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
MAJOR ARTICLE A Hospitalwide Intervention Program to Optimize the Quality of Antibiotic Use: Impact on Prescribing Practice, Antibiotic Consumption, Cost Savings, and Bacterial Resistance Carlos Bantar,1 Beatriz Sartori,2 Eduardo Vesco,3 Claudia Heft,2 Mariano Saúl,2 Francisco Salamone,4 and Marı́a Eugenia Oliva1 1 Committee for Prevention and Control of Nosocomial Infection and Departments of 2Pharmacy, 3Internal Medicine, and 4Microbiology, Hospital San Martı́n, Paraná, Entre Rı́os, Argentina Several findings from Argentina provide compelling evidence of the need for more rational use of antimicrobial agents. Thus, a multidisciplinary antimicrobial treatment committee for the development of a hospital-wide intervention program was formed to optimize the quality of antibiotic use in hospitals. Four successive steps were developed during 6-month periods: baseline data collection, introduction of a prescription form, education, and prescribing control. Sustained reduction of drug consumption was shown during the study (R 2 p 0.6885; P p .01). Total cost savings was US$913,236. To estimate the consumption of cefepime and aminopenicillin-sulbactam in relation to that of the third-generation cephalosporins, 2 indices were calculated: Icfp and Iams, respectively. Decreasing resistance to ceftriaxone by Proteus mirabilis and Enterobacter cloacae proved to be associated with increasing Icfp. Decreasing rates of methicillin-resistant Staphylococcus aureus were related to increasing Iams. The present study indicates that a systematic program performed by a multidisciplinary team is a cost-effective strategy for optimizing antibiotic prescribing. Appropriate antibiotic use is one of the main goals of the medical community [1]. Overuse of antimicrobial agents has been described worldwide in both community [2, 3] and hospital [4, 5] settings. In addition to the effect on patients [6, 7], antibiotic misuse can provoke the emergence of bacterial resistance [4, 8] and increase health care costs [9]. It is evident that optimizing antibiotic use is a challenge that deserves to be undertaken. Received 17 December 2002; accepted 13 March 2003; electronically published 9 July 2003. Reprints or correspondence: Dr. Carlos Bantar, Colón 128, (3100) Paraná, Entre Rı́os, Argentina ([email protected]). Clinical Infectious Diseases 2003; 37:180–6 2003 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2003/3702-0004$15.00 180 • CID 2003:37 (15 July) • Bantar et al. It has been observed that the infectious diseases physician plays a crucial role in controlling antibiotic use in the hospital [10], as does a multidisciplinary team approach, with the active involvement of clinical microbiologists and pharmacists [8, 9, 11–13]. Bantar et al. [14] published alarming rates of bacterial resistance in a surveillance study involving 27 Argentinean health care centers, and Bantar and others have noted high rates of nosocomial infection, surgical prophylaxis errors leading to unnecessary cost increases in the hospital [15], and confirmation of misuse of antibiotics in the same hospital [5]. These findings provide compelling evidence of the need for more-rational use of antimicrobial agents in Argentina. To our knowledge, a systematic strategy for control of antibiotic use in our country has not been published. Thus, we designed a multidisciplinary antimicrobial treatment team for de- veloping a hospital-wide intervention program to optimize antibiotic use in our hospital. Here, we report the impact of this program on prescribing practice, antibiotic use, cost savings, and bacterial resistance. METHODS Setting. The Hospital San Martı́n is a 250-bed public teaching hospital for adult patients in Paraná, a city in Argentina of ∼350,000 inhabitants. It has a 10-bed intensive care unit (ICU) and several surgical wards (including orthopedic, abdominal, thoracic, gynecological, and neurological units) and lacks facilities for solid-organ or bone marrow transplantation and cardiac surgery. Program design. An antimicrobial treatment committee (ATC), which comprised an infectious diseases physician (M.E.O.), a clinical microbiologist with experience in pharmacokinetics and pharmacodynamics (C.B.), a laboratory microbiologist (F.S.), 2 pharmacists (B.S. and C.H.), an internal medicine specialist (E.V.), and a computerized system analyst (M.S.), was convened in 1999 to design a sequential intervention program to optimize antibiotic use within the hospital. By early 1999 (before the study), specific