Improving the appropriateness of vancomycin use by sequential interventions

Improving the appropriateness of vancomycin use by sequential interventions Benjamin A. Lipsky, MDa,b Catherine A. Baker, PharmDc Linda L. McDonald, RN, MSPHa Norman T. Suzuki, PharmDa Seattle, Washington

Background: Vancomycin usage is directly associated with the incidence of vancomycin-resistant enterococci. Optimal methods to reduce inappropriate use have not been delineated. We determined the appropriateness of vancomycin prescribing at our hospital on the basis of national guidelines and assessed the effect of sequential administrative and educational interventions. Methods: In this prospective 3-phase study conducted in a Veterans Affairs Medical Center, we monitored vancomycin prescribing at baseline and in 2 follow-up periods. Administrative interventions included discussions with service chiefs and revising routine perioperative antibiotic prophylaxis orders. Educational interventions included in-services about vancomycin-resistant enterococci and appropriate vancomycin prescribing. In each monitoring period, 50 consecutive new vancomycin orders that could be evaluated were classified for appropriateness and categorized by indication. Results: At baseline, 70% of vancomycin use was inappropriate. Surgical services accounted for 84% of orders. Interventions targeted services with high or frequently inappropriate vancomycin use. After administrative interventions, inappropriate vancomycin use dropped to 40% of orders (P = .003). Improvements were noted in targeted services. Educational interventions further decreased inappropriate vancomycin use, but the effect appeared transient. Conclusions: The simple, nonrestrictive administrative interventions used resulted in a statistically significant (30%) reduction in inappropriate vancomycin prescribing. However, educational interventions provided only transient benefit on institutional prescribing patterns. (AJIC Am J Infect Control 1999;27:84-90)

The recent increasing incidence of infections caused by vancomycin-resistant enterococci (VRE) has focused national attention on containing this worrisome pathogen. Spread of VRE isolates to virtually all 50 of the United States is considered by some authorities to be “perhaps the greatest threat to public health.”1 The limited treatment options for VRE and the possibility of vancomycin resistance genes spreading to other bacteria highlight the importance of halting the spread of this nosocomial pathogen. Among the identified factors associated with VRE isolation2-7 is a relatively high rate of vancomycin use.4,6,8-12 One large case-control study found that previously receiving vancomycin increased the odds of being colonized or infected with an enterococcal isolate that was vancomycin resistant by 2.35 times.13

From the Veterans Affairs Puget Sound Health Care System,a and the Departments of Medicineb and Pharmacy Services,c University of Washington. Presented in part at the Infectious Diseases Society of America 34th Annual Meeting (abstract 40), September 19, 1996, New Orleans, La. Reprint requests: Benjamin A. Lipsky, MD, VA Puget Sound HCS, GIMC/Antibiotic Research Clinic (111M), 1660 S Columbian Way, Seattle, WA 98108-1532. 17/46/94338

84

Two major approaches have been advocated to prevent the appearance or curtail the spread of vancomycin-resistant organisms: environmental containment (eg, use of single rooms or cohorting for patients with VRE) and use of contact isolation techniques, and restricting vancomycin usage.7,10,14-16 To encourage these practices, the Hospital Infection Control Practices Advisory Committee (HICPAC) of the Centers for Disease Control and Prevention issued guidelines promoting prudent vancomycin use by delineating criteria for “appropriate” and “inappropriate” therapy (Table 1).2 Previous studies have documented frequent inappropriate vancomycin prescribing.5,15,16-26 Some authors have advocated studying the effect of controlling antibiotic use on nosocomial VRE acquisition.1,5,7-9,12,14,27 However, altering prescribing habits is difficult, and effective methods have yet to be delineated.28 At the Veterans Affairs Puget Sound Health Care System, Seattle Division, a 446-bed university-affiliated tertiary care facility, we have isolated VRE from more than 70 patients during a 3-year period. Vancomycin use in our institution increased approximately fivefold in the 7 years preceding this study. This increase has been caused, in part, by approximately 30% of Staphylococcus aureus isolates and 50% of coagulase-negative staphylococci at our institution being methicillin-resistant. Therefore, we prospectively studied the appropriateness

AJIC

Lipsky et al 85

Volume 27, Number 2

Table 1. Summary of criteria for vancomycin use* Appropriate

Inappropriate

Treatment of serious infections caused by β-lactam-resistant gram-positive microorganisms

