of LIS/DE and disability quartiles in the manner described above. 12 The Challenges of Pharmacy Star Ratings • 87 Mean Overall Grouping of contract, based on % Unadjusted Star LIS/DE and % Disabled Rating LIS/DE 1st quartile & disability 1st quartile LIS/DE 1st quartile & disability 2nd4th quartiles LIS/DE 2nd quartile & disability 1st quartile LIS/DE 2nd quartile & disability 2nd-4th quartile LIS/DE 3rd quartile & disability 1st2nd quartiles LIS/DE 3rd quartile & disability 3rd4th quartiles LIS/DE 4th quartile & disability 1st3rd quartiles LIS/DE 4th quartile & disability 4th quartile Mean Adjusted Star Rating Mean Difference in Star Rating (Adjusted - Unadjusted) Indirect Standardization Indirect standardization, an alternative proposal, would be applied to a subset of the individual Star Ratings measure scores; measure stars are not used because that would incorrectly assume assignments of measure stars are linear in the underlying measure and thus, lead to measurement error. The focus of the adjustment is the within-contract LIS/DE and/or disability status difference while allowing for the existence of true differences in quality by contract. The standardization would employ the current year’s ratings. An expected measure score would be calculated using the percent of LIS/DE and/or disabled and non-LIS/DE and/or non-disabled beneficiaries per measure multiplied by the adjusted mean national performance for each subgroup. As above, this method could use a variety of grouping methodologies to determine appropriate subgroups. These contract-adjusted LIS/DE and/or disabled and non-LIS/DE and/or non-disabled national performance means would be calculated such that they differed by the national performance mean within-contract LIS/DE versus non-LIS/DE difference (and/or disabled versus non-disabled difference). The expected measure score is a weighted average based on the composition of the enrollees of the contract and the national adjusted mean measure score for the subgroups of interest (e.g., LIS/DE and/or disabled versus non-LIS/DE and/or non-disabled beneficiaries). For example: Using indirect standardization at the measure score level, the expected measure score for diabetes control would use the adjusted national means for performance on diabetes control for the subgroups of interest (LIS/DE and non-LIS/DE and/or disabled and non-disabled beneficiaries). To simplify the example, the focus of the standardization is disability status, but the same method could be expanded to more than two subgroups and include standardization for LIS/DE and non-LIS/DE. If the adjusted national mean performance on diabetes control is 50% for disabled beneficiaries and 60% for non-disabled beneficiaries, a contract that served only disabled beneficiaries would be expected to perform at 50% and a contract that served only non-disabled beneficiaries would be expected to perform at 60%. Contracts that had both disabled and non-disabled beneficiaries would be expected to perform at a specific level 13 88 • AIS’s Management Insight Series between 50% and 60%, as a weighted average based on the composition of the enrollees in the contract for that measure. Each contract would be judged against its expected performance. Next, the ratio of the observed-to-expected measure score would be calculated. (The observed value is the measure score based on the rating year’s data). The observed-to-expected ratio would equal one for contracts that performed at the level expected given their percentages of LIS/DE and/or disabled beneficiaries and indicates average performance. In contrast, a ratio less than one would indicate lower observed performance than expected given the contract percentage of LIS/DE and/or disabled beneficiaries; similarly, ratios greater than one would indicate better than expected performance. The adjusted measure score would then be calculated. The adjusted measure score is the product of the observed-to-expected ratio for a contract and the adjusted national mean performance for all Medicare beneficiaries. Two contracts with identical observed measure performance but different expected measure performance would receive different adjusted measure scores. The adjusted measure score would be converted to a measure-level Star Rating using the measure thresholds for the given Star Ratings year. The adjusted measure level stars would then be used in the determination of the Overall and/or Summary Star Rating. The tables below summarize for a single measure the information needed and the overall methodology to determine the adjusted measure score. The first table provides the national values needed to indirectly standardize a measure score. The second table and corresponding information below the table provide a high-level overview of the process of indirect standardization to determine the adjusted measure score. For ease of presentation, the tables focus on an adjustment for the subgroups of LIS/DE and non-LIS/DE, but the same method would be applied for the distinct groupings of LIS/DE, non-LIS/DE, disabled, and non-disabled combinations. Measure-specific National Adjusted Values Contract 1 Proportion of LIS/DE Beneficiaries D Adjusted Pass Rate for LIS/DE Beneficiaries A Proportion of non-LIS/DE Beneficiaries E Adjusted Pass Rate for non-LIS/DE Beneficiaries B Overall Pass Rate (Observed) F Expected Rate (Based on composition of plan) G Adjusted Overall National Mean C Ratio of Observed to Expected H Adjusted Measure Score I The contract’s proportion of LIS/DE and non-LIS/DE per measure (values D and E) would be multiplied by the corresponding national adjusted rates for LIS/DE and non-LIS/DE (values A and B) to determine the expected performance for the measure score (value G). The formula applied would be as follows: Expected Rate = A×D + B×E. 14 The Challenges of Pharmacy Star Ratings • 89 Next, the ratio of the observed (actual) would be calculated (Ratio = F/H). The ratio calculated would be multiplied by the adjusted overall national mean (value C) to determine the adjusted measure score (value I). The adjusted measure score would be converted to a measure-level Star Rating, and the traditional Star Ratings methodology would ensue to determine the adjusted Overall and/or Summary Star Rating for the year. Additional response to address lack of an LIS indicator for enrollees in Puerto Rico Notably, Puerto Rican beneficiaries are not eligible for LIS, which is an important element of both potential methodologies. (Beneficiaries in the 50 states are eligible for LIS in the mainland if their income is less than 150% of the Federal Poverty Level). To make the proposed analytical adjustments equitable, CMS is considering an additional adjustment for contracts in Puerto Rico to identify beneficiaries in Puerto Rico’s contracts whose incomes would result in an LIS designation in the mainland. The contract-level modified LIS/DE proportion for Puerto Rico would be developed from two sources of information: (1) the overall proportion of beneficiaries in Puerto Rico with incomes less than 150% of the FPL and (2) each contract’s proportion of DE beneficiaries. A linear regression model would be developed to predict the percentage of LIS in a contract using the percentage of DE using all MA contracts except those in Puerto Rico. Preliminary evidence suggests this model has very high accuracy in predicting contract-level LIS from contract-level DE in the 50 states and the District of Columbia, even when restricted to lower-income subsets of states. This model would then be adjusted for use in Puerto Rican contracts (i.e., contracts with a service area only in Puerto Rico) using Puerto Rico’s mean percentage of DE and mean of LIS (using the percentage of Puerto Rico’s population with incomes less than 150% of the FPL). Using the model developed, each contract’s proportion of DE beneficiaries in Puerto Rico would have a corresponding proportion of LIS to create a contract-level measure of LIS/DE percentage to be used in the Categorical Adjustment Index or Indirect Standardization adjustments. We welcome comments on this proposed approach to approximate the percentage of LIS by contract in Puerto Rico, or other possible suggestions of ways to estimate this percentage. We also considered options to address the unique challenges that Puerto Rican contracts face in improving medication adherence. It has been shown that beneficiaries’ out-of-pocket costs may adversely affect medication adherence, which presents an additional barrier for Puerto Rican contracts serving beneficiaries whose incomes would result in an LIS designation in the mainland. One option we considered was to reduce the weights of the three Part D Medication Adherence measures for Puerto Rican contracts. A prior proposal in the 2015 draft Call Letter to reduce the weight of the three Part D Medication Adherence measures to 1.5 as access measures for all Part D sponsors was not supported by the majority of commenters. MA plans and PDP sponsors expressed concerns that this type of change would be contrary to efforts to encourage coordination of care, as well as decrease performance in other quality measures. We simulated this change to assess the potential impact. The highest ratings for substantially all of the Puerto Rican contracts remained unchanged based on simulation of the 2015 Star Ratings data. We commend the Puerto Rican contracts on their improved performance overall across the 2016 Star Ratings and the Part D Medication Adherence 15 90 • AIS’s Management Insight Series measures in particular. We found no changes to the highest ratings for Puerto Rican contracts when we simulated reducing the weight of the adherence measures using 2016 Star Ratings data. We welcome other proposals to account for the barriers for Puerto Rican contracts to improve medication adherence. Summary The potential interim analytical adjustments to address the LIS/DE/disability effect, the Categorical Adjustment Index and Indirect Standardization, share common features. These adjustments were developed while focusing on measuring within-contract differences, while allowing for the existence of true differences in quality by contract. In addition, both methods rely on an adjustment external to the measure specification. The methods result in an adjustment to the measure scores for a subset of measures. The adjusted measure scores are converted to a measure-level Star Rating employing the original measure thresholds for the Star Ratings year before adjustment. For the improvement measure we would use unadjusted scores. Both analytical adjustments under consideration would require additional processing of the data by CMS. Both methods are applied to the measure score and employ a measure-specific proportion of LIS/DE/disabled which is not constant across measures given measure specification exclusions. Further, both adjustments are based on the current year’s data and thus, values needed for the application of the methods (Categorical Adjustment Index values and the adjusted overall national mean per measure and for each subgroup for Indirect Standardization) would not be available until after the submission of the data for the year. Given the additional data processing steps, the analytical adjustments may result in a compressed timeframe for Part C and D contracts’ review process of the ratings. The Categorical Adjustment Index values would be incorporated in the Star Ratings Technical Specifications and would be a set of adjustment factors to the overall and/or summary ratings. For the Indirect Standardization method, contracts would need to review a larger volume of intermediate calculations for each measure during the preview periods. Both of these methods adjust for within-contract differences. Based on our research, we would not expect either approach to generate major adjustments in the overall Star Ratings, though any adjustment may be significant for individual contracts. We recognize adjustments that account for the full between-contract differences (e.g., summary level indirect standardization) would make a considerably larger difference overall, but these approaches could risk over-crediting poor quality contracts. Implementing such an approach coupled with adjustments both upward (for contracts with a high proportion of duals) and downward (for contracts with a lesser proportion of duals) would likely result in larger swings both positively and negatively for contracts. CMS is interested in understanding comments about this approach, including its impact on accurate measurement of quality. CMS also notes that ASPE will continue to explore options that could be implemented in future years. 16 The Challenges of Pharmacy Star Ratings • 91 E. 2017 CMS Display Measures Display measures on www.cms.gov are not part of the Star Ratings. These may include measures that have been transitioned from the Star Ratings, new measures that are being tested before inclusion into the Star Ratings, or measures displayed for informational purposes. Similar to the 2016 display page, organizations and sponsors have the opportunity to preview their data for the display measures prior to release on CMS’ website. Data for measures moved to the display page will continue to be collected and monitored, and poor scores on display measures are subject to compliance actions by CMS. It is expected that all 2016 display measures will continue to be shown on www.cms.gov. CMS will continue to provide advance notice regarding measures considered for implementation as future Star Ratings measures. Other display measures may be provided as information only. 1. Timely Receipt of Case Files for Appeals (Part D) & Timely Effectuation of Appeals (Part D). For the 2016 display measures, the data time frame for both measures was 01/01/2015 – 06/30/2015. CMS proposes to change the data time frame from the first six months of the current year to January 1 – December 31 of the previous year. For example, the 2017 display measures would be based on IRE data from January 1, 2015-December 31, 2015. This change will allow the appeal display measures to match the same timeframe used for the Part D Appeal Star Ratings measures. The following are a number of new measures for the 2017 display page. 2. Medication Reconciliation Post Discharge (Part C). The Medication Reconciliation PostDischarge (MRP) measure assesses the percentage of discharges from acute or non-acute inpatient facilities for members 66 years of age and older for whom medications were reconciled within 30 days of discharge. NCQA made two changes: 1) expanded the coverage on this measure from Medicare Special Needs Plans only to all of MA; and 2) expanded the age range to members 18 years and older. Both of these changes for HEDIS 2016 are seen as an important step to measure the quality of care coordination post-discharge for MA beneficiaries as well as ensuring patient safety. CMS is planning to include this measure on the 2017 display page and is planning to include it in the 2018 Star Ratings. 3. Hospitalizations for Potentially Preventable Complications (Part C). NCQA added to HEDIS 2016 a risk-adjusted measure of hospitalization for ambulatory care sensitive conditions based on the NQF-endorsed Prevention Quality Indicators (PQI), developed by AHRQ. This measure assesses the rate of hospitalization for complications of chronic and acute ambulatory care sensitive conditions. The intent of the measure is to assess the quality of ambulatory care—including coordination of that care—to prevent the complications of chronic and acute conditions that result in hospitalization. CMS is planning to include this measure on the 2017 display page and is planning to include it in the 2018 Star Ratings. 4. Statin Therapy for Patients with Cardiovascular Disease (Part C). NCQA has added two sets of statin therapy measures to HEDIS aligned with the 2013 ACC/AHA blood cholesterol guidelines. These measures are focused on two of the major statin benefit groups described in 17 92 • AIS’s Management Insight Series the guidelines: patients with clinical atherosclerotic cardiovascular disease and patients with diabetes. Since some of these HEDIS measures overlap with the measures developed by the PQA, CMS is planning to include only one of the HEDIS measures on the 2017 display page and is planning to include this measure in the 2018 Star Ratings. This measure focuses on statin therapy for patients with cardiovascular disease. It is the percentage of males 21 to 75 years of age and females 40 to 75 years of age who were identified as having clinical atherosclerotic cardiovascular disease and were dispensed at least one high or moderate-intensity statin medication during the measurement year. 5. Asthma Measures (Part C). NCQA has expanded their asthma measures to include older adults. HEDIS 2016 includes two measures for older adults. Medication Management for People with Asthma is the percentage of members 5 to 85 years of age who were identified as having persistent asthma and were dispensed appropriate medications that they remained on during the treatment period (i.e., first prescription date through end of measurement year). The Asthma Medication Ratio is the percentage of members who were identified as having persistent asthma and had a ratio of controller medications to total asthma medications of 0.50 or greater during the measurement year. CMS is planning to include these on the 2017 display page and will consider these for inclusion in Star Ratings for future years. 6. Statin Use in Persons with Diabetes (SUPD) (Part D). This new PQA-endorsed measure, Statin Use in Persons with Diabetes (SUPD), calculates the percentage of patients between 40 and 75 years old who received at least two diabetes medication fills and also received a statin medication during the measurement period. Beneficiaries in hospice according to the Enrollment Database (EDB) will be excluded from the denominator of the SUPD measure for the entire year. Part D sponsors have received year of service 2015 SUPD measure reports on a monthly basis through the Patient Safety Analysis website, and we will add the SUPD measure to the 2017 display page (using 2015 data). We propose adding the SUPD measure to the 2018 Star Ratings (using 2016 data). Forecasting to 2018 and Beyond The following describes changes to existing measures and potential new measures. CMS will also monitor any additional measures developed by NCQA or PQA for potential incorporation into the Star Ratings. F. New Measures: See section E above which describes a number of new measures under consideration for the 2018 Star Ratings that will be reported as 2017 display measures. The following are additional measures under consideration for the Star Ratings or display measures for 2018 and beyond. 1. Care Coordination Measures (Part C). Effective care coordination contributes to improved health outcomes. CMS believes that 5-star contracts perform well on our Star Ratings measures because they understand how to effectively coordinate care for their enrollees. Our assumption about plans, however, is based largely on anecdote and discussions with high performing plans, as well as on data we collect from CAHPS surveys which reflect enrollees’ experiences with the care they receive. 18 The Challenges of Pharmacy Star Ratings • 93 CMS is working to expand efforts in this area. To identify potential new care coordination measures, CMS is utilizing experts to conduct targeted research, extensive literature reviews, and data analysis, and to engage in discussions with expert panels and high performing plans. As part of this effort, we are considering various data sources; whether the measures should be focused on subgroups of MA enrollees or all MA enrollees; the activities that best represent care coordination such as ensuring seamless transitions across settings, appropriate follow up after inpatient and emergency department visits, communication across providers, and comprehensive assessments; and the relationship between the plan and provider in care coordination activities. NCQA, using administrative and medical record data, will begin testing the following proposed measures using 2015 data: primary care provider (PCP) notification of inpatient admissions, summary of care record in PCP chart, follow-up with PCP/specialist following hospital discharge or emergency department visit, and in the ambulatory setting whether there is a comprehensive assessment performed and documented by the PCP/specialist and whether there is a specialist visit summary in the PCP chart. Additionally, CMS has recently awarded another contract to develop care coordination measures using administrative data, including MA encounter data and Part D data. CMS welcomes comments on measures that could be developed using MA encounter data. We will continue to provide updates to the industry as this work progresses. Measures developed and tested may be considered for future inclusion on the display page and in Star Ratings. 2. Depression Measures (Part C). NCQA has adapted a provider-level depression outcome measure developed by Minnesota Community Measurement for use in HEDIS. Depression Remission or Response in Adolescents and Adults (DRR) uses a patient-reported outcome measure, the PHQ-9 tool, to assess whether patients with depression have achieved remission or have an improvement in their symptoms. The measure assesses the percentage of individuals age 12 and older with depression and an elevated PHQ-9 score (greater than 9) who achieve a PHQ-9 score of less than 5 at six months or have a 50% reduction in their PHQ9 score. This measure also uses a new data collection methodology for HEDIS, relying on data coming from electronic clinical data systems (e.g., EHRs, clinical registries, case management records). If approved, the new measure would be published in HEDIS 2017. 