computerized systems for data recording and analysis were implemented in the departments of pharmacy, microbiology, and prevention and control of nosocomial infection. Thereafter, 4 successive intervention steps were developed during 6-month periods, as follows: period 1 (the last 6 months of 1999; baseline), introduction of an optional antibiotic order form and collection of baseline data (i.e., data on bacterial resistance, antibiotic use, prescribing practice, and nosocomial infection and crude mortality rates); period 2 (the first 6 months of 2000; initial intervention period), feedback activities, according to the baseline findings, and transformation of the optional antibiotic order form into an obligatory requirement for procuring the drug from the pharmacy; period 3 (the last 6 months of 2000; education phase), like period 2, but with the addition of policy activity for every antibiotic prescription (except for surgical prophylaxis, which was already standardized) and bedside discussion among the infectious diseases physicians, the clinical microbiologist, and the attending physicians; and period 4 (the first 6 months of 2001; active control period), like period 3 but included modification of prescription, if necessary, by the ATC. There were no restrictions on antibiotic-prescribing habits. However, during the education phase, physicians were informed about the increased risk for the selection of bacterial resistance associated with the overuse of the third-generation cephalosporins and carbapenems, as well as the potential benefits of their replacement (when appropriate) with aminopenicillin-sulbactam or with the available fourth-generation cephalosporin, cefepime. To evaluate the impact of the antimicrobial-control program, no specific actions to control and prevent nosocomial infections (beyond the standard precautions that were being undertaken before starting the study) were introduced by the pertinent committee during the 24 months of the study. Strategy and data collection. Use of antibiotics was recorded as total number of grams of the drug on a homemade computerized system (SOUFARMA) [5] and then converted into defined daily doses (DDDs) per 1000 patient-days, in accordance with the World Health Organization recommendation [16]. Only the expenditure for drugs that are administered intravenously was analyzed. Data regarding antibiotic use associated with surgical prophylaxis were excluded. To estimate the rates of use of cefepime and aminopenicillin-sulbactam in relation to that for the third-generation cephalosporins, 2 indices were calculated: Icfp and Iams. Icfp was calculated as consumption of cefepime/consumption of ceftriaxone and ceftazidime ⫻ 100. Iams was calculated as consumption of aminopenicillin-sulbactam/consumption of ceftriaxone plus ceftazidime ⫻ 100. Consumption was measured in DDDs per 1000 patient-days. Bacterial resistance rates were recorded and analyzed using a homemade computerized system (SIR) described elsewhere [14] that was developed by one of the authors (C.B.). The antibiotic order form required the physician to state the certainty of infection diagnosis, as well as the most relevant epidemiological data. The infectious diseases physician filled out a unique form with the same data as the antimicrobial order form and additional data, such as any interruption or modification of treatment suggested by the ATC. Criteria used to define the need for antimicrobial therapy were adopted from the current edition of Principles and Practices of Infectious Diseases [17]. Therapy was considered appropriate if it was included in any of the primary or alternative regimens suggested in The Sanford Guide to Antimicrobial Therapy [18]. Data were finally recorded and analyzed using Epi Info software, version 6.0 (Centers for Disease Control and Prevention). Density of nosocomial infection was estimated as described elsewhere [15]. Data on the number of patient admissions and discharges, number of patient-days, and crude mortality were kindly supplied by S. Hayy from the Department of Statistics (Hospital San Martı́n). Bacterial resistance and antibiotic use. To evaluate the impact of changes in antibiotic use on bacterial resistance, we selected certain worrisome multidrug-resistant species detected in the baseline period for analysis, such as third-generation cephalosporin–resistant Escherichia coli, Enterobacter cloacae, Klebsiella pneumoniae, and Proteus mirabilis; imipenem-resistant Pseudomonas aeruginosa; and methicillin-resistant Staphylococcus aureus (MRSA). In