Routine perioperative prophylaxis other than in a patient with a life-threatening allergy to β-lactam antibiotics

Treatment of gram-positive infections in patients with a serious allergy to β-lactam antimicrobials

Empiric therapy for a febrile neutropenic patient unless strong evidence is present for a gram-positive infection and the prevalence of infections caused by MRSA in the institution is substantial

Antibiotic-associated colitis that fails to respond to metronidazole therapy or is severe and potentially life-threatening Prophylaxis for endocarditis as recommended by the American Heart Association for certain procedures in high-risk persons Prophylaxis for postoperative infections before major surgical procedures involving implantation of prosthetic/foreign materials at institutions with high rates of infections caused by MRSA or methicillin-resistant Staphylococcus epidermis (Note: A single dose immediately before surgery is sufficient unless the procedure lasts >6 hours, in which case a single additional dose should be administered; prophylaxis should not exceed 2 doses)

Treatment for a single blood culture positive for a coagulasenegative Staphylococcus if other blood cultures drawn in the same time frame are negative Continued empiric use for presumed infections in patients whose cultures are negative for β-lactam-resistant grampositive microorganisms Systemic or local prophylaxis for infection or colonization of indwelling central or peripheral intravascular catheters Selective decontamination of the digestive tract Eradication of MRSA colonization Primary treatment of antibiotic-associated colitis Routine prophylaxis for patients on continuous ambulatory peritoneal dialysis or hemodialysis Treatment (chosen for dosing convenience) of infections caused by β-lactam-sensitive gram-positive microorganisms in patients with renal failure Topical application or irrigation of a wound Routine prophylaxis for very low-birth-weight infants

*Modified from Reference 2.

of vancomycin prescribing at our medical center and assessed the effect of sequential interventions designed to increase the appropriateness of vancomycin usage. METHODS Data collection We monitored vancomycin prescribing habits at 3 intervals: at baseline, after administrative interventions, and, finally, after educational interventions (Fig 1). In our medical center, vancomycin is available on formulary with no restrictions. Data were collected to determine whether the indication for vancomycin prescribing was appropriate (per the HICPAC guidelines), to identify specific services or areas within the institution with high rates of inappropriate vancomycin use, and to then assess the impact of the 2 types of interventions. Data were collected by reviewing medical

records and microbiology laboratory reports, interviewing patients, and consulting with prescribing clinicians, if necessary. In each monitoring period, consecutive new orders for vancomycin were reviewed until 50 orders that could be evaluated were obtained. Intravenous vancomycin orders were tracked through the central pharmacy, and oral vancomycin orders were obtained from computerized records. Orders for vancomycin in our medical center are mostly written by house staff, who rotate among the University of Washington–affiliated hospitals, usually on a monthly schedule, with some orders written by nurse practitioners or physicians assistants. Data collected included the patients’ location (ward), the ordering service and program, and the reason for prescribing vancomycin. Each vancomycin order was classified in 2 ways: first, as being for prophylaxis, empiric therapy, or definitive therapy, and second, as appropri-

AJIC

86 Lipsky et al

April 1999

Fig 1. Timeline for study procedures, including monitoring and interventions.

ate, inappropriate, or not applicable, according to the HICPAC criteria. We judged the appropriateness of the indication for vancomycin usage by the circumstances prevailing at the time of the order and the duration of therapy only for major surgical procedures involving prosthetic devices, as specified in the HICPAC guidelines. Administrative interventions Baseline monitoring assessed vancomycin prescribing before any formal interventions. Targeted administrative interventions were then undertaken in an effort to promote appropriate vancomycin use. Administrative interventions consisted of a combination of discussions with chiefs of clinical programs and committees (infection control, pharmacy, pharmacy and therapeutics committee, infectious disease, quality assurance, and services with the highest inappropriate use of vancomycin) and revising the routine preprinted orders for perioperative antibiotic prophylaxis. Discussions involved 15-minute to 30-minute sessions led by the chair of the infection control committee (B. A. L.) or a clinical pharmacist (C. A. B.) with use of handouts that detailed baseline vancomycin usage data for each service and identified areas of frequent inappropriate use for each service. Literature was supplied supporting the HICPAC guidelines for those criteria with high levels of inappropriate use or in cases in which existing service protocols or prescribing patterns differed from HICPAC criteria. In addition, routine cardiothoracic operative antibiotic prophylaxis orders were revised to remove vancomycin for coronary artery bypass grafts and reserve it for implantation of prostheses or for patients with a serious allergy to β-lactam antibiotics. Approximately 1 month after these administrative interventions, monitoring of vancomycin usage by the same methods used at baseline was repeated for 50 orders. To maintain a similar sample, routine preprint-