3. Appropriate Pain Management (Part C). NCQA is exploring opportunities to develop a new measure(s) focusing on appropriate pain management. The intent is to assess the quality of pain management and treatment. There is no definite timeline established for the development of this measure. 4. Use of Opioids from Multiple Providers or at High Dosage in Persons without Cancer (Part D). In the 2016 Call Letter, we noted that three opioid overutilization measures were in development by the PQA. We further stated that if these measures were endorsed by the PQA prior to the 2017 bid deadline in June 2016 that we may adopt them as future display measures or alternatively use them in the Overutilization Monitoring System (OMS). The measures were endorsed by the PQA in May 2015. PQA’s three opioid measures examine multi-provider, high dosage opioid use among individuals 18 years and older without cancer and not in hospice care. 19 94 • AIS’s Management Insight Series Measure 1 (Opioid High Dosage): The proportion (XX out of 1,000) of individuals without cancer or hospice receiving prescriptions for opioids with a daily dosage greater than 120mg morphine equivalent dose (MED) for 90 consecutive days or longer. Measure 2 (Multiple Prescribers and Multiple Pharmacies): The proportion (XX out of 1,000) of individuals without cancer or hospice receiving prescriptions for opioids from four (4) or more prescribers AND four (4) or more pharmacies. Measure 3 (Multi-Provider, High Dosage): The proportion (XX out of 1,000) of individuals without cancer or hospice receiving prescriptions for opioids with a daily dosage greater than 120mg morphine equivalent dose (MED) for 90 consecutive days or longer, AND who received opioid prescriptions from four (4) or more prescribers AND four (4) or more pharmacies. We tested the measures using the PQA specifications. We will develop new patient safety opioid overutilization measure reports (beginning with 2016 dates of service) to provide to Part D sponsors on a monthly basis through the Patient Safety Analysis website, similar to the other patient safety measures. The website also includes the OMS. The reports will allow sponsors to track their performance over time and allow for contract level trending and outlier analyses. We will also add these three measures to the 2018 Part D display page (using 2016 data). We do not recommend adding these measures to the Star Ratings at this time due to concerns (1) about the current lack of consensus clinical guidelines for the use of opioids to treat chronic, non-cancer pain and potential exceptions due to medical necessity and (2) pending additional analysis on diagnosis data sources, such as newly available encounter data for Medicare Part C and resolving timing issues of RAPS file updates, which are used to identify exclusions for certain cancer conditions. Additionally, NCQA is adapting the three opioid overuse measures developed by the PQA for potential use in HEDIS. 5. Antipsychotic Use in Persons with Dementia (APD) (Part D). CMS has been particularly concerned with the unnecessary use of antipsychotic drugs in nursing homes and, as a result, has pursued strategies to increase awareness of antipsychotic use in long term care settings. In 2013, we began to calculate a general atypical antipsychotic utilization rate, called Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes, for inclusion in the Part D display measures. The average rates decreased from approximately 24.0% in 2011 to 21.4% in 2013. There continues to be increased attention on this important issue. The United States Government Accountability Office (GAO) released a report4 in January 2015 describing the inappropriate use of antipsychotics in Part D beneficiaries with dementia, in both community (i.e., outside of nursing homes) and long-stay nursing home residents during 2012, with 4 Antipsychotic Drug Use: HHS Has Initiatives to Reduce Use among Older Adults in Nursing Homes, but Should Expand Efforts to Other Settings. http://www.gao.gov/products/GAO-15-211. GAO-15-211: Published: Jan 30, 2015. Publicly Released: March 2, 2015 20 The Challenges of Pharmacy Star Ratings • 95 recommendations for CMS to address this problem. The GAO conducted this study due to concerns raised regarding the use of antipsychotic drugs to address the behavioral symptoms associated with dementia, the FDA’s boxed warning that these drugs may cause an increased risk of death when used by older adults with dementia, and because the drugs are not approved for this use. In addition, the PQA endorsed the measure, Antipsychotic Use in Persons with Dementia (APD). This provides CMS with a new measure developed through a consensus process to monitor the inappropriate use of antipsychotics in both the nursing home and community settings across Medicare Part D plans. We tested this measure based on the PQA specifications. The APD measure rate was calculated for all contracts, MA-PDs, PDPs, and at the individual contract-level for all beneficiaries, community-only residents (never a nursing home resident), and both short-term and long-term nursing home residents that met the inclusion and exclusion criteria. Beneficiaries were identified as long-stay nursing home residents if they had stays greater than 100 cumulative days in a nursing home during the year based data in the Long Term Care Minimum Data Set (MDS). Each beneficiary was counted in only one category for the entire measurement period within a contract, not considered separately for time spent in different settings (e.g., a beneficiary who experienced both short-term and long-term nursing home stays was included only in the long-term category). To identify the numerator and denominator populations, we used diagnosis data obtained from inpatient (IP), outpatient (OP), and carrier claims from the Common Working File (CWF) and RxHCCs from the RAPS. OP and Carrier claims are available for PDP contracts only. We also adjusted rates based on the number of months beneficiaries are enrolled in each Part D contract (i.e., member-years adjustment). We conducted reliability testing using mixed effect logistic regression with varying intercept. The testing results indicate that the rate variations at the contract level are statistically significant, providing evidence that the measure is reliable. We will develop new patient safety APD measure reports to provide to Part D sponsors on a monthly basis through the Patient Safety Analysis website beginning with year of service 2016. We also recommend adding the overall APD measure plus breakout rates for community-only residents, short-term nursing home residents, and long-term nursing home stay residents to the 2018 Part D display measure set (using 2016 data) to continue to draw attention to the inappropriate use of antipsychotics in persons with dementia without an appropriate mental health diagnosis in both the community and nursing home settings. The APD measure will replace the Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes display measure. However, we do not recommend adding this measure to the Star Ratings pending additional research on diagnosis data sources, such as newly available encounter data for Medicare Part C and resolving timing issues of RAPS file updates. 21 96 • AIS’s Management Insight Series G. Changes to Existing Star Ratings and Display Measures and Potential Future Changes: 1. Colorectal Cancer Screening (Part C Star Rating). The Colorectal Cancer Screening (COL) measure assesses the percentage of adults age 50-75 years of age who had appropriate screening for colorectal cancer. This measure is based on the U.S. Preventative Services Task Force (USPSTF) guideline on colorectal cancer screening in adults age 50-75. NCQA is monitoring updates to the guideline as the USPSTF has released a draft recommendation statement for public comment. We have discussed the guideline timing with the Agency for Healthcare Research and Quality team in charge of the USPSTF process, and they note the final release is not likely to occur until late 2016. NCQA will consider revisions to the COL measure once the USPSTF final recommendation statement is published. 2. Fall Risk Management (Part C Star Rating). The Fall Risk Management (FRM) measure, collected through the Health Outcomes Survey, consists of the following two indicators: 1) Discussing Fall Risk assesses the percentage of Medicare members 75 years of age and older or 65-74 years of age with a balance or walking problem or fall in the past 12 months who discussed falls or problems with balance or walking with their current practitioner; and 2) Managing Fall Risk assesses the percentage of Medicare members 65 years of age and older who had a fall or had problems with balance or walking in the past 12 months and received fall risk intervention from their current practitioner defined as suggesting use of a cane or walker, a vision or hearing test, physical therapy or exercise, or taking of a postural blood pressure. NCQA is currently re-evaluating this measure to align with the most current U.S. Preventive Services Task Force (USPSTF) guidelines. NCQA is proposing to 1) revise the denominator in the Discussing Fall Risk indicator to include all Medicare members age 65 and older and 2) revise the numerator for the Managing Fall Risk indicator to include use of vitamin D and remove vision or hearing test and taking of postural blood pressure. These proposed changes, if approved, would be published in HEDIS 2017 or HEDIS 2018. 3. Pneumococcal Vaccination Status for Older Adults (Part C Display). The Pneumococcal Vaccination Status for Older Adults (PNU) measure, collected through the Medicare CAHPS survey, assesses the percentage of Medicare members 65 years of age and older who have ever received a pneumococcal vaccination. In 2014, The Advisory Committee on Immunization Practices (ACIP) released new recommendations that all adults 65 years of age and older should receive sequential administration of both PCV13 and PPSV23. NCQA is considering changes to the measure to align with the most current guidelines. There is no definite timeline established for these changes. This measure is on the CMS display page. 4. CAHPS measures (Part C & D). The current MA & PDP CAHPS Survey includes the core CAHPS 4.0 Health Plan Survey. CMS conducted an experiment in 2015 to understand how CAHPS measures differ between 4.0 and 5.0, and based on the results we propose to update the survey for future years to reflect AHRQ’s CAHPS 5.0 Health Plan Survey. The findings from the experiment suggest that these changes are associated with a small increase in scores for several evaluative MA measures. These small increases did not significantly differ across contracts. Since there are no longer fixed thresholds for Star Ratings and they are based on the actual distribution of scores, there should be no shifts in Star Ratings due to transition to the version 5.0 instrument compared to what would have been the case with 4.0. Every 22 The Challenges of Pharmacy Star Ratings • 97 contract would have the same expected Star Rating whether version 4.0 or 5.0 is used, and the correlation between this year’s Star Ratings and next year’s Star Ratings should be the same regardless of whether 4.0 or 5.0 is used next year. The 5.0 update applies recent improvements in survey design that resulted from development and testing of the Clinician & Group Surveys. The 5.0 version of the CAHPS Health Plan Survey incorporates some minor changes into the wording of core items, and a change in the placement of one core item that also resulted in the deletion of a screener item. The following are the changes in the 5.0 version of the Health Plan Surveys: The items about access to urgent and non-urgent appointment items were modified to ask respondents if they were able to get an appointment as soon as they needed, as opposed to as soon as they thought they needed. Non-urgent appointments are described as a check-up or routine care rather than health care. In addition, the phrase, “…not counting the times you needed care right away” was deleted from these questions. These revisions simplify the items and make them consistent with questions in other CAHPS surveys. The item about how often it was easy to get appointments with specialists was revised to ask respondents if they got an appointment to see a specialist as soon as they needed. This revision makes the item consistent with other CAHPS items that ask about access to care. The item about how often it was easy to get care, tests, or treatment was moved from the Your Health Plan section to the Your Health Care in the Last 6 Months section, because respondents had difficulty attributing this item to the health plan. The screener item about getting care, tests, or treatment through the health plan was deleted because the subsequent question was moved to an earlier section of the survey and no longer required a screener. These changes would take effect for the 2017 CAHPS survey administration (2018 Star Ratings) based on OMB approval. Since we are modifying question wording, we propose the following standard for deciding that a specification change has occurred for a CAHPS measure for the purposes of excluding it from the improvement measure calculation: (1) at least one item within the measure changed in wording, had a wording change in its screener, or had a wording change in the immediately preceding item, and (2) the measure score in version 5.0 was significantly different from the measure score in version 4.0 in the 5.0 experiment. Three MA measures met this standard: Getting Care Quickly, Customer Service, and Care Coordination; thus, these three measures would be excluded from the Part C improvement measure for the 2018 Star Ratings. We are also considering changing the sampling for CAHPS when a contract is listed in HPMS as a consolidation, merger, or novation between July of the prior year and January of the current year when the CAHPS sample is drawn. We are considering changing the sampling frame for 23 98 • AIS’s Management Insight Series the surviving contract to include the enrollees for all members of all contracts involved if the two or more contracts merging, consolidating or novating are under the same parent organization. This would go into effect for the 2017 sample draw. 5. Medication Adherence for Hypertension (RAS Antagonists) (Part D Star Rating). Based on PQA specification change, the measure will exclude from the denominator those patients with one or more claims for sacubitril/valsartan. 6. MPF Price Accuracy (Part D Star Rating). As stated in the 2016 Call Letter, CMS is considering a few updates to this measure for the 2018 Star Ratings. The first proposed change is related to the method by which claims are excluded from the measure. Currently, the measure is limited to 30-day claims filled at pharmacies reported by sponsors as retail only or retail and limited access only in their MPF Pharmacy Cost files. That is, claims filled for near 30 days supplies, or claims filled for 60 and 90 days supplies are excluded. Additionally, claims filled by retail pharmacies that are also long term care, mail order, or home infusion pharmacies are excluded. These restrictions result in the exclusion of many PDEs, thus potentially biasing the reliability of the measure. We propose to include claims with 28-34 days supplies, as we believe it would be appropriate to compare their PDE costs to MPF’s fixed display of 1 month pricing. We also propose to include 60-62 and 90-93 day supplies. Beginning with CY2015 MPF submissions, plans must provide brand and generic dispensing fees for 60 and 90 day supply claims in the Pharmacy Cost file. CMS can use these data, along with 60 and 90 day supply Pricing File data, to compare MPF and PDE costs. While the majority of claims are for a 30 day supply, we found that claims with a 90 day supply account for almost one-fifth of available PDE data, thus allowing for a more comprehensive evaluation of PDE claims. Additionally, we propose to use the PDE-reported Pharmacy Service Type code in conjunction with the MPF Pharmacy Cost data to identify retail claims. CMS began requiring pharmacies to populate the Pharmacy Service Type field on all PDEs at the end of February 2013. We recommend expanding the retail claims identification process to include all PDEs that are from retail pharmacies according to the Pharmacy Cost data and have a Pharmacy Service Type of either Community/Retail or Managed Care Organization (MCO). Although some sponsors cited concern about the accuracy of these data as reported by pharmacists, Part D sponsors are ultimately responsible for the accuracy of their submitted PDE to CMS. According to PDE requirements, CMS expects “…sponsors and their network pharmacies to develop and implement controls to improve the accuracy of this information during 2013…” This methodology change would increase the number of PDEs eligible for inclusion in the Price Accuracy Scores while continuing to identify only retail claims. We are also considering changes to the methodology by which price accuracy is calculated. Because the current methodology measures the magnitude of a contract’s overpricing relative to its overall PDE costs, the Price Accuracy Scores do not reflect the frequency of accurate price reporting, and can be significantly impacted by high cost PDEs. As a result, contracts with divergent accurate price reporting and/or consistency can receive the same Price Accuracy Score. CMS is interested in modifying the methodology to also factor in how often 24 The Challenges of Pharmacy Star Ratings • 99 PDE costs exceeded MPF costs. The frequency of inaccuracy by a contract would be the percent of claims where PDE cost is greater than MPF cost. The numerator is the number of claims where PDE cost is greater than MPF cost, and the denominator is the total number of claims. This ratio is then subtracted from 1 and multiplied by 100 to calculate the Claim Percentage Score, with 100 as the best possible score and 0 as the worst possible score. The contract’s accuracy score would be a composite of the Price Accuracy Score and the Claim Percentage Score. By capturing the frequency of inaccuracy as well as the magnitude, the measure would better depict the reliability of a contract’s MPF advertised prices. CMS is aware that while the MPF display is updated every two weeks, real time pricing, at the point of sale, can change as often as every day. Some sponsors have expressed concern that in order to perform well in the Price Accuracy measure, they cannot offer lower prices at point of sale in real time than the prices are displayed on MPF. We would note that PDEs priced lower than MPF displayed pricing do not lower a contract’s score in this measure. CMS’ simulation of this proposal found little change in the range of contracts’ accuracy scores. Other options we explored include measuring the magnitude of inaccuracy as a percentage cost difference, instead of the current measure’s use of absolute cost difference. Testing however found this method may overstate small differences between PDE and MPF costs for low-cost claims. For example, when using percentage cost differences, a claim with a $2.00 PDE cost and a $1.00 MPF cost would be considered equally overpriced as a claim with a $200.00 PDE cost and a $100.00 MPF cost. As noted in the 2016 Call Letter, we propose that these changes are implemented for the 2018 Star Ratings (using 2016 PDE and MPF data). We believe the proposed changes will greatly improve the Price Accuracy Scores, making them a more comprehensive assessment of contracts’ price reporting for Part D beneficiaries. 7. Drug-Drug Interactions (DDI) (Part D Display). The PQA-endorsed DDI measure is currently a Part D display measure. This measure is defined as the percent of Medicare Part D beneficiaries who received a prescription for a target medication during the measurement period and who were dispensed a prescription for a contraindicated medication with or subsequent to the initial prescription. The PQA has conducted an extensive review of the drug-drug pairs included in the DDI measure. They engaged a DDI expert panel convened by the University of Arizona on PQA's behalf, which completed the review, including a comparison to the DDI list developed for the Office of the National Coordinator for Health Information Technology (ONC). Next, the Expert Panel's recommendations will be reviewed by the PQA’s Measure Update Panel for consideration by the PQA’s Quality Metrics Expert Panel (QMEP). We anticipate that there will be extensive changes to the DDI measure specifications. We will closely monitor any updates to this measure, test updated specification when available, and propose changes in the future for the Part D display measure and patient safety reporting. 