addition, 2 phenotypes were selected to assess the impact of the increase of Iams and Icfp indices, aminopenicillin-sulbactam–resistant E. coli and cefeIntervention on Antibiotic Prescription • CID 2003:37 (15 July) • 181 pime-resistant P. aeruginosa. To date, vancomycin-resistant enterococci (VRE) has never been isolated in our hospital; thus, this organism was not included in the analysis. Variations in the resistance phenotype above, which were estimated as a 6month prevalence rate from each period, were compared with those for every antibiotic showing any sustained trend over the study period by linear regression analysis. Statistical analysis. Rates were analyzed by comparison of proportions with the x2 or Fisher’s exact tests using the Epi Info statistical package. Trends over time and correlation among variables were assessed by simple linear regression analysis, performed using Statistix software for Windows, version 2.0 (Analytical Software). P ⭐ .05 was regarded as statistically significant. All values are in US dollars. RESULTS Antibiotic use and cost savings. Variations in intravenous antibiotic consumption during the different periods are shown in table 1. The numbers of patient-days for periods 1, 2, 3, and 4 were 37,556, 38,086, 39,954, and 38,905, respectively. A sig- nificant global decrease in the use of antibiotics was observed between period 1 (baseline period) and period 2 (initial intervention period). Simple linear regression analysis showed that this trend continued during the remaining periods (adjusted R 2 p 0.6885; P p .01). Linear regression analysis revealed a sustained increase in the Icfp and Iams indices throughout the study (adjusted R 2 p 0.88 [P p .04] and 0.81 [P p .06], respectively). The evolution of both indices along the 4 periods is shown in figure 1. Cost savings from period 1 (baseline) to period 2, from period 2 to period 3, and from period 3 to period 4 were $261,955, $57,245, and $12,881, respectively. The cumulative total savings during the 18 months of intervention was $913,236. Prescribing practice. In all, 450 and 349 successive unique antimicrobial order forms (corresponding to 1 patient each) were analyzed during periods 1 and 4, respectively. Antibiotics to be given orally were also included in this analysis. Because the ATC members might modify a considerable number of antimicrobial orders during period 4, prescription rates during this period were divided between the prescribing intention and definitive treatment (after correction by the ATC members). Table 1. Variations in use of intravenous antibiotics after implementation of successive intervention steps for optimizing the quality of antimicrobial prescription in the hospital. Consumption, divided daily doses a per 1000 patient-days Antibiotic Period 1: baseline Period 2: initial intervention Period 3: education phase Period 4: active control phase P for period 4 vs. period 1b Amikacin 40.18 25.40c 26.22 32.04 NS AMP-SUL 24.75 17.05 18.82 22.90 NS Carbapenem 13.54 7.79 6.39 6.18 .03 3.86 3.31 8.96d 8.80 NS Ceftazidime 29.46 21.67 24.60 30.50 NS Ceftriaxone 62.85 35.63c 26.63 11.77 !.0001 Cefepime Cefuroxime 9.48 4.83 4.77 2.88 Cephalothin 80.68 62.59 59.54 45.43 !.0001 Ciprofloxacin 16.02 14.11 12.75 17.54 NS 7.87 3.49 3.10 3.02 NS Clindamycin 42.75 31.14 19.66 21.99 .003 Gentamicin 51.13 43.03 33.04 36.72 NS Metronidazole 19.78 15.20 17.85 15.87 NS Vancomycin 28.53 27.81 24.98 20.70 430.89 313.06c 287.31 276.35 Clarithromycin Total NOTE. a b c d AMP-SUL, aminopenicillin-sulbactam; NS, not significant. See the section “Program design” in Methods for detailed definitions. Determined by x2 test. Significant decrease was observed between period 1 and period 2 (P ! .05). Significant increase was observed between period 2 and period 3 (P ! .05). 182 • CID 2003:37 (15 July) • Bantar et al. .04 NS !.0001 Figure 1. Variation of 2 antibiotics indices during a 4-period systematic intervention program. Iams () and Icfp () represent the respective ratios between the use (divided daily doses per 1000 patient-days) of aminopenicillin-sulbactam or cefepime and third-generation cephalosporins (ceftriaxone and ceftazidime). See Methods for details. Data are shown in table 2. An alarmingly high rate of carbapenem prescription (44%) was noted at baseline. In addition, an appreciable use of ceftriaxone (17%) was observed. On the other hand, cefepime and aminopenicillin-sulbactam were seldom ordered. After the education phase, a dramatic decrease was observed in the prescribing intention of carbapenem and