ed orders for perioperative prophylaxis, which were revised during this phase, were included as orders that could be evaluated from this monitoring period forward. Educational interventions The second intervention consisted of educational efforts aimed at prescribing clinicians and other health care workers. These efforts included 20-minute to 30minute in-services throughout the institution, led by a representative from the infection control program (B. A. L. or L. L. M.) or a clinical pharmacist (C. A. B.), to various health care professionals, including surgeons, nurses, pharmacists, physicians, and representatives from radiology and respiratory therapy. Audiovisual presentations and handouts focused on VRE: the implications, epidemiology (both nationwide and within the institution), associated risk factors, and available methods to prevent emergence and reduce transmission. We reviewed techniques of body substance isolation29 and appropriate vancomycin prescribing. We summarized our institutional data on vancomycin prescribing, accentuating identified target areas and potential strategies for improvement. Providers were encouraged to “think VRE” when ordering vancomycin: “V” = verify that the indication is appropriate, “R” = reassess daily if vancomycin is still required, and “E” = ensure appropriate use by avoiding topical application and irrigation with vancomycin solutions and obtaining adequate serum vancomycin concentrations when indicated. A final monitoring period initiated during the later part of educational interventions assessed vancomycin usage by the methods previously described. Statistical analysis Statistical analysis of the differences between proportions of orders were performed by a χ2 test for nonparametric, independent sample data, using 2-tailed P val-

AJIC

Lipsky et al 87

Volume 27, Number 2

ues. Use of the χ2 test for the categoric data was supported in that each cell’s expected frequency was ≥5. For comparisons of inappropriate prescribing rates, orders classified as “not applicable” were grouped with appropriate orders, and a χ2 test with 1 degree of freedom was used. Significance was defined as a P value < .05. RESULTS Sample This study was conducted during an 11-month period ending August 1996. To obtain 50 vancomycin orders that could be evaluated in each of the 3 monitoring phases, we examined 66 to 70 orders in each phase. The main reasons for excluding orders throughout the study period were incomplete data (11 orders), unavailable medical records (21 orders), outpatient or renal dialysis unit administration (8 orders), and unclear history of penicillin reaction (12 orders). No more than 3 orders were classified as “not applicable” during any 1 monitoring period. Prescribing characteristics The distribution of vancomycin orders among the various medical center programs was similar for each of the 3 monitoring periods. At baseline, 66% of orders were for surgical prophylaxis, 22% for empiric therapy, and 12% for definitive therapy of infection after culture results were available. The various surgical services were responsible for 84% of the vancomycin orders. Overall, 70% of new vancomycin orders were for inappropriate indications. The percentage of all vancomycin orders written and the rate of inappropriate usage for the most frequently prescribing services were: cardiothoracic surgery (48% of orders; 92% inappropriate), general surgery (16% of orders; 62% inappropriate), and medicine services (14% of orders; 86% inappropriate). Thus target interventions were focused on these programs. Impact of administrative interventions After the administrative interventions, overall inappropriate prescribing of vancomycin was statistically significantly reduced, from 70% to 40% of orders (P = .003) (Fig 2). Both the total number of orders and the proportion of inappropriate orders prescribed by target programs were reduced (Fig 3). The proportion of orders prescribed by cardiothoracic surgery decreased from 48% to 36%; only 39% were inappropriate, compared with 92% at baseline (P < .001). This decrease was largely accounted for by improved perioperative prophylaxis orders. General surgery accounted for 12% (16% at baseline) of vancomycin orders, 50% (62% at baseline) of which were inappropriate. Medicine accounted for 6% (14% at baseline) of orders, 67% (86%

Fig 2. Overall percentage of inappropriate vancomycin prescribing per HICPAC guidelines.