8. Center for Medicare and Medicaid Innovation Model Tests. The CMS Center for Medicare and Medicaid Innovation has announced the Medicare Advantage Value-Based Insurance Design (MA-VBID) and the Part D Enhanced MTM model tests. Beginning January 1, 2017, in a 25 100 • AIS’s Management Insight Series limited number of states, CMS will give MA only, MA-PD or Part D plans participating in these tests additional flexibilities intended to improve the quality of care and reduce costs in the Medicare Advantage or Part D programs, respectively. More information about the specific flexibilities offered in these model tests is available at https://innovation.cms.gov/initiatives/HPI. Some stakeholders have expressed to CMS the potential for the improvements in quality in these models to favorably influence the Star Ratings of contracts with participating plans, as compared to the performance of those ineligible to participate. The goal is to not penalize participants or non-participants. As the model tests are implemented, we will closely monitor performance trends of participating plans across individual measures and determine if any changes are warranted. We welcome any comments on how to address any potential differences in performance between participating and non-participating plans. The Part D plans participating in the Part D Enhanced MTM model test will be waived from the MTM requirements under Section 1860D–4(c)(2) and 42 CFR 423.153(d) and the Part D reporting requirements for MTM. However, Part D sponsors will not be waived from establishing MTM programs in compliance with current requirements and reporting data for the remaining plans under each Part D contract. Therefore, the MTM Program CMR Completion Rates will be calculated using available plan-reported data from the remaining plans under the Part D contract. H. Measurement and Methodological Enhancements CMS is committed to continuing to improve the Part C and D Star Ratings by identifying new measures and methodological enhancements. Feedback or recommendations can help CMS’ continuing analyses, as well as our collaboration with measurement development entities such as NCQA and PQA. We welcome comments and input on issues not described in earlier sections. 26 The Challenges of Pharmacy Star Ratings • 101 Appendix Improvement measures (Part C & D): Part C or D C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C D D D D D D D D D D D D D D Measure Breast Cancer Screening Colorectal Cancer Screening Annual Flu Vaccine Improving or Maintaining Physical Health Improving or Maintaining Mental Health Monitoring Physical Activity Adult BMI Assessment Special Needs Plan (SNP) Care Management Care for Older Adults – Medication Review Care for Older Adults – Functional Status Assessment Care for Older Adults – Pain Assessment Osteoporosis Management in Women who had a Fracture Diabetes Care – Eye Exam Diabetes Care – Kidney Disease Monitoring Diabetes Care – Blood Sugar Controlled Controlling Blood Pressure Rheumatoid Arthritis Management Reducing the Risk of Falling Plan All-Cause Readmissions Getting Needed Care Getting Appointments and Care Quickly Customer Service Rating of Health Care Quality Rating of Health Plan Care Coordination Complaints about the Health Plan Members Choosing to Leave the Plan Beneficiary Access and Performance Problems Health Plan Quality Improvement Plan Makes Timely Decisions about Appeals Reviewing Appeals Decisions Call Center – Foreign Language Interpreter and TTY Availability Call Center – Foreign Language Interpreter and TTY Availability Appeals Auto–Forward Appeals Upheld Complaints about the Drug Plan Members Choosing to Leave the Plan Beneficiary Access and Performance Problems Drug Plan Quality Improvement Rating of Drug Plan Getting Needed Prescription Drugs MPF Price Accuracy Medication Adherence for Diabetes Medications Medication Adherence for Hypertension (RAS antagonists) Medication Adherence for Cholesterol (Statins) MTM Program Completion Rate for CMR Measure Type Process Measure Process Measure Process Measure Outcome Measure Outcome Measure Process Measure Process Measure Process Measure Process Measure Process Measure Process Measure Process Measure Process Measure Process Measure Intermediate Outcome Measure Intermediate Outcome Measure Process Measure Process Measure Outcome Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Measures Capturing Access Improvement Measure Measures Capturing Access Measures Capturing Access Measures Capturing Access Measures Capturing Access Measures Capturing Access Measures Capturing Access Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Measures Capturing Access Improvement Measure Patients’ Experience and Complaints Measure Patients’ Experience and Complaints Measure Process Measure Intermediate Outcome Measure Intermediate Outcome Measure Intermediate Outcome Measure Process Measure Weight 1 1 1 3 3 1 1 1 1 1 1 1 1 1 3 3 1 1 3 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 5 1.5 1.5 1 3 3 3 1 Improvement Measure Yes Yes Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No Yes Yes Yes Yes 27 102 • AIS’s Management Insight Series February 19, 2016 Summary of Comments to the Request for Comments on 2017 Star Ratings and Beyond On November 12, 2015, CMS released a memo, Request for Comments: Enhancements to the Star Ratings for 2017 and Beyond, to Part C and D sponsors, stakeholders and advocates. The memo described CMS’ proposed methodology for the 2017 Star Ratings for Medicare Advantage (MA) and Prescription Drug Plans (PDP). We received approximately 90 comments representing plan sponsors, associations, consumer groups, and measurement development organizations. Concerns about and requests for clarification of specifications have been passed along to measure developers and stewards. This document provides a summary of the comments received and how we addressed these comments in the draft 2017 Call Letter. A. Changes to Measures for 2017 1. Improvement measures (Part C & D). Summary of Comments: Our proposal to use the 2014 CAHPS measure score (used in 2015 Star Ratings) as the baseline for the 2017 improvement calculation for that measure if a contract’s CAHPS measure score moved to very low reliability with the exclusion of the enrollees with less than 6 months of continuous enrollment for the 2015 survey administration was not supported by the majority of commenters. The following reasons were cited: The proposal does not provide an accurate reflection of more recent efforts that may have been made to impact CAHPS performance. Sponsors may be disadvantaged by use of dated 2014 CAHPS measure scores. These commenters were either against using prior year CAHPS data or proposed to exclude the measure from the improvement measure. About half requested additional clarification. A small number of additional comments related to the improvement measures included: Excluding the Call Center measure or applying revisions due to transparency in the reporting of plan performance or to increase the metric’s sample size or tailor the calls/ make test calls to test availability of interpreters. Including either the Part C or the Part D improvement measure to benefit the rating, as well as reducing the current weight of 5 to 3. Excluding the MTM measure from the improvement measure. Considering achievement levels along with grading on the curve for some measures or receiving the improvement measures above a 3-star rating. Response: CMS appreciates these concerns. We will use the 2014 CAHPS data only if there is a significant improvement from 2014 to 2016 when they do not have 2015 data due to very low reliability. This policy would affect very few contracts, but this would hold contracts harmless from missing CAHPS data. 1 The Challenges of Pharmacy Star Ratings • 103 February 19, 2016 1. Reviewing Appeals Decisions/Appeals Upheld measures (Part C & D). Summary of Comments: Almost all commenters supported the proposed change to include appeal cases that are reopened and decided prior to May 1, 2016 in the upheld measure. Additionally, some commenters suggested: Extending the time frame to include reopened cases past May 1st. Removing cases where the IRE obtains new or different information in making a decision, Providing sponsors with the same reports that CMS receives. Including reopening cases that occur in the following year’s data. Accounting for the volume of cases appealed. Adjusting the threshold for contracts to be excluded based on the contract’s membership . Response: CMS will move forward with changing the reopening deadline to May 1. CMS will review other suggestions. 2. Contract Enrollment Data (Part C & D). Summary of Comments: Most commenters supported the proposed change. Some commenters did not understand that we used the enrollment in the Special Needs Plans (SNP) Care of Older Adults (COA) measures only for submissions that do not contain a valid eligible population element. Response: Further review by CMS has shown that the proposed change is no longer necessary due to changes in the way enrollment data are processed over time. 3. Transition from ICD-9 to ICD-10 (Part C & D). Summary of Comments: Almost all commenters suggested the transition to ICD-10 is appropriate. Some suggested errors are likely because of the complexity of ICD-10 and that plans should be held harmless for errors the transition might cause. Many noted NCQA’s guidance and asked CMS to provide similar guidance and/or to prompt PQA to provide similar guidance. Some suggested PQA has not anticipated likely challenges, while others worry PQA measures may require reference to diagnostic codes that PDPs and pharmacies would not be able to access. Response: PQA measures currently used in Star Ratings do not reference ICD-9 diagnostic codes, so CMS will clarify that the transition to ICD-10 is not relevant to those measures at this time. CMS will encourage measure stewards to provide guidance about the transition, if and as needed. 4. Appeals Upheld measure (Part D). Summary of Comments: Several commenters disagreed with the proposal to again include cases for beneficiaries enrolled in hospice in this measure. Sponsors were concerned that data from hospice-enrolled beneficiaries are unreliable and not reflective of plan performance. 2 104 • AIS’s Management Insight Series February 19, 2016 Commenters also noted that appeals from members in hospice do not relate to services rendered by the MA; including those in the calculation may result in a rate which does not represents the true compliance of the MA. Additionally commenters are concerned that they may be negatively impacted as they may not have the visibility to hospice status at the time of the initial review due to the lag in timing of receipt of hospice indicators and/or retroactive changes to hospice status as sent on the TRR. A few commenters agreed with the proposed change. One commenter also suggested removing cases where the IRE obtains new or different information in making a decision and to align time frames and processes for plan sponsors and IREs to make a more equitable evaluation of plan sponsor decisions. Response: As noted in the 2016 Call Letter, this exclusion was only necessary for the 2016 measure as it was based on 2014 data that may have been affected by policy changes in 2014. CMS policy has not changed since 2014, and there is no reason to exclude hospice appeal cases from the 2017 Star Rating Appeals Upheld measure. 5. Medication Therapy Management (MTM) Program Completion Rate for Comprehensive Medication Reviews (CMR) measure (Part D). Summary of Comments: We received only positive feedback for CMS’ inclusion of more detailed data during HPMS plan previews. Sponsors supported CMS program audits of MTM, but many voiced concerns about the specific audit standards to be used, if sponsors could be penalized due to MTM program variations, and how these audits would differ from the current data validation activities for plan-reported data. Response: In the Data Integrity section, CMS will clarify that more information about MTM audit criteria will be released soon; questions should be directed by email to [email protected]. We will also clarify that we will not apply any relevant MTM program audit findings that could demonstrate data integrity issues for sponsors that participate in the MTM audit during the pilot. We will also clarify that Data Validation standards assess compliance to CMS’ reporting requirements and technical specifications while CMS program audits are more comprehensive assessments of the contracts’ MTM programs. B. Removal of Measures from Star Ratings 1. Improving Bladder Control (Part C). Summary of Comments: Most commenters expressed some doubt about the validity, reliability or utility of this measure and supported removing the measure from Star Ratings, and giving plans advance notice if the measure were to return to the Star Ratings. Some plans thought expecting improvement in bladder control unrealistic for their (Special Needs) population or not meaningful given the other conditions (e.g., receiving dialysis for kidney failure) their enrollees may have. Many asked for clarification if CMS intends to return the measure to Star Ratings in 2018, which would be the normal course of action. Some stated they would prefer this measure were dropped entirely. Some commenters thought the focus of the measure should move beyond receipt of treatment, while others preferred the focus remains on receipt of treatment. 3 The Challenges of Pharmacy Star Ratings • 105 February 19, 2016 There was also confusion about whether this measure was a cross-sectional measure or a twoyear change score. Response: CMS will clarify that the HEDIS measures derived from the HOS survey, like Improving Bladder Control, are cross-sectional and do not require comparison of responses from the same cohort two years later. CMS will forward comments to the measure steward, NCQA. 2. High Risk Medication (Part D). Summary of Comments: A majority of the commenters supported removing the High Risk Medication (HRM) measure from Star Ratings and moving to the display measures for 2017. Over one-quarter of the commenters supported removal of the HRM measure from Star Ratings, but for 2018 or beyond, noting that plan sponsors have made significant investments to improve HRM performance. Few commenters did not support this change. A number of commenters recommended sufficient lead time before future updates to the measure specifications are implemented to consider formulary and bid timelines. Some commenters suggested exclusion of hospice patients from the measure calculation. Response: CMS will move forward with the proposal to remove the HRM measure from the Star Ratings and move to the display measures for 2017. We shared specification comments with the PQA. The timing of implementation of updates to the measure specifications in the future will provide sufficient lead time ahead of the formulary and bid deadlines. C. Data Integrity Summary of Comments: Many supported the development of program audits of Part D sponsors’ MTM programs, and voiced similar concerns as described in section A6 above. A few commenters complained about CMS policy to reduce measures due to Data Validation failures at an element level, or in general that this policy is unwarranted, and that reductions to 1 star should only be made if sponsors intentionally made data errors. One commenter suggested that if CMS or its contractors made errors, then affected sponsors should be assigned 5 stars, and/or be able to reuse the prior year’s data. Response: As noted in A6’s response, we will clarify that information about the MTM audit criteria will be released soon, and that findings identified during pilots of the MTM audit criteria would not be used. We will clarify that reductions are based on systemic failures. A reporting section’s overall DV score may be high enough to consider the contract has passed, but specific element-level failures show inaccurate data that would be used in the measure calculation. For example, if the Data Validation found errors in the numbers of beneficiaries enrolled in the MTM, or receiving CMR – but the overall MTM DV score was above 95, CMS would still have concerns about the MTM CMR numerator and denominator. CMS and its contractors also share in the responsibility for accurate collection and calculation of Star Ratings. Each year, CMS reviews the quality of the data across all measures, variation among organizations and sponsors, and the accuracy and validity of measures before making a final determination about inclusion in the Star Ratings. Rarely, as a result of these reviews, CMS has needed to exclude specific measures from the Star Ratings because systemic errors were found in 4 106 • AIS’s Management Insight Series February 19, 2016 the data; these errors may not have been caused by sponsors’ processes or actions. While CMS considers all options prior to exclusion of a measure, it would be incorrect to infer contracts’ performances from historical data, or reuse prior years’ data as individual contract measure performance changes from year to year. This would also unfairly disadvantage contracts, as use of prior year scores may result in lower Star Ratings, versus if the measure was excluded. Using data from a prior collection year for some but not all contracts (i.e., holding contracts harmless for downward adjustment), or assigning 5 stars for all contracts could also appear arbitrary. D. Impact of Socio-economic and Disability Status on Star Ratings Summary of Comments: There was widespread appreciation of the attention and careful examination of the LIS/DE/disabled effects that CMS has completed at this time and continues to research. Respondents agreed with the overall approach and focus of the research conducted. Many commenters applauded our efforts and were grateful for the acknowledgement of the issue and the development of the two proposed interim analytical adjustments set forth in the Request for Comments: a Categorical Adjustment Index or Indirect Standardization. Respondents valued the numerous opportunities afforded for stakeholder input. In addition, many commenters supported the engagement of the measure stewards and look to their research as part of the answer to the long-term solution to address any sensitivity of the Star Ratings measures to the enrollment of beneficiaries. Further, several re spondents complimented CMS for maintaining the integrity of the core of the Star Ratings Program and the transparent manner in which CMS approached the work related to the issue. Many respondents mentioned the value of the information provided during the December 3rd User Call, however, some still desired additional clarity of the methodologies. Most respondents requested simulated results so that sponsors could better understand the methodologies employed for the analytical adjustments and the impact of the adjustments on their scores. Further, numerous respondents requested the additional details of the interim adjustments and simulation results in advance of the draft 2017 Call Letter. Overall, the reaction to the proposed analytical adjustments was mix ed. The comments received did indicate a need for further outreach so that stakeholders would be comfortable supporting one of the options proposed to address the LIS/DE/ disabled effect. Many respondents felt they were unable to state a preference of one method over the other until they were able to review simulation results. Of the respondents that did express a preference to a single analytical adjustment, more commenters preferred the Categorical Adjustment Index (CAI) over the method of Indirect Standardization (IS). There was a request for the release of any research that focused on PDPs. Some commenters expressed a need for details and discussion on the application of the analytical adjustments for PDPs including the availability of the disability status of beneficiaries. Some respondents expressed the need for immediate action and financial relief for plans that served vulnerable populations. Several commenters wanted the interim adjustments to result in meaningful differences in the ratings. Respondents were unsure of the mechanics of the application of the adjustments for plans where beneficiaries were exclusively LIS/DE and others, requested the minimum proportion of beneficiaries in the subgroups of LIS/DE and/or disabled status to qualify for an adjustment. There were inquiries regarding the stability of the adjustments over time, the details of operationalizing the methods, and the strengths and weaknesses of each method. 5 The Challenges of Pharmacy Star Ratings • 107 February 19, 2016 There were a number of respondents that encouraged CMS to wait rather than to move forward with either of the proposed interim adjustments. Some respondents who preferred a delay referenced either looking to the work of the measure developers or the release of the Assistant Secretary of Planning and Evaluation’s (ASPE) report due in the upcoming year as required by the IMPACT Act. One commenter mentioned the need to ‘proceed with caution to avoid either creating a double standard of care or lowering standards for chronic disease management.’ A respondent questioned the value of an interim adjustment given the increase in the administrative burden for both CMS and sponsors. Another reason cited for waiting for implementation was that the proposed analytical adjustments did not adequately addre ss the issue. In terms of the timing for a policy response, one commenter suggested optional participation in 2017, while another suggested a two-year delay in implementation. Numerous commenters believed that CMS should further investigate or control for factors beyond LIS/DE and/or disabled status. The attributes that respondents mentioned for additional examination and/or consideration included: age, race, health literacy, household income, homelessness, hunger, transportation challenges, low literacy, unemployment, number of providers, risk scores, medical complexity, chronic conditions, and geographical or regional disparities. Some respondents mentioned that risk adjustment of quality measures would penalize plans that deliver high quality of care irrespective of the population served. There were a limited number of comments to help guide the selection of the measures to be adjusted. Some commenters mentioned that only measures that revealed a meaningful withincontract difference in our research should be candidates for adjustment. Other commenters believed all Star Ratings measures, regardless of whether they are already adjusted, should be included as candidates for adjustment. A respondent did note that if all measures were adjusted, it would still result in only a small proportion of the measures in the program. There was a comment that stated that any risk adjustment must be rooted in evidence and not disincentivize plans from enrolling vulnerable beneficiaries. Another respondent believed that some risk adjustment done prudently is better than none. One commenter would not support adjusting for medication adherence measures and believed that LIS beneficiaries in their plan do better and thus, the plan would be negatively impacted by such an adjustment. There were several commenters inquiring about the method of reporting – hybrid or census – and its impact on the interim adjustment methodologies. Some commenters expressed the need for accounting for LIS-look-alikes in the adjustments. Respondents believe that there are beneficiaries that do not quality for LIS but share common characteristics with LIS beneficiaries, such as community factors and income levels, and may in fact experience worse outcomes due to their non-LIS status. The look-alikes therefore impact the within-contract differences due to the modeling that uses a dichotomous variable for LIS status. Many of the general comments related to the two options CMS is exploring for interim analytical adjustments expressed concern about the possible bias an adjustment may introduce in the Star Ratings program. A respondent expressed the need to capitalize on the heterogeneity of the data. One commenter specifically discussed the importance of maintaining the quality signal that the Star Ratings provide. To compensate for the perceived advantage plans with a high percentage of LIS/DE/disabled enrollees may realize, many commenters supported a hold harmless provision. (Such a provision would ultimately result in only positive adjustments to the 6 108 • AIS’s Management Insight Series February 19, 2016 Overall and Summary Star Ratings.) Several commenters expressed concern regarding the use of the pre-adjusted cut points for the conversion of the adjusted measure score to a measure -level Star Rating and requested justification for this aspect of the methodology. One commenter felt that mixing unadjusted and adjusted measure scores and stars would result in apple to orange comparisons. A respondent asked for simulations using both adjusted and unadjusted cut points. Numerous respondents believed that the adjustments would increase the complexity of the Star Ratings and impinge on its transparency. In addition, there was concern about the possibility of a shortened plan-preview period. Several commenters suggested an additional plan preview period if CMS were to move forward with one of the interim adjustments. A number of commenters wanted justification of the use of unadjusted measure cut points for the conversion of the adjusted measures score to measure-level Star Ratings. Some commenters were unsure how the adjustment would be applied to plans that were almost exclusively comprised of Dual beneficiaries. Several respondents were unsure when the proposed adjustment would be applied in relationship to the Reward factor. Further, several comments inquired about which measures scores would be used for determining the integration factor – adjusted or unadjusted measures scores. Many of the comments received that were specific to the CAI were related to the number of initial and final adjustment categories employed in the method. Commenters were concerned about the collapsing of the initial groups and final adjustment categories and the impact of the groupings on the adjustments and possible misclassification. Respondents who preferred CAI over IS cited reasons such as: easier to understand, similarity to CAHPS, greater transparency, ability to have the adjustment factors in advance of the plan preview, flexibility and accuracy of the method. A number of commenters suggested using additional covariates in the model such as race, gender, and community factors. Some commenters wanted to know more about the stability of the adjustment factors over time. A number of the comments received that were specific to IS noted the increased complexity of the method. Many commenters expressed concern about the validation needed if the method was employed and the impact on the plan preview period. Some respondents felt the adjustment was more tailored to their specific plan and perhaps, more accurate. Several commenters believe geographic comparisons instead of national comparisons should be used for the standardization process. There were a number of comments regarding the determination of the national means and the impact of its value on measures when the information is calculated based on mixed reporting methods (census or hybrid). There were a limited number of responses related to the additional response to address lack of an LIS indicator for enrollees in Puerto Rico. Overall, there was appreciation to CMS for addressing the unique circumstances and challenges in Puerto Rico. Respondents expressed concern about the accuracy of the proposed LIS Indicator used for adjustment based on the relationship developed using mainland data and modifying it for use in Puerto Rico. Further, the commenters suggested an adjustment of the medication adherence measures for contracts operating in Puerto Rico. Response: While the measure stewards are undertaking a comprehensive review of their measures used in the Star Ratings Program and ASPE is continuing its work under the IMPACT Act, CMS is proposing to implement the CAI as an interim analytical adjustment for 2017 Star 7 The Challenges of Pharmacy Star Ratings • 109 February 19, 2016 Ratings to take into account the impact of LIS/DE and/or disability status on Star Ratings measures. The CAI is a factor that would be added or subtracted to a contract’s Overall and/or Summary Star Rating to adjust for the average within-contract disparity. The proposed interim solution adheres to certain core CMS principles, such as not permitting a lower standard of care for vulnerable beneficiaries, proposing adjustments that reflect the actual magnitude of the differences observed in the data, and recognizing the need for options that are both transparent and feasible for the plans and CMS to implement. The proposal relies on an adjustment external to the measure specifications, as well as, maintaining the integrity of the Star Ratings and the core of its methodology. For contracts operating in Puerto Rico, we plan to proceed with estimating an LIS Indicator, while other data sources continue to be explored. E. 2017 CMS Display Measures 1. Timely Receipt of Case Files for Appeals (Part D) & Timely Effectuation of Appeals (Part D). Summary of Comments: All commenters agreed to change the time frame of the Timely Receipt of Case Files for Appeals (Part D) & Timely Effectuation of Appeals (Part D) from the first six months of the current year to entire twelve months of the previous year. Response: CMS will move forward with making this change. 2. Medication Reconciliation Post Discharge (Part C). Summary of Comments: One third of the commenters were supportive of this measure being included on the 2017 display page and 2018 Star Ratings. However, almost half of the commenters had concerns about the timing of the measure and requested it delayed by either leaving it on the display page for at least two years or delay it on the display page and Star Ratings for at least a year. Many commenters wanted more clarification on the measure, specifically on the roles to complete the measure (i.e., social workers, pharmacists), data collection, and clarification or changes to technical specifications such as eligible members included in the denominator. There were also a few comments questioning the measure’s construction, assumptions and validity. Examples included difficulty in collecting accurate information that medications were reconciled post discharge for D-SNP population, physicians coding accurately and addressing where members are utilizing multiple providers. A few thought the measure did not differ much from readmissions or the existent MRP measure required for physicians. Two comments requested benchmarking and cut point proposals including case mix and SES adjustments. Response: CMS is planning to proceed to include this measure as part of the 2017 display page and the 2018 Star Ratings. This measure has been collected by SNPs for a number of years. Detailed specifications are available in HEDIS 2016, volume 2. We will monitor the 2016 data submissions for any data issues and modify our plans if needed. 3. Hospitalizations for Potentially Preventable Complications (Part C). Summary of Comments: Almost all commenters were not supportive of the timing of the measure. They recommended it be delayed from the display page and Star Ratings or have it remain on display page for an additional year or two. Some commenters wanted information such as specifically what the ambulatory sensitive conditions are for this measure or requested 8 110 • AIS’s Management Insight Series February 19, 2016 narrowing the scope on ambulatory sensitive conditions. A few of the comments had validity concerns about the measure specifically about the risk adjustment. One commenter had concerns with comparing the ESRD population with a general Medicare Advantage population in this measure. Response: CMS is planning to proceed to include this measure as part of the 2017 display page and the 2018 Star Ratings. Detailed specifications are available in HEDIS 2016, volume 2. We will monitor the 2016 data submissions for any data issues and modify our plans if needed. 4. Statin Therapy for Patients with Cardiovascular Disease (Part C). Summary of Comments: Almost all comments were negative, citing concerns with the measure’s validity. Some noted a general lack of consensus with the 2013 ACC/AHA blood cholesterol guidelines, with others identifying detailed methodological issues. Several commenters note d that the measure does not account for members for whom statins are contraindicated, not well tolerated, not recommended, or refused. Others stated the measure does not account for alternate therapies or a wider range of statin dosages. A few commenters recommended moving this to a Part D measure like the related Statin Use in Persons with Diabete s (SUPD) measure and similarly excluding hospice beneficiaries. While a few agreed with the age ranges, just as many disagreed. Some commenters requested more specifics from CMS for the diagnostic codes and for “high or moderate statin.” Because of concerns with the measure’s validity, along with plans having limited time to implement quality improvement, several asked CMS to omit from the display page in 2017. Many others asked for the measure to stay on the display page for at least two years and/or to hold off reporting measure in the 2018 Star Ratings. Very few commenters agreed with CMS on this measure. Response: Comments have been shared with NCQA. We are aware that treatments for cardiovascular disease are evolving and we will continue to monitor best practices for standards of care. We will keep the measure on the display page for an additional year to gain more experience with new treatment guidelines and metric and then add to the 2019 Star Ratings. 5. Asthma Measures (Part C). Summary of Comments: The majority of commenters argued against adding the two proposed asthma measures as 2017 Star Rating display measures or as Star Rating measures in the future. These two measures expand NCQA’s current asthma measures to include older adults and are defined on the basis of medication utilization. The reasons against included: 1) belief that these are inappropriate measures for the majority of the Medicare population who, being ages 65 and over, are more likely to receive a COPD or other pulmonary disease diagnosis rather than an asthma diagnosis and comments argue f or focusing efforts on areas that impact greater numbers of members; 2) concern about the difficulty for physicians to distinguish asthma from COPD in the senior population; 3) a few recent studies have indicated that medication management for people with asthma has not been shown to correlate with improved health outcomes, lower costs or lower utilization; 4) concern with the measures being constructed on the basis of medication utilization rather than on diagnosis given that some of the same medications are used for both asthma and COPD thus drug claims do not provide an accurate picture of which members have persistent asthma; 5) difficulty distinguishing between 9 The Challenges of Pharmacy Star Ratings • 111 February 19, 2016 ‘seasonal’ versus ‘persistent’ asthma; 6) focusing on patients remaining on medications ‘during the treatment period through the end of measurement year’ requires defining an appropriate period of treatment for a condition which exhibits seasonality; 7) measures are not in line with NIH recommendations for step-down controller therapy or management of patients who exhibit seasonal variation in asthma symptoms; and 8) concern that for a population of persons often with multiple health conditions (seniors and the disabled), there are considerations of possible adverse medical consequences due to medication interactions. A number of commenters requested CMS delay the addition of these 2 asthma measures for a few years until: 1) the measures are fully specified; 2) there is some evidence that asthma medication management is shown to correlate with improved health outcomes; 3) physicians are more experienced using the ICD-10 coding system which is thought to better delineate asthma from COPD; and 4) NCQA has fully specified, vetted and published results of these measures. The only commenters arguing for the inclusion of the proposed asthma measures were pharmaceutical organizations. Lastly, a number of comments argued that those plans which serve large numbers of the under age 65 Medicare population, the dual eligibles, would be further burdened by the inclusion of these measures due to socioeconomic factors which cannot be controlled but impact the frequency and severity of asthma events. Response: CMS appreciates the comments received on this section. CMS shared the comments with NCQA and will continue to monitor development of these measures. CMS is planning to include these on the 2017 and possibly 2018 display page and will consider these for inclusion in Star Ratings for future years. 6. Statin Use in Persons with Diabetes (SUPD) (Part D). Summary of Comments: Comments similar to those received for the Part C Statin Therapy for Patients with Cardiovascular Disease measure were submitted. There was mixed support to add the SUPD measure to the 2017 display page and to the 2018 Star Ratings. While there was support for this measure in general, a majority of commenters cited concerns with the measure’s validity, or asked for the measure to stay on the display page for at least two years to gain experience or until methodological issues could be resolved. Several commenters noted that the measure does not account for members for whom statins are contraindicated, not well tolerated, not recommended, or refused. Others stated that the specifications should account for or exclude beneficiaries taking PCSK-9 therapies. Several commenters were concerned that prescription claims, not diagnosis codes, are used to determine the presence of diabetes, questioned the age criteria, or recommended excluding ESRD patients. Response: Comments regarding the measure’s technical specifications have been shared with the PQA. We will consider keeping the measure on the display page for an additional year to gain more experience with new treatment guidelines and metric and then add to the 2019 Star Ratings. 10 112 • AIS’s Management Insight Series February 19, 2016 Forecasting to 2018 and Beyond F. New Measures: 1. Care Coordination Measures (Part C). Summary of Comments: Support for new care coordination measures was mostly positive, although several commenters recommended against measures that involve chart review because of increased administrative cost. A few commenters requested more detail and for CMS to put specific proposed measures out for public comment before they become display measures. A couple of commenters recommended that CMS take into account mental health issues and dual/LIS/disabilities and consider a population-based approach. A few mentioned that measures should consider multiple provider types or suggested that CMS use existing reporting requirements related to care coordination. Several encouraged CMS to look at care coordination measures that are linked to improved outcomes. Some stated that it is critical for CMS to validate encounter data to ensure they are complete and accurate before relying on them. Response: CMS appreciates the comments received on this section. We shared comments with contractors developing care coordination measures and will continue to provide updates to the industry as the work progresses. 2. Depression Measures (Part C). Summary of Comments: Many commenters expressed concerns about privacy laws, as well as readiness of electronic systems to transmit clinical data from behavioral health providers. Several suggested CMS should first focus on depression screening measures before including a depression outcome measure. A few requested that other depression or mental health screening tools be included. A couple expressed concern that the 6 months measurement window is too short to demonstrate impact. Several stated that the measure should be on display for several years following HEDIS approval. Response: CMS shared comments received on this topic with NCQA and will continue to monitor development of the measure. NCQA is also working on a Depression Screening and Follow-Up measure which may be proposed for HEDIS in the future. 3. Appropriate Pain Management (Part C). Summary of Comments: NCQA is exploring opportunities to develop a new measure focusing on appropriate pain management. The intent is to assess the quality of pain management and treatment. Commenters expressed appreciation for the exploration of this important topic. However, commenters stressed that the experience of pain is subjective and varies across individuals, conditions, and stages of a condition. Commenters stressed that these considerations need to be addressed in specifying an appropriate pain management measure so that valid comparisons can be made across plans and member populations. 11 The Challenges of Pharmacy Star Ratings • 113 February 19, 2016 Commenters requested full measure specification to enable plans to provide meaningful comment. Further, that if a measure is created and implemented, it remain on the display page for at least 2 years. One commenter suggested that a standardized pain screening tool be developed / employed to assess, document, and monitor the experience of pain. Another commenter suggested that in terms of pain control, the measure address not only medication use but also alternative treatments for managing pain. Lastly, two commenters indicated that appropriate pain management can be at odds with controlling/monitoring for opiate use. Response: CMS appreciates the comments received on this section. CMS shared the comments with NCQA and will continue to monitor development of these measures. 4. Use of Opioids from Multiple Providers or at High Dosage in Persons without Cancer (Part D). Summary of Comments: There was general agreement on the importance of resolving opioid overutilization and not adding the measures to the Star Ratings at this time. While there was some support to add these three PQA opioid overutilization measures to the 2018 Part D display page (using 2016 data), most commenters were opposed or expressed concerns with the proposal. Concerns about the proposal included: 1) neither clinical guidelines nor experience support the validity of the measures, 2) the measures do not measure and will not support plan performance, particularly #1 and #2, 3) more detailed descriptions of the measure specifications are needed in order to consider support, 4) some sponsors will be disadvantaged based upon the characteristics of their enrollees (e.g., high numbers of disabled enrollees), 5) Part D sponsors have limited ability to influence prescriber behavior, and 6) if the measures are not suitable for Star Ratings, they should not be display measures. Response: Due to concerns raised by commenters, CMS will implement the three PQA opioid overutilization measures as Patient Safety measures, not display measures, for one year to gain experience with the measures and pending new guidelines (e.g., from CDC) and current research on opioid prescribing, overutilization, and interventions. We shared specification comments with the PQA. 5. Antipsychotic Use in Persons with Dementia (APD) (Part D). Summary of Comments: There was general agreement that this measure addresses an important issue. Over half of the commenters agreed with the proposal, and one -third were neutral. Changes to the proposal were suggested by two commenters, and over two-thirds of the commenters noted specific concerns, including 1) lack of access to diagnosis data required for the measure, 2) limitations on the ability of sponsors to intervene because antipsychotic drugs have protected class status, 3) a desire to review the complete measure specifications, and 4) this measure is primarily associated with nursing homes, the nursing home quality rating reporting system includes a related measure, and the facilities are responsible, not plans. Response: CMS will proceed as proposed. CMS will include a link to the CMS.gov webpage for an APD measure analysis report that provides detailed specifications and testing results. We shared specification comments with the PQA. 12 114 • AIS’s Management Insight Series February 19, 2016 G. Changes to Existing Star Ratings and Display Measures and Potential Future Changes: 1. Colorectal Cancer Screening (Part C). Summary of Comments: This measure is under consideration for revision. The USPSTF will release its revised guidelines in late 2016. At that time, NCQA will consider revising the Star Rating Colorectal Cancer Screening HEDIS measure. The comments supported waiting for the USPSTF’s final evidence -based recommendations for colorectal cancer screening methods. The commenters stressed the need for plans to be provided, in advance of implementing any changes to the Star Ratings, additional detail of specification changes to allow stakeholders to provide meaningful comment. The comments also stressed the need for advance notice of the timeline for implementing changes so as to educate providers and beneficiaries. It was suggested that any revised measure not be implemented until the 2019 Star Rating year. One commenter requested there be no change to the measure’s age limits. One commenter requested the measure be broken down by age groups. Another commenter stressed that universal colorectal cancer screening is not supported for persons with ESRD. Response: CMS appreciates the comments received on this section. CMS shared the comments with NCQA and will continue to monitor development of these measures. 