ceftriaxone, as shown in period 4. On the other hand, prescription rates for cefepime and aminopenicillin-sulbactam and the respective indices, Icfp and Iams, increased in this period. All of these trends were reinforced by the active intervention of the ATC members at the time of definitive treatment (table 2). The characteristics of the control intervention by ATC members in the last period were also examined. A significant increase was found in the rate of microbiologically based prescribing intention in comparison with that of the baseline period (62.8% vs. 27%, respectively; P ! .0001 ). In all, 87 (24.9%) of 349 prescriptions were modified, and 13 order forms (3.7%) were withdrawn by the ATC. Therapy was reduced, in terms of either dose or duration, in 11.5% of cases. Furthermore, 86.1% of the corrections involved less-expensive therapy, and 47% used a drug with a narrower antimicrobial spectrum. Among the 87 order forms that were modified by the ATC, 5, 43, and 39 corresponded to mild, moderate, and severe infections, respectively. The presumed or documented sources of the infection corresponding to these orders were lower respiratory tract (44.2%), abdomen (12.6%), catheter (12.6%), skin and soft tissue (9.2%), urinary tract (8.0%), CNS (6.9%), bone (3.4%), and other sources (3.1%). Nosocomial infection, crude mortality rates, and hospital and ICU stay. The nosocomial infection rates were 8.0, 8.5, 8.5, and 8.9 cases per 1000 patient-days for periods 1, 2, 3, and 4, respectively. The corresponding rates of crude mortality were 9.4, 7.3, 6.6, and 7.4 deaths per 1000 patient-days. The mean lengths of hospital stay (SD) for infected patients decreased significantly (P p .04 ) between period 1 and period 4, as follows: 47.4 40.4, 28.6 28.1, 25.9 24.6, and 25.5 23.3 days. The mean durations of total stay in the ICU (SD) were 261.7 52.8, 228.5 47.3, 233.2 43.1, and 218.8 40.5 days. Decreasing antibiotic use correlated with duration of both hospital and ICU stay (R 2 p 0.99 [P p .004] and R 2 p 0.86 [P p .047], respectively), whereas a trend was observed for crude mortality (R 2 p 0.82 [P p .06]). Correlation between bacterial resistance and antibiotic use. Data on bacterial resistance and antibiotic use are summarized in table 3. No correlation was found between variations in antibiotic use and resistance to third-generation cephalosporins by E. coli or K. pneumoniae, whereas decreasing rates of this kind of resistance for P. mirabilis and E. cloacae proved to be associated with the increase in the Icfp index over time. Likewise, decreasing rates of MRSA were related to an increasing Iams index, as well as to a sustained reduction in vancomycin use. No correlation was shown between bacterial resistance variations and the consumption of any one antibiotic included in the formulas for the above indices. The frequency of imipenem-resistant P. aeruginosa strains decreased to 0% in the last period. This finding was strongly associated with a reduction in carbapenem consumption over time. No changes Table 2. Variation in the frequency of prescription of different antibiotics during a systematic intervention program. Frequency of prescription, %b Period 1: baseline (n p 450) a Antibiotic Period 4: active control (n p 349) Prescribing intention Definitive treatment P for period 1 vs. definitive treatmentc Carbapenem 44 5d 2.3 !.0001 Ceftriaxone 17 9.4d 8 !.0001 Ceftazidime 8 8.4 6 NS Cefepime 0.5 AMP-SUL 6 d 4.2 35.7d 7.3 !.0001 38 !.0001 Iams 24 190d 271 !.0001 Icfp 2 22d 52 !.0001 NOTE. AMP-SUL, aminopenicillin-sulbactam; NS, not specific. a Icfp was calculated as frequency of prescription of cefepime/ceftriaxone and ceftazidime ⫻ 100. Iams was calculated as frequency of prescription of AMP-SUL/ceftriaxone plus ceftazidime ⫻ 100. b See the section “Program design” in Methods for detailed definitions of periods. c Determined by x2 test. d P ⭐ .05 vs. baseline. Intervention on Antibiotic Prescription • CID 2003:37 (15 July) • 183 Table 3. Variation of bacterial resistance rates after the implementation of successive intervention steps for optimizing the quality of antimicrobial prescription and association with antibiotic use over time. Resistance rate, % (total no. of strains) Organism/phenotypic resistance Escherichia coli/third-generation cephalosporins Period 1: baseline 8 (204) Period 2: initial intervention 7 (195) Period 3: education phase 12 (128) a Period 4: active control 8 (131) Associated antibiotic or index (type of variation, adjusted R 2)b P None — E. coli/AMP-SUL 22 (204) 18 (195) 19 (128) 18 (131) None — Klebsiellaspecies/third-generation cephalosporins 45 (64) 63 (32) 21 (43)c 42 (50) None — Proteus mirabilis/third-generation cephalosporins 40 (47) 45 (49) 30 (46) 27 (41) Icfp (increase, 0.85) .05 Enterobacter cloacae/third-generation cephalosporins 50 (16) 50 (6) 29 (17) 10 (10) Iams (increase, 0.92) Icfp (increase, 0.91) .03 .03 Pseudomonas aeruginosa/imipenem 19 (88) 8 (62) 3 (100) 0 (56) Carbapenem (decrease, 0.93) .02 8 (88) 0 (62) 3 (100) 7 (56) None 47 (165) 50 (150) 40 (199) 30 (101) P. aeruginosa/cefepime Staphylococcus aureus/methicillin NOTE. Iams (increase, 0.89) Vancomycin (decrease, 0.94) .04 .02 AMP-SUL, aminopenicillin-sulbactam. a See the section “Program design” in Methods for detailed definitions of periods. b Linear regression analysis between resistance rate evolution and antibiotic consumption throughout the study (table 1). Icfp was calculated as consumption of cefepime/consumption of ceftriaxone and ceftazidime ⫻ 100 . Iams was calculated as consumption of AMP-SUL/consumption of ceftriaxone plus ceftazidime ⫻ 100. Consumption was measured in defined daily doses per 1000 patient-days. c P ! .05 for period 3 vs. period 2. were observed in the resistance rates to aminopenicillin-sulbactam and cefepime by E. coli and P. aeruginosa, respectively. DISCUSSION Several strategies for regulating antimicrobial prescribing practices have been proposed, such as formulary replacement or restriction [19], introduction of order forms [20], health care provider education, feedback activities [21], and required approval from an infectious diseases physician for drug prescription [22]. Although most of these interventions have been assessed separately, data from prospective studies evaluating the impact of the different strategies applied systematically over time in the same hospital setting remain scarce. In addition, results of a coordinated approach by a multidisciplinary team composed of infectious diseases physicians, microbiologists, and pharmacists have rarely been reported [13]. We have shown a significant reduction in antibiotic use when feedback activities were performed and when we made the optional antibiotic order form obligatory for getting the drug from pharmacy (i.e., period 2, the initial intervention period). Our findings are opposite to those of Gyssens et al. [20], who reported a 25% increase in antibiotic use after this kind of intervention at a 948-bed university hospital in The Netherlands. This is not surprising, because we noted a number of unjustified prescriptions during the baseline phase. A further decrease in drug consumption was reached during the last 2 intervention periods 184 • CID 2003:37 (15 July) • Bantar et al. (education and active control phases). This fact might be attributable to the combination of both kinds of interventions. Because of the alarming prevalence of bacterial resistance found in the baseline period, our education strategy emphasized 3 major issues: making an effort to document the infection microbiologically before starting antimicrobial therapy, avoiding use of certain antibiotics known to be associated with emergence of bacterial resistance [23, 24], and increasing the use of antimicrobials thought to reduce the frequency of multidrugresistant organisms [25]. Our results challenge the connection between heavy or inappropriate antibiotic use and bacterial resistance. One of the antibiotics we wanted to use instead of third-generation cephalosporins (when appropriate) was aminopenicillin-sulbactam, a drug active against anaerobic organisms. Although several authors have suggested that use of antibiotics with antianaerobic activity is a strong predictor for VRE acquisition, most studies referred to clindamycin [26] and metronidazole [27], rather than aminopenicillin-sulbactam, or found that use of third-generation cephalosporins and vancomycin was an additional VRE predictor [27, 28]. During our study, use of clindamycin, ceftriaxone, and vancomycin decreased, whereas use of metronidazole remained stable. However, because we did not detect VRE during any part of the study period, we cannot assert that this situation had an impact on selection of VRE. We are unable to assert that the sole substitution of thirdgeneration cephalosporins by aminopenicillin-sulbactam or ce- fepime results in cost containment. Nevertheless, an indirect contribution to the cost savings by the association between this switch and the sustained reduction of resistance rates over time can be inferred. For instance, third-generation cephalosporin resistance