at baseline) of which were inappropriate. Overall, the percentage of orders prescribed for routine surgical prophylaxis in patients without a life-threatening allergy to a β-lactam antibiotic was reduced from 34% to 12% of orders (P = .009), and the proportion for empiric therapy was reduced from 22% to 12% (P = .357). Impact of educational interventions Monitoring after the second phase of interventions (educational in-services) demonstrated that the overall institutional inappropriate prescribing rate had increased from 40% to 50% of orders (Fig 2). However, this monitoring period must be evaluated as 2 separate phases. The first half (25 orders) of the educational monitoring period examined orders written by house staff exposed to the educational in-services. Midway through the monitoring period, resident physicians rotated off service, and new physicians who had not been exposed to the educational interventions arrived. Among the prescribers exposed to the special education, 32% of vancomycin orders were for inappropriate indications compared with 68% for the orders of the prescribers not exposed (P = .011). The later rate is very similar to the baseline rate of 70%. Vancomycin usage and VRE isolations Total (oral and intravenous) vancomycin usage decreased during the study period from approximately 6200 g per year to 5000 g per year. The number of VRE isolates varied greatly, but averaged 3.5 per month during the baseline monitoring and 4.3 per month during the interventions. DISCUSSION Increased prescribing of vancomycin in the past decade has been linked to the isolation and spread of

AJIC

88 Lipsky et al

Fig 3. Percentage of inappropriate orders within targeted services for each monitoring period. CT Surg, cardiothoracic surgery; Gen Surg, general surgery.

VRE,4,12,16,30 and our institution has demonstrated this pattern. This finding has led to efforts to reduce total vancomycin use, first and foremost by decreasing its inappropriate usage. Although previous studies have shown that a substantial proportion of vancomycin therapy is for inappropriate indications,15,19,20,24,26,28,31,32 optimal methods to alter prescribing patterns have not been defined. The effectiveness of educational interventions for this problem has been a topic of considerable debate.14,32 Restrictions on prescribing have been shown to reduce vancomycin use and potentially alter the incidence of VRE isolates in some, but not all, reports.5,6,14,15,26,28 Efforts by Quale et al14 to encourage staff at a Veterans Affairs Medical Center to voluntarily reduce the use of vancomycin failed. They found that traditional infection control measures did not interrupt transmission of VRE,3,14 but imposing restrictions on vancomycin (and cefotaxime) prescribing helped curb their VRE outbreak.14 Similarly, Belliveau et al28 reported that after nonbinding guidelines failed to affect vancomycin prescribing, a restriction program reduced its use. To the contrary, Morris et al5 and Morgan et al26 did not observe a significant reduction in the prevalence of VRE isolates after instituting vancomycin restrictions. This finding is probably caused at least in part by other factors also associated with the emergence and spread of VRE.13,27 Furthermore, many physicians dislike prescribing restrictions. Our surgeons specifically requested that educational efforts be tried before restrictive methods were invoked. Less imposing interventions, such as education and administrative efforts, may be more readily accepted. At baseline, 70% of the vancomycin orders evaluated in our institution were for inappropriate indications, which is slightly higher than the rates reported from previous studies, ranging from 24% to 65% inappropri-

April 1999

ate use.15,16,19-21,23-26,31,32 In most of these studies, the same types of inappropriate uses of vancomycin that we saw were reported (eg, unneeded use for surgical prophylaxis, continued empiric use unsupported by culture results, unnecessarily prolonged surgical prophylaxis, and topical use of vancomycin irrigations). In another acute-care Veterans Affairs-affiliated hospital, Evans et al19 evaluated 101 vancomycin orders during a 1-month survey and found that 65% of the orders were inappropriate. Although their interpretation of the HICPAC guidelines differed somewhat from ours and a majority of their orders were for empiric therapy, this figure is remarkably similar to ours at baseline. Similarly, Watanakunakorn’s24 prospective survey of 97 vancomycin orders in a teaching institution that used educational efforts to promote appropriate vancomycin use found that 79% of orders were inappropriate. As in our study, surgical prophylaxis was identified as the major source of inappropriate use. Misan et al20 performed a prospective 2-phase drug usage review based on criteria developed at a large metropolitan teaching hospital. Their assessment of educational interventions implemented between 2 monitoring periods demonstrated no reduction in inappropriate prescribing (64% at baseline vs 65% after the intervention). Although educational interventions in this study included developing specific guidelines, presentation at hospital rounds, and publication in the hospital bulletin, they found that directly consulting with prescribing clinicians—as we did—was the most effective strategy. Another review of the appropriateness of vancomycin therapy in a large teaching institution after educational efforts and implementation of a vancomycin restriction policy was reported by Morgan et al.26 The policy required approval of an infectious disease specialist if vancomycin use continued past 72 hours. Despite these efforts, 35% of vancomycin courses ordered during a 1month period were inappropriate, and the incidence of VRE remained unchanged for 2 years after implementing the policy. The authors suggested that to decrease vancomycin use, the initiation of vancomycin therapy should be evaluated, and educational programs should be targeted toward high-use services, 2 tactics that were used in the current study. Recently, Anglim et al15 evaluated the impact of an approach combining restrictions on vancomycin use through a computerized physician order-entry system and instituting infection control measures. The overall use of vancomycin was reduced, and the percentage of inappropriate orders decreased from 61% to 30% orders (P < .001). As in our study, the main benefit derived from a reduction in vancomycin orders for routine surgical prophylaxis, from 31% to 3.7% of orders (P < .001). Of note was that the effect on ordering prac-