2. Fall Risk Management (Part C). Summary of Comments: Commenters were equally split between supportive, negative and neutral comments. Many commenters would like more time for research or for plans to prepare for changes. Some commenters would prefer CMS not use measures derived from surveys of beneficiaries. A large number of commenters suggested that while it was appropriate to update specifications to reflect changing guidelines, there is still not enough evidence to consider vitamin D as a treatment to reduce falls. Others commented that specification changes can change the focus of the measure and “eliminate its principal focus of prevention of falls among members by eliminating physical barriers.” A few Special Needs Plans serving dialysis patients expressed concerns that vitamin D is not appropriate treatment for patients also getting dialysis (it risks leading to hypercalcemia) and therefore some exclusion is needed for those plans or for most of their enrollees. Response: CMS shared these comments to NCQA and highlighted the concerns commenters have about the evidence for use of vitamin D as a treatment, as well as the concern that Special Needs Plans that focus on End Stage Renal Disease may not have enrollees appropriate for this measure. 3. Pneumococcal Vaccination Status for Older Adults (Part C). Summary of Comments: Almost all comments were negative, citing concerns that the measure relies on Medicare CAHPS survey data. Self-reports, according to commenters, are unreliable because members may not recall getting a specific vaccine, especially if administered a long time ago. Instead, most commenters recommended CMS use claims data instead. Still, a few cautioned that claims data is imperfect and ask for lower cut points. Some plans asked CMS to 13 The Challenges of Pharmacy Star Ratings • 115 February 19, 2016 keep the measure as display and not add to Star Ratings until tested further, while a pharmaceutical company encouraged CMS to include it the Star Ratings. Response: We appreciate the concerns of commenters. We will monitor data during the display period for issues with validity. 4. CAHPS measures (Part C & D). Summary of Comments: Commenters were mostly positive about the proposed change to the 5.0 version of CAHPS. Several requested that the CAHPS survey be shortened in order to increase response rates and stated that questions should focus on questions related to consumer satisfaction. A few commenters expressed concern with CAHPS in general as they feel it is subjective. One expressed concern about benchmark rate changes and CAHPS scores. Comments on the sampling proposal were mixed. Response: Patient experience surveys such as CAHPS focus on how patients experienced or perceived key aspects of their care, not how satisfied they were with their care. CAHPS surveys follow scientific principles in survey design and development. The surveys are designed to reliably assess the experiences of a large sample of patients. They use standardized questions and data collection protocols to ensure that information can be compared across healthcare settings. CAHPS surveys are developed with broad stakeholder input, including a public solicitation of measures and a technical expert panel, and the opportunity for anyone to comment on the survey through multiple public comments period through the Federal Register. Regarding survey length, analyses suggest that the relationship between survey length and response rate for the MA CAHPS survey is only weakly negative. Specifically, the use of 12 supplemental items as compared to none was associated with a 2.5% reduction in response rate (Beckett et al, Public Opinion Quarterly, in press) 1. To examine the impact of decreasing benchmarks on CAHPS measure scores, we conducted an analysis of changes in CAHPS stars with changes in proxy standardized MA benchmarks at the contract level. The benchmarks were developed as an average of plan-level BPT benchmarks weighted by July enrollment for each year. Analyses on the impact on CAHPS scores did not find a systematic relationship between CAHPS scores and declining benchmarks. The table below includes a comparison of contracts changing their CAHPS measure star rating from 2014 to 2015 comparing the quartile with the biggest drop in benchmarks ($81 or more) with the quartile with the smallest changes in benchmarks (+$30 to -$47): 1 Beckett MB, Elliott MN, Gaillot S, Haas A, Dembosky JW, Giordano LA, Brown J. (In Press) “Establishing limits for supplemental items on a standardized national survey.” Public Opinion Quarterly 14 116 • AIS’s Management Insight Series February 19, 2016 Measure Care Coordination Quartile with largest drop in benchmarks (% increasing stars/% staying same/% decreasing) 23%/51%/26% Quartile with smallest change in benchmarks (% increasing stars/% staying the same/% decreasing) 24%/51%/25% Customer Service Getting Appointments and Care Quickly Getting Needed Care Rating of Health Care Quality 22%/45%/33% 21%/61%/18% 23%/46%/31% 22%/54%/24% 30%/45%/25% 11%/68%/21% 23%/48%/29% 9%/69%/21% Rating of Health Plan Getting Needed Prescription Drugs Rating of Drug Plan 18%/60%/22% 18%/45%/36% 21%/57%/21% 20%/56%/24% 20%/51%/29% 24%/59%/17% 5. Medication Adherence for Hypertension (RAS Antagonists) (Part D). Summary of Comments: All commenters supported the proposed exclusion. One commenter requested we exclude dual-eligible beneficiaries that have $0 copay. Another requested we emphasize ESRD patients are excluded. A few requested CMS clarify which year this change applies; two requested CMS set upper age limits for all 3 medication adherence measures. Response: We will proceed as planned. We will clarify that this exclusion will be applied for the 2017 Star Ratings. We shared specification change suggestions with the PQA. 6. MPF Price Accuracy (Part D). Summary of Comments: About one-quarter of commenters agreed with the proposed changes, and the remaining were neutral or opposed. The most common comments included: Broaden the cut points because the measure scores are tightly clustered resulting in insignificant/negligible differences and therefore do not inform beneficiaries of any impactful differences among the plans. Retire or move the measure to the display page due to high scores. Change the methodology for scoring this measure. Clarify CMS’ decision making process that led to this change, and provide details of our analysis of this change. Provide best practices for how to achieve a high score. Move to display page for one year due to methodology changes. Other commenters commented that the different frequency of point of sale and MPF pricing updates is a barrier to improving plan performance. Also, if the measure has new scoring methodology, commenters suggested temporarily removing the measure from the Improvement 15 The Challenges of Pharmacy Star Ratings • 117 February 19, 2016 Measure. Lastly, some commenters would like to be able to access detail reports of their scores earlier and more frequently. Technical comments included 1) the amount of difference between PDE prices and MPF prices that constitute an inaccuracy should be broadened to more than one half of one cent, 2) only the Formulary Reference File NDCs should be used when selecting the PDE claims that are being measured because pricing for the same drug, strength, and dosage form can vary from manufacturer to manufacturer, and 3) the “Patient Residence Code” should be used to determine the retail claims from the PDEs. Response: We will proceed with these changes as planned. CMS’ simulations found that the accuracy scores using the new methodology were generally similar to scores calculated using the current methodology. This measure will continue to be excluded from the Improvement Measure. CMS uses accepted mathematical algorithms and practices to formulate the thresholds (cut points) required to earn a particular rating. Currently the majority of contracts receive high Accuracy scores. We do not believe options to further differentiate plans’ performances would be supported by sponsors, as this could entail identifying price differences smaller than one half of one cent. We continue to be open to other changes that will improve the impact of this measure. 7. Drug-Drug Interactions (DDI) (Part D Display). Summary of Comments: Most commenters were neutral, while a few voiced concern about the implementation of future changes to PQA specifications including the timing of adoption by CMS for the display measure. Response: We will proceed as planned. Any future changes will be announced via the annual Request for Comments and draft Call Letter process to provide advance notice. 8. Center for Medicare and Medicaid Innovation Model Tests. Summary of Comments: Commenters appreciated CMS’ attention to the potential for improvements in quality in the MA-VBID and Part D Enhanced MTM models to favorably influence Star Ratings for contracts with participating plans and expressed a desire for contracts not to be disadvantaged for either participating or not participating. Several requested that CMS provide additional information about impact on plans. Several mentioned that only some contracts are permitted to participate in the models, a few suggested that SNPs and territories (e.g., Puerto Rico) should be allowed to participate. Some suggested that waiving MTM reporting requirements for participating plans could impact cut points for this measure, and a few requested that CMS suspend the MTM Star Rating during the model test years. Response: CMS will closely monitor performance of contracts participating in the model to evaluate any effect on Star Ratings. Our goal is to ensure that contracts are not penalized. Some possible options are to establish different cut points for model participants and to case mix adjust scores for the purpose of determining cut points. We will provide more information to stakeholders as it is available and continue to engage with stakeholders on the impact of these models. 16 118 • AIS’s Management Insight Series February 19, 2016 H. Measurement and Methodological Enhancements. Summary of Comments: Comments ranged in topics from general to measure-specific. They included comments about specific Star Rating changes and measures, cut points, development of outcome measures, the Reward Factor, as well as display measures. Response: CMS appreciates all comments and will explore the feasibility of specific proposals for possible future implementation. For example, we will continue to look at the issue of whether to conduct call center monitoring in languages proportional to the prevalence of each language in the 65 and older U.S. population. 17 The Challenges of Pharmacy Star Ratings • 119 2017 Part C & D Star Ratings Measures 2017 ID C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 2016 ID C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 Breast Cancer Screening Colorectal Cancer Screening Annual Flu Vaccine Improving or Maintaining Physical Health Improving or Maintaining Mental Health Monitoring Physical Activity Adult BMI Assessment Special Needs Plan (SNP) Care Management Care for Older Adults – Medication Review Care for Older Adults – Functional Status Assessment Care for Older Adults – Pain Assessment Osteoporosis Management in Women who had a Fracture Diabetes Care – Eye Exam Diabetes Care – Kidney Disease Monitoring Diabetes Care – Blood Sugar Controlled Controlling Blood Pressure Rheumatoid Arthritis Management Reducing the Risk of Falling Plan All-Cause Readmissions Getting Needed Care Getting Appointments and Care Quickly Customer Service Rating of Health Care Quality Rating of Health Plan Care Coordination Complaints about the Health Plan Members Choosing to Leave the Plan Beneficiary Access and Performance Problems Health Plan Quality Improvement Plan Makes Timely Decisions about Appeals Reviewing Appeals Decisions Call Center – Foreign Language Interpreter and TTY Availability HEDIS HEDIS CAHPS HOS HOS HEDIS / HOS HEDIS Part C Plan Reporting HEDIS HEDIS HEDIS HEDIS HEDIS HEDIS HEDIS HEDIS HEDIS HEDIS / HOS HEDIS CAHPS CAHPS CAHPS CAHPS CAHPS CAHPS CTM MBDSS CMS Administrative Data Star Ratings IRE IRE Call Center D01 D01 Call Center – Foreign Language Interpreter and TTY Availability Call Center D02 D03 D04 D05 D06 D07 D08 D09 D10 D11 D12 D13 D14 D15 D02 D03 D04 D05 D06 D07 D08 D09 D10 D11 D12 D13 D14 D15 Appeals Auto–Forward Appeals Upheld Complaints about the Drug Plan Members Choosing to Leave the Plan Beneficiary Access and Performance Problems Drug Plan Quality Improvement Rating of Drug Plan Getting Needed Prescription Drugs MPF Price Accuracy High Risk Medication Medication Adherence for Diabetes Medications Medication Adherence for Hypertension (RAS antagonists) Medication Adherence for Cholesterol (Statins) MTM Program Completion Rate for CMR Measure Primary Data Source Improvement Measure Weight Yes 1 Yes 1 Yes 1 No 3 No 3 Yes 1 Yes 1 Yes 1 Yes 1 Yes 1 Yes 1 Yes 1 Yes 1 Yes 1 Yes 3 Yes 3 Yes 1 Yes 1 Yes 3 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 No 1.5 No 5 Yes 1.5 Yes 1.5 Yes 1.5 Yes 1.5 IRE Yes 1.5 IRE Yes 1.5 CTM Yes 1.5 MBDSS Yes 1.5 CMS Administrative Data No 1.5 Star Ratings No 5 CAHPS Yes 1.5 CAHPS Yes 1.5 PDE data, MPF Pricing Files No 1 Prescription Drug Event (PDE) data Yes 3 Prescription Drug Event (PDE) data Yes 3* Prescription Drug Event (PDE) data Yes 3* Prescription Drug Event (PDE) data Yes 3* Part D Plan Reporting, Medicare Yes 1 Enrollment Database (EDB) File * Note: for contracts whose service area only covers Puerto Rico, the weights for these measures will be zero in the summary and overall rating calculations and remain three for the improvement measure calculations. The Challenges of Pharmacy Star Ratings • 121 Appendix D: Medicare 2016 Part C & D Display Measure Technical Notes Medicare 2016 Part C & D Display Measure Technical Notes Updated – 12/17/2015 122 • AIS’s Management Insight Series Document Change Log: Previous Version - Description of Change Initial Release of the 2016 Display Measure Technical Notes (Last Updated 12/17/2015) Revision Date 12/17/2015 Page i The Challenges of Pharmacy Star Ratings • 123 Table of Contents DOCUMENT CHANGE LOG:............................................................................................................................. I GENERAL ......................................................................................................................................................... 1 CONTACT INFORMATION ............................................................................................................................... 1 PART C DISPLAY MEASURE DETAILS .......................................................................................................... 2 Measure: DMC01 - Follow-up Visit after Hospital Stay for Mental Illness (within 30 days of discharge) ........................... 2 Measure: DMC02 - Call Answer Timeliness ....................................................................................................................... 2 Measure: DMC03 - Antidepressant Medication Management (6 months) .......................................................................... 2 Measure: DMC04 - Continuous Beta Blocker Treatment.................................................................................................... 2 Measure: DMC05 - Appropriate Monitoring of Patients Taking Long-term Medications .................................................... 3 Measure: DMC06 - Osteoporosis Testing ........................................................................................................................... 3 Measure: DMC07 - Testing to Confirm Chronic Obstructive Pulmonary Disease .............................................................. 3 Measure: DMC08 - Doctors who Communicate Well ......................................................................................................... 4 Measure: DMC09 - Call Center – Beneficiary Hold Time ................................................................................................... 4 Measure: DMC10 - Pneumonia Vaccine ............................................................................................................................. 5 Measure: DMC11 - Access to Primary Care Doctor Visits.................................................................................................. 5 Measure: DMC12 - Calls Disconnected When Customer Calls Health Plan ...................................................................... 5 Measure: DMC13 - Pharmacotherapy Management of COPD Exacerbation – Systemic Corticosteroid .......................... 6 Measure: DMC14 - Pharmacotherapy Management of COPD Exacerbation – Bronchodilator ......................................... 6 Measure: DMC15 - Initiation of Alcohol or other Drug Treatment ....................................................................................... 6 Measure: DMC16 - Engagement of Alcohol or other Drug Treatment ................................................................................ 7 Measure: DMC17 - Reminders for Appointments ............................................................................................................... 7 Measure: DMC18 - Reminders for Immunizations .............................................................................................................. 7 Measure: DMC19 - Reminders for Screening Tests ........................................................................................................... 7 Measure: DMC20 - Computer Used during Office Visits .................................................................................................... 8 Measure: DMC21 - Computer Use by Doctor Helpful ......................................................................................................... 8 Measure: DMC22 - Computer Use Made Talking with Doctor Easier ................................................................................ 8 Measure: DMC23 - Improving Bladder Control ................................................................................................................... 9 PART D DISPLAY MEASURE DETAILS ........................................................................................................ 10 Measure: DMD01 - Timely Receipt of Case Files for Appeals ......................................................................................... 10 Measure: DMD02 - Timely Effectuation of Appeals .......................................................................................................... 10 Measure: DMD03 - Calls Disconnected When Customer Calls Drug Plan ....................................................................... 11 Measure: DMD04 - Call Center – Beneficiary Hold Time ................................................................................................. 11 Measure: DMD05 - Drug-Drug Interactions ...................................................................................................................... 12 Measure: DMD06 - Diabetes Medication Dosing .............................................................................................................. 12 Measure: DMD07 - Drug Plan Provides Current Information on Costs and Coverage for Medicare’s Website ............... 13 Measure: DMD08 - MPF – Stability................................................................................................................................... 14 Measure: DMD09 - Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes ............ 14 Measure: DMD10 - Getting Information from Drug Plan ................................................................................................... 15 Measure: DMD11 - Call Center – Pharmacy Hold Time ................................................................................................... 16 Measure: DMD12 - Plan Submitted Higher Prices for Display on MPF ............................................................................ 16 Measure: DMD13 - Transition monitoring - failure rate for drugs within classes of clinical concern ................................ 17 Measure: DMD14 - Transition monitoring - failure rate for all other drugs ........................................................................ 18 Measure: DMD15 - Reminders to Fill prescriptions .......................................................................................................... 18 Measure: DMD16 - Reminders to Take Medications ........................................................................................................ 