in P. mirabilis and E. cloacae was inversely associated with the increase in the Icfp (also Iams for E. cloacae), whereas no correlation was observed with the single components of the index formula. These findings differ from those of Friedrich et al. [29], who found a correlation between use of single antibiotics and changes in the susceptibility patterns of some gramnegative aerobes by using data analysis comparable to that of our study (i.e., simple linear regression). To our knowledge, the present article is the first report to evaluate a ratio representing the relative consumption between ⭓2 antibiotics against variations in resistance rates over time. Thus, our results suggest that, for certain antibiotics, demonstration of the relationship between drug use and a specific phenotype of resistance may require the assessment of a ratio relating to the opposite trends in consumption of different antimicrobial agents. In fact, neither the sole decreasing use of ceftriaxone nor the sole increasing consumption of cefepime or aminopenicillin-sulbactam proved to be related to any variation in resistance rates over time. Of interest, rates of MRSA were inversely associated with increasing Iams, but not with the single drugs included in this index formula. This finding is in agreement with that of Landman et al. [22], who described a reduction in the number of patients infected with MRSA after a dramatic increase in use of ampicillin-sulbactam in parallel with a reduction in thirdgeneration cephalosporin use. We hypothesize that the known in vivo and in vitro inhibitory effect of aminopenicillin-sulbactam over that of the cephalosporins against several MRSA strains might be a suitable explanation for this phenomenon [30, 31]. In summary, the results of the present study support the notion that a systematic program executed by a multidisciplinary team is a cost-effective strategy for optimizing antibiotic use in a hospital and has an evident impact on prescribing practice, antibiotic use, cost savings, and bacterial resistance. Furthermore, this impact may be noted even where there are high rates of bacterial resistance and irrespective of any additional precautions for controlling nosocomial infection. Acknowledgment We deeply thank Mary Forrest for her valuable technical assistance in revising the manuscript. References 1. Polk R. Optimal use of modern antibiotics: emerging trends. Clin Infect Dis 1999; 29:264–74. 2. Guillemot D, Maison P, Carbon C, et al. Trends in antimicrobial use in the community—France, 1981–1992. J Infect Dis 1998; 177:492–7. 3. Gonzalez R, Steiner JF, Sande MA. Antibiotic prescribing for adults with colds, upper respiratory tract infections and bronchitis by ambulatory care physicians. JAMA 1997; 278:901–4. 4. Fridkin SK, Steward CD, Edwards JR, et al. Surveillance of antimicrobial use and antimicrobial resistance in United States hospitals: Project ICARE phase 2. Project Intensive Care Antimicrobial Resistance Epidemiology (ICARE) hospitals. Clin Infect Dis 1999; 29:245–52. 5. Bantar C, Sartori B, Saúl M, Salamone F, Vesco M, Morera G. Alarming misuse of antibiotics in a hospital from Argentina [abstract A-69]. In: Program and abstracts of the 101st General Meeting of The American Society for Microbiology (Orlando, Florida). Washington, DC: American Society for Microbiology, 2001:17. 6. Beilby J, Marley J, Walker D, Chamberlain N, Burke M. Effect of changes in antibiotic prescribing on patient outcomes in a community setting: a natural experiment in Australia. FIESTA Study Group. Clin Infect Dis 2002; 34:55–64. 7. Gross R, Morgan AS, Kinky DE, Weiner M, Gibson GA, Fishman NO. Impact of a hospital-based antimicrobial management program on clinical and economic outcomes. Clin Infect Dis 2001; 33:289–95. 8. Shlaes DM, Gerding DN, John JF Jr, et al. Programmatic role of the infectious diseases physician in controlling antimicrobial costs in the hospital: guidelines for the prevention of antimicrobial resistance in hospitals. Society for Healthcare Epidemiology of America and Infectious Diseases Society of America Joint Committee on the Prevention of Antimicrobial Resistance. Clin Infect Dis 1997; 25:584–99. 9. Barenfanger J, Short MA, Groesch AA. Improved antimicrobial interventions have benefits. J Clin Microbiol 2001; 39:2823–8. 10. John JF Jr, Fishman NO. Programmatic role of the infectious diseases physician in controlling antimicrobial costs in the hospital. Clin Infect Dis 1997; 24:471–85. 11. Knox KL, Holmes AH. Regulation of antimicrobial prescribing practices—a strategy for controlling nosocomial antimicrobial resistance. Int J Infect Dis 2002; 6(Suppl 1):S8–13. 