AJIC

Lipsky et al 89

Volume 27, Number 2

tices was immediate and continued for at least 1 year. Furthermore, perirectal surveillance cultures showed a significant reduction in the incidence of VRE acquisition throughout the implementation period. The effect of a pharmacy-driven vancomycin restriction program on the compliance with locally developed guidelines was reported by Belliveau et al.28 Immediately after restrictions were imposed, 82% of vancomycin courses during an 8-week period were deemed in compliance; during a second 8-week monitoring period, the rate had fallen to 78%. Disappointingly, nearly half of the inappropriate vancomycin orders had been evaluated in advance by a pharmacist. Some institutions have modified the HICPAC guidelines, suggesting that situations such as empiric use for presumed pneumococcal meningitis or presumed pneumococcal infections in critically ill patients are reasonable indications for vancomycin therapy.23,24 Furthermore, it has been suggested that the HICPAC guidelines should include a re-evaluation after 72 hours to assess appropriateness of continuing therapy when additional clinical data is available. Salemi et al25 retrospectively reviewed 758 courses of vancomycin therapy in a health maintenance organization teaching hospital. After 3 days of therapy, adequate clinical or laboratory data was available in 70% of cases to discontinue vancomycin or reclassify use from suspected to documented infections or indications. Of note is that evaluation solely after 3 days would not have revealed 44% of the total days of inappropriate vancomycin use. Previous studies have proposed and tested several approaches for monitoring and improving vancomycin prescribing, with varying degrees of efficacy. Our study employed a prospective study design based on national guidelines to sequentially evaluate the impact of 2 major types of nonrestrictive interventions on institutional prescribing patterns. Our efforts were targeted at the specialty services we found at baseline to have the highest and most frequently inappropriate use of vancomycin. The administrative interventions resulted in 30% reduction in inappropriate vancomycin prescribing, largely attributable to changing the routine perioperative antibiotic prophylaxis orders. The educational interventions in our second phase further reduced inappropriate use in institutional vancomycin prescribing patterns, but only during the time in which the house staff who received the interventions were prescribing. Previous reports have shown an initial, but short-lived, impact of education on vancomycin prescribing.23,26 Given that vancomycin use had steadily increased in our institution during the 7 years preceding this study, the nearly 20% decrease in total vancomycin use during the year of this study is noteworthy. We saw no decrease in the isolation rate of VRE, but that rate is highly variable in our institution.