18 (Last Updated 12/17/2015) Page ii 124 • AIS’s Management Insight Series COMMON PART C & D DISPLAY MEASURE DETAILS ............................................................................... 19 Measure: DME01 - Enrollment Timeliness........................................................................................................................ 19 Measure: DME02 - Grievance Rate .................................................................................................................................. 19 Measure: DME03 - Disenrollment Reasons - Problems Getting Needed Care, Coverage, and Cost Information (MA-PD, MA-only) ............................................................................................................................................................................ 20 Measure: DME04 - Disenrollment Reasons - Problems with Coverage of Doctors and Hospitals (MA-PD, MA-only) .... 21 Measure: DME05 - Disenrollment Reasons - Financial Reasons for Disenrollment (MA-PD, MA-only, PDP) ................ 22 Measure: DME06 - Disenrollment Reasons - Problems with Prescription Drug Benefits and Coverage (MA-PD, PDP) 22 Measure: DME07 - Disenrollment Reasons - Problems Getting Information about Prescription Drugs (MA-PD, PDP) .. 23 ATTACHMENT A: NATIONAL AVERAGES FOR PART C AND D DISPLAY MEASURES ........................... 24 Table A-1: National Averages for Part C Display Measures ......................................................................................... 24 Table A-2: National Averages for Part D Display Measures ......................................................................................... 24 Table A-3: National Averages for common Part C and D Display Measures ............................................................. 25 (Last Updated 12/17/2015) Page iii The Challenges of Pharmacy Star Ratings • 125 General This document describes the metric, data source and reporting time period for each Medicare Part C or Part D Display Measure. All data are reported at the contract level. The data do not reflect information for National PACE, 1833 Cost contracts, Continuing Care Retirement Community demonstrations (CCRCs), End Stage Renal Disease Networks (ESRDs), and Demonstration contracts. All other organization types are included. These display measures are not part of the Star Ratings. Display measures may have been transitioned from the Star Ratings. These can also be new measures being tested before inclusion into the Star Ratings. Lastly, some measures are displayed for informational purposes only. As indicated in the 2015 Call Letter, CMS will give advance notice if display measures are being considered for inclusion to the Star Ratings. Data for display page measures will continue to be collected and monitored, and poor scores on display measures are subject to compliance actions by CMS. For 2016, CMS is Transitioning one Star Rating measures to display: a. Improving Bladder Control (Part C) Reintroducing five display measures a. Call Center Hold Time and Disconnection measures (Part C and D) Contact Information The contact below can assist you with various aspects of the Display Measures. Part C & D Star Ratings: [email protected] If you have questions or require information about the specific subject areas associated with the Display Measures please write to those contacts directly and cc the Part C & D Star Ratings mailbox. CAHPS (MA & Part D): [email protected] Call Center Monitoring: [email protected] Disenrollment Reasons Survey: [email protected] HEDIS: [email protected] HOS: [email protected] Part C Plan Reporting: [email protected] Part D Plan Reporting: [email protected] Part C & D Plan Reporting Data Validation: [email protected] (Last Updated 12/17/2015) Page 1 126 • AIS’s Management Insight Series Part C Display Measure Details Measure: DMC01 - Follow-up Visit after Hospital Stay for Mental Illness (within 30 days of discharge) Title Description HEDIS Label: Follow-Up After Hospitalization for Mental Illness (FUH) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 177 Metric: The percentage of discharges for members 6 years of age and older who were hospitalized for treatment of selected mental health disorders (denominator) and who had an outpatient visit, an intensive outpatient encounter or partial hospitalization with a mental health practitioner within 30 days of discharge (numerator). Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC02 - Call Answer Timeliness Title Description HEDIS Label: Call Answer Timeliness (CAT) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 251 Metric: The percentage of calls received by the organization’s member services call center (during operating hours) during the measurement year that were answered by a live voice within 30 seconds. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC03 - Antidepressant Medication Management (6 months) Title Description HEDIS Label: Antidepressant Medication Management (AMM) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 168 Metric: The percentage of members 18 years of age and older with a diagnosis of major depression (denominator) who were newly treated with antidepressant medication, and who remained on an antidepressant medication treatment (numerator). Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC04 - Continuous Beta Blocker Treatment Title Description HEDIS Label: Persistence of Beta-Blocker Treatment After a Heart Attack (PBH) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 138 Metric: The percentage of members 18 years of age and older during the measurement year who were hospitalized and discharged alive from July 1 of the year prior to the measurement year to June 30 of the measurement year with a diagnosis of AMI (Last Updated 12/17/2015) Page 2 The Challenges of Pharmacy Star Ratings • 127 Title Description (denominator) and who received persistent beta-blocker treatment for six months after discharge (numerator). Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC05 - Appropriate Monitoring of Patients Taking Long-term Medications Title Description HEDIS Label: Annual Monitoring for Patients on Persistent Medication (MPM) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 202 Metric: The percentage of members 18 years of age and older who received at least 180 treatment days of ambulatory medication therapy for a select therapeutic agent (denominator) during the measurement year and at least one therapeutic monitoring event for the therapeutic agent in the measurement year (numerator). Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC06 - Osteoporosis Testing Title Description HEDIS Label: Osteoporosis Testing in Older Women (OTO) Measure Reference: NCQA HEDIS 2014 Specifications for The Medicare Health Outcomes Survey Volume 6, page 37 Metric: The percentage of Medicare women 65 years of age and older (denominator) who report ever having received a bone density test to check for osteoporosis (numerator). Exclusions: None listed. Data Source: HEDIS / HOS Data Source Description: Cohort 15 Follow-up Data collection (2014) and Cohort 17 Baseline data collection (2014). HOS Survey Question 52: Have you ever had a bone density test to check for osteoporosis, sometimes thought of as "brittle bones"? This test may have been done to your back, hip, wrist, heel or finger. Data Time Frame: 04/18/2014 - 07/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC07 - Testing to Confirm Chronic Obstructive Pulmonary Disease Title Description HEDIS Label: Use of Spirometry Testing in the Assessment and Diagnosis of COPD (SPR) Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 114 Metric: The percentage of members 40 or older with a new diagnosis or newly active Chronic Obstructive Pulmonary Disease (COPD) during the measurement year (denominator), (Last Updated 12/17/2015) Page 3 128 • AIS’s Management Insight Series Title Description who received appropriate spirometry testing to confirm the diagnosis (numerator). Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC08 - Doctors who Communicate Well Title Description Metric: This case mix adjusted composite measure is used to assess how well doctors communicate. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) score uses the mean of the distribution of responses converted to a scale from 0 to 100. The score shown is the percentage of the best possible score each contract earned. CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, how often did your personal doctor explain things in a way that was easy to understand? • In the last 6 months, how often did your personal doctor listen carefully to you? • In the last 6 months, how often did your personal doctor show respect for what you had to say? • In the last 6 months, how often did your personal doctor spend enough time with you? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC09 - Call Center – Beneficiary Hold Time Title Description Metric: This measure is defined as the average time spent on hold by the call surveyor following the navigation of the Interactive Voice Response (IVR) or Automatic Call Distributor (ACD) system and prior to reaching a live person for the “Customer Service for Current Members – Part C” phone number associated with the contract. This measure is calculated by taking the sum of the total time (mm:ss) it takes for a caller to reach a Customer Service Representative (CSR) for all eligible calls made to that Part C contract beneficiary customer service call center, divided by the number of eligible calls made to the Part C contract beneficiary customer service call center. For calls in which the caller terminated the call due to being on hold for greater than 10 minutes prior to reaching a live person, the hold time applied is truncated to 10:00 minutes. Note that total time excludes the time navigating the IVR/ACD system and thus measures only the time the caller is placed into the “hold” queue. Exclusions: Data were not collected from contracts that cover U.S territories, 1876 Cost, Employer/Union Only Direct Contract PDP, Employer/Union Only Direct Contract PFFS, National PACE, MSA, employer contracts, and organizations that did not have a phone number accessible to survey callers. Data Source: Call Center Data Source Description: Call Center surveillance monitoring data collected by CMS. The “Customer Service for Current Members – Part C” phone number associated with each contract was monitored. This measure is based on calls to the current enrollee call center. Data Time Frame: 01/19/2015 - 07/03/2015 (Last Updated 12/17/2015) Page 4 The Challenges of Pharmacy Star Ratings • 129 Title Description General Trend: Lower is better Data Display: Time Compliance Standard: 2:00 Measure: DMC10 - Pneumonia Vaccine Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported ever having received a pneumococcal vaccine (numerator). CAHPS Survey Question (question number varies depending on survey type): • Have you ever had a pneumonia shot? This shot is usually given only once or twice in a person’s lifetime and is different from a flu shot. It is also called the pneumococcal vaccine. Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC11 - Access to Primary Care Doctor Visits Title Description HEDIS Label: Adults' Access to Preventive/Ambulatory Health Services Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 233 Metric: The percentage of members 20 years and older (denominator) who had an ambulatory or preventive care visit during the measurement year (numerator). Exclusions: None listed. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Compliance Standard: 85% Measure: DMC12 - Calls Disconnected When Customer Calls Health Plan Title Description Metric: This measure is defined as the number of calls unexpectedly dropped by the sponsor while the call surveyor was navigating the IVR or connected with a customer service representative (CSR) divided by the total number of calls made to the phone number associated with the contract. Exclusions: Data were not collected from contracts that cover U.S territories, 1876 Cost, Employer/Union Only Direct Contract PDP, Employer/Union Only Direct Contract PFFS, National PACE, MSA, employer contracts, and organizations that did not have a phone number accessible to survey callers. Data Source: Call Center Data Source Description: Call Center surveillance monitoring data collected by CMS. The “Customer Service for Current Members – Part C” phone number associated with each contract was monitored. This measure is based on calls to the current enrollee call center. Data Time Frame: 01/19/2015 - 07/03/2015 (Last Updated 12/17/2015) Page 5 130 • AIS’s Management Insight Series Title Description General Trend: Lower is better Data Display: Percentage with 2 decimal points Compliance Standard: 5% Measure: DMC13 - Pharmacotherapy Management of COPD Exacerbation – Systemic Corticosteroid Title Description HEDIS Label: Pharmacotherapy Management of COPD Exacerbation Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 116 Metric: The percentage of COPD exacerbations for members 40 years of age and older who had an acute inpatient discharge or ED encounter on or between January 1–November 30 of the measurement year and who were dispensed a systemic corticosteroid within 14 days of the event. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC14 - Pharmacotherapy Management of COPD Exacerbation – Bronchodilator Title Description HEDIS Label: Pharmacotherapy Management of COPD Exacerbation Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 116 Metric: The percentage of COPD exacerbations for members 40 years of age and older who had an acute inpatient discharge or ED encounter on or between January 1–November 30 of the measurement year and who were dispensed a bronchodilator within 30 days of the event. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC15 - Initiation of Alcohol or other Drug Treatment Title Description HEDIS Label: Initiation and Engagement of Alcohol and Other Drug Dependence Treatment Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 239 Metric: The percentage of members who initiate treatment through an inpatient AOD admission, outpatient visit, intensive outpatient encounter or partial hospitalization within 14 days of the diagnosis. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 6 The Challenges of Pharmacy Star Ratings • 131 Measure: DMC16 - Engagement of Alcohol or other Drug Treatment Title Description HEDIS Label: Initiation and Engagement of Alcohol and Other Drug Dependence Treatment Measure Reference: NCQA HEDIS 2015 Technical Specifications Volume 2, page 239 Metric: The percentage of members who initiated treatment and who had two or more additional services with a diagnosis of AOD within 30 days of the initiation visit. Data Source: HEDIS Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC17 - Reminders for Appointments Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported that they were reminded about appointments (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, did anyone from a doctor’s office or your health plan contact you to remind you to make appointments for tests or treatment? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC18 - Reminders for Immunizations Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported that they were reminded about getting immunizations (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, did anyone from a doctor’s office or your health plan contact you to remind you to get a flu shot or other immunization? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC19 - Reminders for Screening Tests Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported that they were reminded about getting a screening test (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, did anyone from a doctor’s office or your health plan contact you to remind you about screening tests such as breast cancer or colorectal cancer screening? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 7 132 • AIS’s Management Insight Series Measure: DMC20 - Computer Used during Office Visits Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported their doctor used a computer or handheld device during an office visit (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • Doctors may use computers or handheld devices during an office visit to do things like look up your information or order prescription medicines. In the last 6 months, did your personal doctor use a computer or handheld device during any of your visits? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC21 - Computer Use by Doctor Helpful Title Description Metric: This case-mix adjusted measure is used to assess how helpful providers’ computer use is. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) score is the percentage of sampled Medicare enrollees (denominator) who reported that their doctor’s use of a computer or handheld device was helpful “a lot” or “a little”. CAHPS Survey Questions (question numbers vary depending on survey type): • During your visits in the last 6 months, was your personal doctor’s use of a computer or handheld device helpful to you? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMC22 - Computer Use Made Talking with Doctor Easier Title Description Metric: This case-mix adjusted measure is used to assess whether providers’ computer use made talking harder or easier. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) score is the percentage of sampled Medicare enrollees (denominator) who reported that their doctor’s use of a computer or handheld device made talking to them easier. CAHPS Survey Questions (question numbers vary depending on survey type): • During your visits in the last 6 months, did your personal doctor’s use of a computer or handheld device make it harder or easier for you to talk to him or her? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 8 The Challenges of Pharmacy Star Ratings • 133 Measure: DMC23 - Improving Bladder Control Title Description HEDIS Label: Management of Urinary Incontinence in Older Adults (MUI) Measure Reference: NCQA HEDIS 2014 Specifications for The Medicare Health Outcomes Survey Volume 6, page 29 Metric: The percentage of Medicare members 65 years of age or older who reported having a urine leakage problem in the past six months (denominator) and who received treatment for their current urine leakage problem (numerator). Exclusions: None listed. Data Source: HEDIS / HOS Data Source Description: Cohort 15 Follow-up Data collection (2014) and Cohort 17 Baseline data collection (2014). HOS Survey Question 42: Many people experience problems with urinary incontinence, the leakage of urine. In the past 6 months, have you accidentally leaked urine? Data Time Frame: 04/18/2014 - 07/31/2014 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 9 134 • AIS’s Management Insight Series Part D Display Measure Details Measure: DMD01 - Timely Receipt of Case Files for Appeals Title Description Metric: This measure is defined as the percent of case files that were requested by the IRE that were received timely from the plan. (Timely is defined as files being received from the plan within 48 hours for Standard appeals, and within 24 hours for Expedited appeals.) Numerator = The number of case files requested that were received in the required time frame. Denominator = The number of case files requested by the IRE. This is calculated as: [(The number of case files received in the required timeframe) / (The number of case files requested by the IRE)] * 100. Exclusions: None Data Source: IRE Data Source Description: Data were obtained from the IRE contracted by CMS for Part D reconsiderations. These data are limited to appeal cases requested by beneficiaries and the IRE requests files from the plans. Cases auto-forwarded to the IRE are excluded. Data Time Frame: 01/01/2015 - 06/30/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMD02 - Timely Effectuation of Appeals Title Description Metric: This measure is defined as the percent of appeals that required effectuation that the plan effectuated in a timely manner (Timely is defined as within one day of decision notification for Expedited appeals, or three days of decision notification for Standard appeals.). Numerator = The number of appeals that were effectuated timely. Denominator = The number of the dispositions which required effectuation. Appeals with a disposition of “Fully Reverse Plan” or “Partially Reverse Plan” require effectuation. This measure looks at the most recent proceeding where effectuation is required in the event of ALJ’s or Reopenings. This is calculated as: [(The number of appeals that were effectuated timely) / (The number of dispositions that required effectuation)] * 100. Exclusions: None. These data are based on the report generation date. If the IRE does not receive a notice of effectuation before the timeframe has elapsed, the IRE will count the appeal as non-timely. Discrepancies may occur if the IRE receives the effectuation notice late, despite the actual effectuation occurring timely. Re-openings and ALJ decisions may also negate the need for effectuation. Data Source: IRE Data Source Description: Data were obtained from the IRE contracted by CMS for Part D reconsiderations. Timely is defined as within one day of decision notification for Expedited appeals, or three days of decision notification for Standard appeals. For appeals involving plans (Last Updated 12/17/2015) Page 10 The Challenges of Pharmacy Star Ratings • 135 Title Description making payments, timely is defined as payment being made within 30 calendar days of decision notification. Data Time Frame: 01/01/2015 - 06/30/2015 General Trend: Higher is better Data Display: Percentage with 2 decimal points Measure: DMD03 - Calls Disconnected When Customer Calls Drug Plan Title Description Metric: This measure is defined as the number of calls