12. Ibrahim KH, Gunderson B, Rotschafer JC. Intensive care unit antimicrobial resistance and the role of the pharmacist. Crit Care Med 2001; 29(Suppl 4):N108–13. 13. Gums JG, Yancey RW Jr, Hamilton CA, Kubilis PS. A randomized, prospective study measuring outcomes after antibiotic therapy intervention by a multidisciplinary consult team. Pharmacotherapy 1999; 19:1369–77. 14. Bantar C, Famiglietti A, Goldberg M. Three-year surveillance study of nosocomial bacterial resistance in Argentina. The Antimicrobial Committee and the National Surveillance Program (SIR) Participants Group. Int J Infect Dis 2000; 4:85–90. 15. Bustos J, Vesco E, Tosello C, et al. Alarming baseline rates of nosocomial infection and surgical prophylaxis errors in a small teaching hospital from Argentina. Infect Control Hosp Epidemiol 2001; 22:264–5. 16. Maxwell M, Heaney D, Howie JG, Noble S. General practice fund holding: observations on prescribing patterns and costs using the defined daily dose method. BMJ 1993; 307:1190–4. 17. Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas and Benett’s principles and practice of infectious diseases. 4th ed. New York: Churchill Livingstone, 1995. 18. Gilbert DN, Moellering RC, Sande MA. The Sanford guide to antimicrobial therapy. 32nd ed. Vermont: Antimicrobial Therapy, 2002. 19. Lee J, Carlson JA, Chamberlain MA. A team approach to hospital formulary replacement. Diagn Microbiol Infect Dis 1995; 22:239–42. 20. Gyssens IC, Blok WL, van den Broek PJ, Hekster YA, van der Meer JW. Implementation of an educational program and an antibiotic order form to optimize quality of antimicrobial drug use in a department of internal medicine. Eur J Clin Microbiol Infect Dis 1997; 16:904–12. 21. De Santis G, Harvey KJ, Howard D, Mashford ML, Moulds RF. Improving the quality of antibiotic prescription patterns in general practice: the role of educational intervention. Med J Aust 1994; 160:502–5. 22. Landman D, Chockalingam M, Quale JM. Reduction in the incidence Intervention on Antibiotic Prescription • CID 2003:37 (15 July) • 185 23. 24. 25. 26. of methicillin-resistant Staphylococcus aureus and ceftazidime-resistant Klebsiella pneumoniae following changes in a hospital antibiotic formulary. Clin Infect Dis 1999; 28:1062–6. Carmeli Y, Troillet N, Eliopoulos GM, Samore MH. Emergence of antibiotic-resistant Pseudomonas aeruginosa: comparison of risks associated with different antipseudomonal agents. Antimicrob Agents Chemother 1999; 43:1379–82. Patterson JE, Hardin TC, Kelly CA, Garcia RC, Jorgensen JH. Association of antibiotic utilization measures and control of multiple-drug resistance in Klebsiella pneumoniae. Infect Control Hosp Epidemiol 2000; 21:455–8. Bisson G, Fishman NO, Patel JB, Edelstein PH, Lautenbach E. Extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella species: risk factors for colonization and impact of antimicrobial formulary interventions on colonization prevalence. Infect Control Hosp Epidemiol 2002; 23:254–60. Lautenbach E, LaRosa LA, Marr AM, Nachamkin I, Bilker WB, Fishman NO. Changes in the prevalence of vancomycin-resistant enterococci in response to antimicrobial formulary interventions: impact of progressive restrictions on use of vancomycin and third-generation cephalosporins. Clin Infect Dis 2003; 36:440–6. 186 • CID 2003:37 (15 July) • Bantar et al. 27. Carmeli Y, Eliopoulos GM, Samore MH. Antecedent treatment with different antibiotic agents as a risk factor for vancomycin-resistant Enterococcus. Emerg Infect Dis 2002; 8:802–7. 28. Fridkin SK, Edwards JR, Courval JM, et al. The effect of vancomycin and third-generation cephalosporins on prevalence of vancomycinresistant enterococci in 126 US adult intensive care units. Intensive Care Antimicrobial Resistance Epidemiology (ICARE) Project and the National Nosocomial Infections Surveillance (NNIS) System Hospitals. Ann Intern Med 2001; 135:175–83. 29. Friedrich LV, White RL, Bosso JA. Impact of use of multiple antimicrobials on changes in susceptibility of gram-negative aerobes. Clin Infect Dis 1999; 28:1017–24. 30. Kobayashi S, Arai S, Hayashi S, Sakaguchi T. In vitro effects of blactams combined with b-lactamase inhibitors against methicillinresistant Staphylococcus aureus. Antimicrob Agents Chemother 1989; 33:331–5. 31. Chambers HF, Kartalija M, Sande M. Ampicillin, sulbactam, and rifampin combination treatment of experimental methicillin-resistant Staphylococcus aureus endocarditis in rabbits. J Infect Dis 1995; 171: 897–902.