unexpectedly dropped by the sponsor while the call surveyor was navigating the IVR or connected with a customer service representative (CSR) divided by the total number of calls made to the phone number associated with the contract. Exclusions: Data were not collected from contracts that cover U.S. territories, 1876 Cost, Employer/Union Only Direct Contract PDP, Employer/Union Only Direct Contract PFFS, National PACE, MSA, employer contracts, and organizations that did not have a phone number accessible to survey callers. Data Source: Call Center Data Source Description: Call Center surveillance monitoring data collected by CMS. The “Customer Service for Current Members – Part D” phone number associated with each contract was monitored. This measure is based on calls to the current enrollee call center. Data Time Frame: 01/19/2015 - 07/03/2015 General Trend: Lower is better Data Display: Percentage with 2 decimal points Compliance Standard: 5% Measure: DMD04 - Call Center – Beneficiary Hold Time Title Description Metric: This measure is defined as the average time spent on hold by a call surveyor following the navigation of the Interactive Voice Response (IVR) or Automatic Call Distributor (ACD) system and prior to reaching a live person for the “Customer Service for Current Members – Part D” phone number associated with the contract. This measure is calculated by taking the sum of the total time (mm:ss) it takes for a caller to reach a Customer Service Representative (CSR) for all eligible calls made to that Part D contract beneficiary customer service call center divided by the number of eligible calls made to the Part D contract beneficiary customer service call center. For calls in which the caller terminated the call due to being on hold for greater than 10 minutes prior to reaching a live person, the hold time applied is truncated to 10:00 minutes. Note that total time excludes the time navigating the IVR/ACD system and thus measures only the time the caller is placed into the “hold” queue. Exclusions: Data were not collected from contracts that cover U.S. territories, 1876 Cost, Employer/Union Only Direct Contract PDP, Employer/Union Only Direct Contract PFFS, National PACE, MSA, employer contracts, and organizations that did not have a phone number accessible to survey callers. Data Source: Call Center Data Source Description: Call center monitoring data collected by CMS. The “Customer Service for Current Members – Part D” phone number associated with each contract was monitored. Data Time Frame: 01/19/2015 - 07/03/2015 General Trend: Lower is better (Last Updated 12/17/2015) Page 11 136 • AIS’s Management Insight Series Title Description Data Display: Time Compliance Standard: 2:00 Measure: DMD05 - Drug-Drug Interactions Title Description Metric: This measure is defined as the percent of Medicare Part D beneficiaries who received a prescription for a target medication during the measurement period and who were dispensed a prescription for a contraindicated medication with or subsequent to the initial prescription. Numerator = Number of member-years of beneficiaries enrolled during the measurement period who were dispensed a target medication with at least one day overlap with a contraindicated medication. Denominator = Number of member-years of beneficiaries enrolled during the measurement period who were dispensed a target medication. This is calculated as: [(Number of member-years of beneficiaries enrolled during the measurement period who were dispensed a target medication with at least one day overlap with a contraindicated medication) / (Number of member-years of beneficiaries enrolled during the measurement period who were dispensed a target medication)]*100. Exclusions: A percentage is not calculated for contracts with 30 or fewer beneficiary member years (in the denominator). Data Source: PDE data Data Source Description: The Drug-Drug Interaction (DDI) measure is adapted from the measure concept that was first developed by the Pharmacy Quality Alliance (PQA). The data for this measure come from Prescription Drug Event (PDE) data files submitted by drug plans to Medicare for dates of service from January 1, 2014-December 31, 2014, and processed by June 30, 2015. Only final action PDE claims are used to calculate the patient safety measures. PDE adjustments made post-reconciliation were not reflected in this measure. The measure is calculated using the National Drug Code (NDC) lists updated by the PQA. The complete NDC lists will be posted along with these technical notes. Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with 1 decimal point Measure: DMD06 - Diabetes Medication Dosing Title Description Metric: This measure is defined as the percent of Medicare Part D beneficiaries who were dispensed a dose higher than the daily recommended dose for the following diabetes treatment therapeutic categories of oral hypoglycemics: biguanides, sulfonlyureas, thiazolidinediones, and DiPeptidyl Peptidase (DPP)-IV inhibitors. Numerator = Number of member-years of beneficiaries 18 years and older enrolled during the measurement period who were dispensed a dose of an oral hypoglycemic higher than the daily recommended dose. Denominator = Number of member-years of beneficiaries 18 years and older enrolled during the measurement period who were dispensed at least one prescription of an oral hypoglycemic. (Last Updated 12/17/2015) Page 12 The Challenges of Pharmacy Star Ratings • 137 Title Description This is calculated as: [(Number of member-years of beneficiaries 18 years and older enrolled during the measurement period who were dispensed a dose of an oral hypoglycemic higher than the daily recommended dose) / (Number of member-years of beneficiaries 18 years and older enrolled during the measurement period who were dispensed at least one prescription of an oral hypoglycemic)]*100. Exclusions: A percentage is not calculated for contracts with 30 or fewer beneficiary member years (in the denominator). Data Source: PDE data Data Source Description: The Diabetes Medication Dosing (DMD) measure is adapted from the measure concept that was first developed by the Pharmacy Quality Alliance (PQA). The data for this measure come from Prescription Drug Event (PDE) data files submitted by drug plans to Medicare for dates of service from January 1, 2014-December 31, 2014, and processed by June 30, 2015. Only final action PDE claims are used to calculate the patient safety measures. PDE adjustments made post-reconciliation were not reflected in this measure. The measure is calculated using the National Drug Code (NDC) lists updated by the PQA. The complete NDC lists will be posted along with these technical notes. Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with 2 decimal points Measure: DMD07 - Drug Plan Provides Current Information on Costs and Coverage for Medicare’s Website Title Description Metric: This measure is defined as percent of pricing/formulary data file submissions that do not result in suppression of pricing data on www.medicare.gov. Numerator = Number of pricing data file submissions that do not result in suppression of pricing data on www.medicare.gov Denominator = Total number of pricing data submissions This is calculated as: [(Number of pricing data file submissions that do not result in suppression of pricing data on www.medicare.gov) / (Total number of pricing data submissions)]*100. Exclusions: None. Data Source: CMS Administrative Data Data Time Frame: 10/01/2014 - 09/30/2015 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 13 138 • AIS’s Management Insight Series Measure: DMD08 - MPF – Stability Title Description Metric: This measure evaluates stability in a plan’s point of sale prices. The stability price index uses final prescription drug event (PDE) data to assess changes in prices over the contract year. It is defined as the average change in price of a specified basket of drugs each quarter. A basket of drugs defined by quarter 1 PDEs is priced using quarter 1 average prices for each drug first. The same basket is then priced using quarter 2 average prices. The stability price index from quarter 1 to quarter 2 is calculated as the total price of the basket using the quarter 2 average prices divided by the total price of same basket using quarter 1 average prices. This same process is repeated using a quarter 2 basket of drugs to compute the quarter 2 to quarter 3 price index and a quarter 3 basket of drugs to compute the quarter 3 to quarter 4 price index. The overall stability price index is the average of the price index from quarter 1 to 2, quarter 2 to 3, and quarter 3 to 4. A price index of 1 indicates a plan had no increase in prices from the beginning to the end of the year. A stability index smaller than 1 indicates that prices decreased, while an index greater than 1 indicates that prices increased. To convert the index into the stability score, we use the formula below. The score is rounded to the nearest whole number. 100 – ((stability index – 1) x 100). Exclusions: A contract must have at least one drug with at least 10 claims in each quarter for the price stability index. PDEs must also meet the following criteria: • Pharmacy number on PDE must appear in MPF pharmacy cost file • PDE must be for retail pharmacy • Date of service must occur at a time that data are not suppressed for the plan on MPF • PDE must not be a compound claim • PDE must not be a non-covered drug Data Source: PDE data, MPF Pricing Files, HPMS approved formulary extracts, and data from First DataBank and Medi-span Data Source Description: Data were obtained from a number of sources: PDE data, MPF Pricing Files, HPMS approved formulary extracts. Post-reconciliation PDE adjustments are not reflected in this measure Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Higher is better Data Display: Rate with no decimal point Measure: DMD09 - Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes Title Description Metric: This measure is defined as the percent of Medicare Part D beneficiaries 65 years and older who are continuously enrolled in a nursing home and who received atypical antipsychotic (AA) medication fills during the period measured. Denominator = Number of beneficiaries who meet all of the following: • Had Long-Term Institutional (LTI) status * for all months of the measurement period or until death, • Were alive for at least 90 days at the beginning of the measurement period, • Were enrolled in Part D for all months of the measurement period that they were alive, (Last Updated 12/17/2015) Page 14 The Challenges of Pharmacy Star Ratings • 139 Title Description and • Whose first reason for Medicare enrollment was aging-in. Numerator = Number of Part D beneficiaries in the denominator who received at least a 90 day supply of AA medication(s) during the nursing home stay in the measurement period This rate is calculated using a list of AA National Drug Codes (NDC) maintained by CMS. The complete medication list will be posted along with these technical notes. * See Notes under Data Source for definition of LTI Exclusions: A percentage is not calculated for contracts with 10 or fewer beneficiaries in the denominator and will be shown as “No Data Available.” Data Source: PDE data, Enrollment data, Minimum Date Set( (MDS) Assessments Data Source Description: Data Source: Prescription Drug Event (PDE) data, Enrollment data, Minimum Data Set (MDS) Assessments Notes: Beneficiaries are defined as LTI for payment purposes under the Medicare Risk Adjustment program. The algorithm that creates monthly flags for each LTI-defined beneficiary is described below. Monthly LTI flags are created to identify, by month, a beneficiary’s institutional versus community status. The flags are used to determine the appropriate CMS- risk scores for calculating Part C and Part D risk payments, and for resolving risk scores for analysis purposes. The monthly LTI flags are created based on an analysis of MDS assessments. A nursing home resident (beneficiary) is stepped through their MDS assessments chronologically. For each month, if a quarterly, annual, or significant change assessment is encountered and the nursing home length of stay on the date of that assessment is more than 90 days, then an LTI flag is turned on for the following month. An LTI flag is established for all subsequent months until the beneficiary dies, a discharge assessment is encountered, or if an assessment is not encountered within 150 days of a prior assessment. Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with 2 decimal points Measure: DMD10 - Getting Information from Drug Plan Title Description Metric: This case-mix adjusted composite measure is used to assess how easy it is for members to get information from the plan about prescription drug coverage and cost. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) score uses the mean of the distribution of responses converted to a scale from 0 to 100. The score shown is the percentage of the best possible score each contract earned. CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, how often did your health plan’s customer service give you the information or help you needed about prescription drugs? • In the last 6 months, how often did your plan's customer service staff treat you with courtesy and respect when you tried to get information or help about prescription drugs? (Last Updated 12/17/2015) Page 15 140 • AIS’s Management Insight Series Title Description • In the last 6 months, how often did your health plan give you all the information you needed about which prescription medicines were covered? • In the last 6 months, how often did your health plan give you all the information you needed about how much you would have to pay for your prescription medicine? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMD11 - Call Center – Pharmacy Hold Time Title Description Metric: This measure is defined as the average time spent on hold by a call surveyor following the navigation of the Interactive Voice Response (IVR) or Automatic Call Distributor (ACD) system and prior to reaching a live person for the “Pharmacy Technical Help Desk” phone number associated with the contract. This measure is calculated by taking the sum of the total time (mm:ss) it takes for a caller to reach a Customer Service Representative (CSR) for all eligible calls made to that Part D contract pharmacy technical help desk divided by the number of eligible calls made to the Part D contract pharmacy technical help desk. For calls in which the caller terminated the call due to being on hold for greater than 10 minutes prior to reaching a live person, the hold time applied is truncated to 10:00 minutes. Note that total time excludes the time navigating the IVR/ACD system and thus measures only the time the caller is placed into the “hold” queue. Exclusions: Data were not collected from contracts that cover U.S. territories, 1876 Cost, Employer/Union Only Direct Contract PDP, Employer/Union Only Direct Contract PFFS, National PACE, MSA, employer contracts, and organizations that did not have a phone number accessible to survey callers. Data Source: Call Center Data Source Description: Call center data collected by CMS. The Pharmacy Technical Help Desk phone number associated with each contract was monitored. Data Time Frame: 01/18/2015 - 07/02/2015 General Trend: Lower is better Data Display: Time Compliance Standard: 2:00 Measure: DMD12 - Plan Submitted Higher Prices for Display on MPF Title Description Metric: This measure evaluates the accuracy of drug prices posted on the MPF tool. A contract’s score is based on the accuracy index. The accuracy price index compares point-of-sale PDE prices to plan-reported MPF prices and determines the magnitude of differences found. Using each PDE’s date of service, the price displayed on MPF is compared to the PDE price. The accuracy index considers both ingredient cost and dispensing fee and measures the amount that the MPF price is higher than the PDE price. Therefore, prices that are understated on MPF—that is, the reported price is lower than the actual price—will not count against a plan’s accuracy score. (Last Updated 12/17/2015) Page 16 The Challenges of Pharmacy Star Ratings • 141 Title Description The index is computed as: (Total amount that PF is higher than PDE + Total PDE cost) / (Total PDE cost). The best possible accuracy index is 1. An index of 1 indicates that a plan did not have PDE prices less than MPF prices. A contract’s score is computed using its accuracy index as: 100 – ((accuracy index - 1) x 100). Exclusions: A contract must have at least 30 claims over the measurement period for the price accuracy index. PDEs must also meet the following criteria: • Pharmacy number on PDE must appear in MPF pharmacy cost file • Drug must appear in formulary file and in MPF pricing file • PDE must be for retail and/or specialty pharmacy • PDE must be a 30 day supply • Date of service must occur at a time that data are not suppressed for the plan on MPF • PDE must not be a compound claim • PDE must not be a non-covered drug • PDE must be for retail pharmacy (pharmacies marked retail and mail order/HI/LTC are excluded) Data Source: PDE data, MPF Pricing Files, HPMS approved formulary extracts, and data from First DataBank and Medi-span Data Time Frame: 01/01/2014 - 09/30/2014 General Trend: Higher is better Data Display: Rate with no decimal point Measure: DMD13 - Transition monitoring - failure rate for drugs within classes of clinical concern Title Description Metric: The numbers of failures (numerator) were divided by the number of claims sampled (denominator) to calculate an overall compliance score. If the number of failures resulted in more than a 10% failure rate, CMS determined that an overall compliance failure occurred for this area. Exclusions: Contracts with fewer than 15 claims sampled; Contracts not listed in active status in HPMS; MMPs that did not have a start date on or before January 2015; Contracts that are involved in other transition oversight activities; Contracts that do not offer Part D coverage or did not utilize a formulary. Data Source: Part D Sponsor, PDE data, CME data and HPMS approved formularies Data Source Description: Data was obtained from the Part D Sponsor, PDE data, CME data and HPMS approved formulary extracts. Data Time Frame: January 4 – 24, 2015 General Trend: Lower is better Data Display: Percentage with 1 decimal point Compliance Standard: >10% (Last Updated 12/17/2015) Page 17 142 • AIS’s Management Insight Series Measure: DMD14 - Transition monitoring - failure rate for all other drugs Title Description Metric: The numbers of failures (numerator) were divided by the number of claims sampled (denominator) to calculate an overall compliance score. If the number of failures resulted in more than a 20% failure rate, CMS determined that an overall compliance failure occurred for this area. Exclusions: Contracts with fewer than 15 claims sampled; Contracts not listed in active status in HPMS; MMPs that did not have a start date on or before January 2015; Contracts that are involved in other transition oversight activities; Contracts that do not offer Part D coverage or did not utilize a formulary. Data Source: Part D Sponsor, PDE data, CME data, and HPMS approved formularies Data Source Description: Data was obtained from the Part D Sponsor, PDE data, CME data, and HPMS approved formulary extracts. Data Time Frame: January 4 – 24, 2015 General Trend: Lower is better Data Display: Percentage with 1 decimal point Compliance Standard: >20% Measure: DMD15 - Reminders to Fill prescriptions Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported that they were reminded about filling or refilling a prescription (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, did anyone from a doctor’s office, pharmacy or your prescription drug plan contact you to make sure you filled or refilled a prescription? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DMD16 - Reminders to Take Medications Title Description Metric: The percentage of sampled Medicare enrollees (denominator) who reported that they were reminded about taking medications as directed (numerator). CAHPS Survey Questions (question numbers vary depending on survey type): • In the last 6 months, did anyone from a doctor’s office, pharmacy or your prescription drug plan contact you to make sure you were taking medications as directed? Data Source: CAHPS Data Time Frame: 02/15/2015 - 05/31/2015 General Trend: Higher is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 18 The Challenges of Pharmacy Star Ratings • 143 Common Part C & D Display Measure Details Measure: DME01 - Enrollment Timeliness Title Description Metric: Numerator = The number of plan generated enrollment transactions submitted to CMS within 7 calendar days of the application date Denominator = The total number of plan generated enrollment transactions submitted to CMS Calculation = [(The number of plan generated enrollment transactions submitted to CMS within 7 calendar days of the application date) / (The total number of plan generated enrollment transactions submitted to CMS)] * 100 Exclusions: 1. Contracts with 25 or fewer enrollment submissions during the measurement period, when summed. 2. Election Types: ICEP, IEP, IEP2 and AEP. 3. Employer/Union enrollments. 4. 1876 Cost Contract MA-only members. 5. Special Needs Plans. 6. Transaction Reply Codes 1-5 (TRC1, TRC2, TRC3, TRC4, TRC5) equal to any of the below: TRC’s: ('001', '002', '003', '004', '006', '007', '008', '009', '019', '020', '032', '033', '034', '035', '036', '037', '038', '039', '042', '044', '045', '048', '056', '060', '062', '102', '103', '104', '105', '106', '107', '108', '109', '110', '114', '116', '122', '123', '124', '126', '127', '128', '129', '130', '133', '139', '156', '157', '162', '166', '169', '176', '184', '196', '200', '201', '202', '203', '211', '220', '257', '258', '263', '600', '601', '602', '603', '605', '611') TRCs are defined in the Plan Communication Users Guide Appendix Table I-2. Data Source: Medicare Advantage and Prescription Drug System (MARx) Data Source Description: The data timeframe is the monthly enrollment files for January - June, 2015, which represents submission dates of 01/01/2015 - 06/30/2015. Data Time Frame: 01/01/2015 - 06/30/2015 General Trend: Higher is better Data Display: Percentage with no decimal point Measure: DME02 - Grievance Rate Title Description Metric: This measure is defined as the number of grievances filed with the health plan per 1,000 enrollees per month. Numerator = (Quarter 1 Total Grievances + Quarter 2 Grievances + Quarter 3 Grievances + Quarter 4 Grievances) * 1,000 * 30 Denominator = Average Enrollment * Number of days in period For MAOs, Total Grievances includes grievances reported per the Part C Reporting Requirements. For PDPs, Total Grievances includes grievances reported per the Part D Reporting Requirements. For MA-PDs, Part C and Part D grievances are combined in order to report a single contract-level rate. Exclusions: Part C grievances reported in the “CMS issues” category (Element 5.10: CMS issues grievances) are excluded from the Total Grievances count. Part D grievances reported in the “CMS issues” category (Element T: CMS issues grievances) are excluded from the Total Grievances count. A contract must have an average enrollment of 800 or more enrollees to have a rate calculated. Contracts with fewer than 800 enrollees are listed as "Plan too small to be (Last Updated 12/17/2015) Page 19 144 • AIS’s Management Insight Series Title Description measured.” Contracts and plans with an effective terminate date on or before the deadline to submit data validation results to CMS (June 30, 2014) are listed as “Plan not required to report measure.” Rates are not calculated for contracts that did not score at least 95% on data validation for the Grievances reporting section(s). Rates are also not calculated for contracts that scored 95% or higher on data validation for Grievance section(s) but that were not compliant with data validation standards/sub-standards for at least one of the following Grievance data elements: Part C (MA only and MA-PDs) • Fraud grievances (Element 5.1) • Enrollment/disenrollment grievances (Element 5.2) • Benefit package grievances (Element 5.3) • Access grievances (Element 5.4) • Marketing grievances (Element 5.5) • Customer service grievances (Element 5.6) • Privacy issue grievances (Element 5.7) • Quality of care grievances (Element 5.8) • Appeals grievances (Element 5.9) • Other grievances (Element 5.11) Part D (PDPs and MA-PDs) • Enrollment, plan benefits, or pharmacy access – Total number of grievances (Element A) • Customer Service – Total number of grievances (Element C) • Coverage determinations and Redeterminations process – Total number of grievances (Element E) • Other – Total number of grievances (Element I) These contracts excluded from the measure due to data validation issues are shown as “Data issues found.” Data Source: Part C & D Plan Reporting Data Source Description: Data were reported by contracts to CMS through the Health Plan Management System (HPMS). Validation of these data was performed retrospectively during the 2014 Data Validation cycle. Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Rate with 2 decimal points Measure: DME03 - Disenrollment Reasons - Problems Getting Needed Care, Coverage, and Cost Information (MA-PD, MA-only) Title Description Metric: “Problems Getting Needed Care, Coverage, and Cost Information” is a composite of the following survey questions (question numbers vary depending on survey type): (a) Did you leave the plan because you were frustrated by the plan’s approval process for care, tests, or treatment? (b) Did you leave the plan because you had problems getting the care, tests, or treatment you needed? (c) Did you leave the plan because you had problems getting the plan to pay a claim? (Last Updated 12/17/2015) Page 20 The Challenges of Pharmacy Star Ratings • 145 Title Description (d) Did you leave the plan because it was hard to get information from the plan -- like which health care services were covered or how much a specific test or treatment would cost? Each of these questions asked about a reason for disenrollment that was related to the beneficiary’s experiences with getting needed health care services and cost information and getting claims paid for these services. Scores range from 0 to 100 and a lower mean indicates that problems getting needed care, coverage and cost information reasons were endorsed less frequently by disenrollees from your contract. Exclusions: Contracts with less than 30 responses are excluded. Data Source: Disenrollment Reasons Survey Data Source Description: Survey of members who disenrolled from the contract during the measurement time frame with the following disenrollment reason codes: disenrollment reason codes: 11 - Voluntary Disenrollment through plan, 13 - Disenrollment because of enrollment in another Plan, 14 - Retroactive or 99 - Other (not supplied by beneficiary). Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with no decimal point Measure: DME04 - Disenrollment Reasons - Problems with Coverage of Doctors and Hospitals (MA-PD, MA-only) Title Description Metric: “Problems with Coverage of Doctors and Hospitals” is a composite of the following survey questions (question numbers vary depending on survey type): (a) Did you leave the plan because the doctors or other health care providers you wanted to see did not belong to the plan? (b) Did you leave the plan because clinics or hospitals you wanted to go to for care were not covered by the plan? Each of these questions asked about a reason for disenrollment that was related to the coverage of doctors and hospitals by the plan. Scores range from 0 to 100 and a lower mean indicates that problems with coverage of doctors and hospitals reasons were endorsed less frequently by disenrollees from your contract. Exclusions: Contracts with less than 30 responses are excluded. Data Source: Disenrollment Reasons Survey Data Source Description: Survey of members who disenrolled from the contract during the measurement time frame with the following disenrollment reason codes: disenrollment reason codes: 11 - Voluntary Disenrollment through plan, 13 - Disenrollment because of enrollment in another Plan, 14 - Retroactive or 99 - Other (not supplied by beneficiary). Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 21 146 • AIS’s Management Insight Series Measure: DME05 - Disenrollment Reasons - Financial Reasons for Disenrollment (MA-PD, MA-only, PDP) Title Description Metric: “Financial Reasons for Disenrollment” is a composite of the following survey questions (question numbers vary depending on survey type): (a) Did you leave the plan because the monthly fee that the health plan charges to provide coverage for health care and prescription medicines went up? (b) Did you leave the plan because the dollar amount you had to pay each time you filled or refilled a prescription went up? (c) Did you leave the plan because you found a health plan that costs less? (d) Did you leave the plan because a change in your personal finances meant you could no longer afford the plan? Each of these questions asked about a reason for disenrollment that was related to the cost or affordability of services. Scores range from 0 to 100 and a lower mean indicates that financial reasons were endorsed less frequently by disenrollees from your contract. Exclusions: Contracts with less than 30 responses are excluded. Data Source: Disenrollment Reasons Survey Data Source Description: Survey of members who disenrolled from the contract during the measurement time frame with the following disenrollment reason codes: disenrollment reason codes: 11 - Voluntary Disenrollment through plan, 13 - Disenrollment because of enrollment in another Plan, 14 - Retroactive or 99 - Other (not supplied by beneficiary). Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with no decimal point Measure: DME06 - Disenrollment Reasons - Problems with Prescription Drug Benefits and Coverage (MA-PD, PDP) Title Description Metric: “Problems with Prescription Drug Benefits and Coverage” is a composite of the following survey questions (question numbers vary depending on survey type): (a) Did you leave the plan because they changed the list of prescription medicines they cover? (b) Did you leave the plan because the plan refused to pay for a medicine your doctor prescribed? (c) Did you leave the plan because you had problems getting the medicines your doctor prescribed? (d) Did you leave the plan because it was difficult to get brand name medicines? (e) Did you leave the plan because you were frustrated by the plan’s approval process for medicines your doctor prescribed that were not on the plan’s list of medicines that the plan covers? Each of these questions asked about a reason for disenrollment that was related to prescription drug benefits and coverage. Scores range from 0 to 100 and a lower mean indicates that problems with prescription drug benefits and coverage reasons were endorsed less frequently by disenrollees from your contract. Exclusions: Contracts with less than 30 responses are excluded. Data Source: Disenrollment Reasons Survey Data Source Description: Survey of members who disenrolled from the contract during the measurement time frame with the following disenrollment reason codes: disenrollment reason codes: 11 - Voluntary Disenrollment through plan, 13 - Disenrollment because of enrollment in (Last Updated 12/17/2015) Page 22 The Challenges of Pharmacy Star Ratings • 147 Title Description another Plan, 14 - Retroactive or 99 - Other (not supplied by beneficiary). Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with no decimal point Measure: DME07 - Disenrollment Reasons - Problems Getting Information about Prescription Drugs (MA-PD, PDP) Title Description Metric: “Problems Getting Information about Prescription Drugs” a composite of the following survey questions (question numbers vary depending on survey type): (a) Did you leave the plan because you did not know whom to contact when you had a problem filling or refilling a prescription? (b) Did you leave the plan because it was hard to get information from the plan -- like which prescription medicines were covered or how much a specific medicine would cost? (c) Did you leave the plan because you were unhappy with how the plan handled a question or complaint? (d) Did you leave the plan because you could not get the information or help you needed from the plan? (e) Did you leave the plan because their customer service staff did not treat you with courtesy and respect? Each of these questions asked about a reason for disenrollment that was related to the beneficiary’s experiences with getting information about prescription drugs. Scores range from 0 to 100 and a lower mean indicates that problems with getting information about prescription drug reasons were endorsed less frequently by disenrollees from your contract. Exclusions: Contracts with less than 30 responses are excluded. Data Source: Disenrollment Reasons Survey Data Source Description: Survey of members who disenrolled from the contract during the measurement time frame with the following disenrollment reason codes: disenrollment reason codes: 11 - Voluntary Disenrollment through plan, 13 - Disenrollment because of enrollment in another Plan, 14 - Retroactive or 99 - Other (not supplied by beneficiary). Data Time Frame: 01/01/2014 - 12/31/2014 General Trend: Lower is better Data Display: Percentage with no decimal point (Last Updated 12/17/2015) Page 23 148 • AIS’s Management Insight Series Attachment A: National Averages for Part C and D Display Measures The tables below contain the average of the numeric values for each measure reported in the 2016 Display measures. Table A-1: National Averages for Part C Display Measures Measure ID Measure Name Average DMC01 Follow-up Visit after Hospital Stay for Mental Illness (within 30 days of discharge) 55% DMC02 Call Answer Timeliness 82% DMC03 Antidepressant Medication Management (6 months) 56% DMC04 Continuous Beta Blocker Treatment 90% DMC05 Appropriate Monitoring of Patients Taking Long-term Medications 91% DMC06 Osteoporosis Testing 75% DMC07 Testing to Confirm Chronic Obstructive Pulmonary Disease 36% DMC08 Doctors who Communicate Well 91% DMC09 Call Center – Beneficiary Hold Time 0:20 DMC10 Pneumonia Vaccine 69% DMC11 Access to Primary Care Doctor Visits DMC12 Calls Disconnected When Customer Calls Health Plan DMC13 Pharmacotherapy Management of COPD Exacerbation – Systemic Corticosteroid 72% DMC14 Pharmacotherapy Management of COPD Exacerbation – Bronchodilator 80% DMC15 Initiation of Alcohol or other Drug Treatment 33% DMC16 Engagement of Alcohol or other Drug Treatment DMC17 Reminders for Appointments 59% DMC18 Reminders for Immunizations 45% DMC19 Reminders for Screening Tests 39% DMC20 Computer Used during Office Visits 85% DMC21 Computer Use by Doctor Helpful 94% DMC22 Computer Use Made Talking with Doctor Easier 53% DMC23 Improving Bladder Control 35% 95% 1.62% 3% Table A-2: National Averages for Part D Display Measures Measure ID Measure Name Average DMD01 Timely Receipt of Case Files for Appeals DMD02 Timely Effectuation of Appeals DMD03 Calls Disconnected When Customer Calls Drug Plan DMD04 Call Center – Beneficiary Hold Time 0:20 DMD05 Drug-Drug Interactions 5.7% DMD06 Diabetes Medication Dosing 0.66% DMD07 Drug Plan Provides Current Information on Costs and Coverage for Medicare’s Website 100% DMD08 MPF – Stability DMD09 Rate of Chronic Use of Atypical Antipsychotics by Elderly Beneficiaries in Nursing Homes DMD10 Getting Information from Drug Plan 82% DMD11 Call Center – Pharmacy Hold Time 0:09 DMD12 Plan Submitted Higher Prices for Display on MPF DMD13 Transition monitoring - failure rate for drugs within classes of clinical concern 2.0% DMD14 Transition monitoring - failure rate for all other drugs 3.2% DMD15 Reminders to Fill prescriptions 38% DMD16 Reminders to Take Medications 22% (Last Updated 12/17/2015) 86% 92.73% 1.65% 99 17.54% 97 Page 24 The Challenges of Pharmacy Star Ratings • 149 Table A-3: National Averages for common Part C and D Display Measures Measure ID Measure Name Average DME01 Enrollment Timeliness DME02 Grievance Rate 2.97 DME03 Disenrollment Reasons - Problems Getting Needed Care, Coverage, and Cost Information (MA-PD, MA-only) 17% DME04 Disenrollment Reasons - Problems with Coverage of Doctors and Hospitals (MA-PD, MA-only) 26% DME05 Disenrollment Reasons - Financial Reasons for Disenrollment (MA-PD, MA-only, PDP) 28% DME06 Disenrollment Reasons - Problems with Prescription Drug Benefits and Coverage (MA-PD, PDP) 11% DME07 Disenrollment Reasons - Problems Getting Information about Prescription Drugs (MA-PD, PDP) 12% (Last Updated 12/17/2015) 95% Page 25 News and Analysis of Medicare Advantage, Medicare Part D and Managed Medicaid March 27, 2014 Volume 20, Number 6 Medicaid D and Managed age, Medicare Part is of Medicare Advant News and Analys Contents Senate Passes Patients Deal to Serve MA CareMore-Emory erg’ in New Linkages a Welliceberg” and that May Be ‘Tip of Iceb 3 Oklahoma Bare-Bones Medicaid Medicare Advantage News is your most reliable source for valuable strategic information to boost revenues, increase enrollees, cut costs and improve outcomes. calls the “tip of the System unit, industry observer In a move that one for its CareMore Health serve e terms a “new strategy” 19 that they will work together to Point, Inc. executiv tells MAN Healthcare said March r, CareMore’s CEO CareMore and Emory Georgia. Moreove ents and exge beneficiaries in similar arrangem Medicare Advanta other states about negotiations in five weeks.” coming “the in that her unit is in place not like additional deal in Leeba Lessin are pects to have one by CareMore CEO plans MA outlined start ents s provider The kinds of arrangem which insurers helped alliance is in recent years in nding in the Emory deals that surfaced Lessin, the understa three existing 1). Indeed, notes be working with (MAN 4/25/13, p. and Emory instead will of Georgia unit — start an MA plan. that Emory won’t Cross and Blue Shield with it. g WellPoint’s Blue share risk and savings and e MA plans — includin CareMor services from t’s prior managewill both purchase than what WellPoin 6/16/11, p. role for CareMore in June 2011 (MAN While this is a different acquire CareMore t rates and when it agreed to MA plan paymen declining of s ent ment envisioned observer in the current environm ties in MA, Lessin and industry 3), it makes sense care Spe-insurer for close provider institutional or chronic heightened need starting new MA tells MAN. lead to CareMore as early as 2016, Lessin agree. And it could next January, with provider systems to get its first patients cial Needs Plans (SNPs) ent, which is slated Emory does ints. arrangem standpo new the She says eous from several continued on p. 6 Emory is advantag with highly regarded Since 1994, thousands of busy health plan executives and other health care managers have received reliable intelligence from Medicare Advantage News. Also Available from AIS: Published biweekly, the • Health Plan Week newsletter is the health • AIS’s Directory of Health Plans care industry’s #1 source of • Drug Benefit News timely news and business • Medicare-Medicaid Dual Eligibles strategies about MA plans, Database product design, marketing, enrollment, market expansions, CMS audits, and countless Visit www.AISHealth.com or call federal initiatives in this hotly contested area of health 800-521-4323 for more information insurance. Managed Care Pilot Bill Suggests Employer 4 MedPAC Group MA Plans Should Bid Like Other MA Plans n of Compariso Plans 4 Table: Employer Group MA And Nonemployer Plans Plan Uses Many MA Efforts 5 Tufts Stars Improvement With Providers CMS Pullback, PDPs 6 Despite Second Enhanced May Nearly Disappear 8 News Briefs Looking for a back issue of MAN? PDF issues, plus a searchable article on database, are archived Web your subscriber-only back page — all the way to 2008! Log in at www. click AISHealth.com and in the on the newsletter title ions” gray “My Subscript box on the right. Managing Editor James Gutman om [email protected] Associate Editor BJ Taylor Executive Editor Jill Brown Duals yed OK for CMS Mich. Awaits DelaSlowly in Starting Its Own Demo; Ariz. Goes of programs for Medicare-Medicaid dual eligi-l a Michigan proposa future direction An indicator of the whether to approve care despite the when CMS will decide from physical health bles could come soon al health care separate Michigan selected eight health plans g” that would keep behavior , p. 6). When the process of “finalizin care (MAN 11/21/13 goal of integrated state said it was in months last November, the But more than four CMS. with for its duals initiative (MOU) obstacle has of Understanding tell MAN the major its Memorandum industry insiders and MOU, no later, there has been a facn of behavioral health. n already has been separatio been the separatio this to large out issues related nt of duals in two The delay in working date for passive enrollme the Oct. 1 the projected start 1. And now even tor in pushing back out ) from July 1 to Oct. The executive points (Wayne and Macomb Michigan counties plan executive says. by summer. significant risk,” one would need to occur start date “is at very date, CMS site visits new Michigan Medicaid that to make the October complicated by a the situation is further to beneficiaries, with the aid of health In the meantime, costs some that aims to shift redesign program om • www.AISHealth.c DC • 800-521-4323 , Inc., Washington, nts or associations vendors, consulta Information Services Published by Atlantic publication not affiliated with insurers, An independent Start your no-risk newsletter subscription today! To order, use the form below, or call 800-521-4323. 4 Yes, I need timely news and proven strategies on MA, ❏ Part D and managed Medicaid. Please start my 1-year (24-issue) subscription to Medicare Advantage News. Name �������������������������������������������������������������������� Title ���������������������������������������������������������������������� Organization��������������������������������������������������������������� Address (no PO boxes, please)������������������������������������������������� ������������������������������������������������������������������������� City/State/Zip ������������������������������������������������������������� Phone__________________________________ Fax ������������������������������� Email (optional)������������������������������������������������������������ Three Convenient Ways to Order 1. C all toll-free 800-521-4323 (202-775-9008 in D.C.) 2. Fill out this order form and mail it to: AIS 1100 17th St, NW, Suite 300, Washington, DC 20036 3. V isit the MarketPlace at www.AISHealth.com for secure online ordering Three Payment Options ❏ Check enclosed for $586 Make checks payable to Atlantic Information Services, Inc. D.C. residents add 5.75% sales tax. ❏ Charge my credit card $586: AmEx Visa MC *Required for electronic delivery Card #��������������������������������������������������� Signature and phone are required to process credit card orders Exp. date������������������������������������������������� Signature����������������������������������������������������������������� Phone�������������������������������������������������������������������� Charges will appear as Atlantic Information Services, Inc. ❏ Bill Me $636 11PUB/0616 Atlantic Information Services • 1100 17th Street, NW, Suite 300 • Washington, DC 20036 • 800-521-4323 • www.AISHealth.com