Our study had several limitations. A majority of vancomycin use in our medical center is for surgical prophylaxis; definitive therapy was the indication for only 10% to 12% of our orders compared with 26% to 35% reported in previous studies.17,19,21 We focused only on the appropriateness of the indication for vancomycin therapy at the time it was prescribed. Initially, appropriate therapy may become inappropriate if it is continued. Based on our findings, promoting appropriate duration of therapy could reduce unnecessary vancomycin use in our institution by about 25%. Additional limitations of our study include the relatively small sample sizes and monitoring periods that overlapped the tenure of physicians who were variably exposed to the educational interventions. Nevertheless, we demonstrated a statistically significant reduction in inappropriate vancomycin prescribing with our interventions. Although we did not have the benefit of a control group and vancomycin usage patterns may fluctuate for reasons other than our interventions, the trends seen at our institution suggest that simple administrative interventions can substantially reduce inappropriate vancomycin prescribing. A mathematical model has suggested that reducing vancomycin use may decrease the number of VRE infections between 3% and 34%; this suggestion remains to be correlated with clinical data.27 Future studies of vancomycin prescribing should address the portion of its administration that is appropriate, the long-term followup on the effectiveness of control measures, and the impact of reduced use on the incidence of VRE. The most successful efforts to combat VRE probably will entail an institutionwide combination of both infection control measures and reduced antimicrobial use. Edward Boyko, MD, helped with statistical analysis, and Randy Grekowicz, RPh, assisted in data retrieval from a computerized pharmacy database.

References 1. Jarvis WR. Preventing the emergence of multidrug-resistant microorganisms through antimicrobial use controls: the complexity of the problem. Infect Control Hosp Epidemiol 1996;17:490-5. 2. Hospital Infection Control Practices Advisory Committee (HICPAC), Centers for Disease Control and Prevention. Preventing the spread of vancomycin resistance: a report from the Hospital Infection Control Practices Advisory Committee prepared by the subcommittee on prevention and control of antimicrobial-resistant microorganisms in hospitals. Infect Control Hosp Epidemiol 1995;16:105-13. 3. Quale J, Landman D, Atwood E, Kreiswirth B, Willey BM, Ditore V, et al. Experience with a hospital-wide outbreak of vancomycinresistant enterococci. AJIC Am J Infect Control 1996;24:372-9. 4. Handwerger S, Raucher B, Altarac D, Monka J, Marchione S, Singh KV, et al. Nosocomial outbreak due to Enterococcus faecium highly resistant to vancomycin, penicillin, and gentamicin. Clin Infect Dis 1993;16:750-5. 5. Morris JG, Shay DK, Hebden JN, McCarter RJ, Perdue BE, Jarvis W, et al. Enterococci resistant to multiple antimicrobial agents,

AJIC

90 Lipsky et al

including vancomycin. Ann Intern Med 1995;123:250-9. 6. Rubin LG, Tucci V, Cercenado E, Eliopoulos G, Isenberg HD. Vancomycin-resistant Enterococcus faecium in hospitalized children. Infect Control Hosp Epidemiol 1992;13:700-5. 7. Slaughter S, Hayden MK, Nathan C, Hu TC, Rice T, Van Voorhis J, et al. A comparison of the effect of universal use of gloves and gowns with that of glove use on acquisition of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med 1996;125:448-56. 8. Gordts B, Van Landuyt H, Ieven M, Vandamme P, Goossens H. Vancomycin-resistant enterococci colonizing the intestinal tracts of hospitalized patients. J Clin Microbiol 1995;33:2842-6. 9. Boyce JM. Vancomycin-resistant enterococci: pervasive and persistent pathogens. Infect Control Hosp Epidemiol 1995;16:676-9. 10. Hospital Infection Control Practices Advisory Committee (HICPAC), Centers for Disease Control and Prevention. Interim guidelines for prevention and control of staphylococcal infection associated with reduced susceptibility to vancomycin. MMWR Morb Mortal Wkly Rep 1997;46:626-8,635. 11. Santibanez S, Riley DK, Slifkin M, Chan-Tompkins NH, Babinchak TJ, Rotheram EB. Emergence of vancomycin-resistant enterococci at a large teaching hospital. In: Programs and abstracts of the 34th Annual Meeting of the Infectious Diseases Society of America (abstract 53); September 18-20, 1996; New Orleans, La. 12. Dever LL, China C, Eng RH, O’Donovan C, Johanson WG. Vancomycin-resistant Enterococcus faecium in a Veterans Affairs Medical Center: association with antibiotic usage. AJIC Am J Infect Control 1998;26:40-6. 13. Tornieporth NG, Roberts RB, John J, Hafner A, Riley LW. Risk factors associated with vancomycin-resistant Enterococcus faecium infection or colonization in 145 matched case patients and control patients. Clin Infect Dis 1996;23:767-72. 14. Quale J, Landman D, Saurina G, Atwood E, DiTore V, Patel K. Manipulation of a hospital antimicrobial formulary to control an outbreak of vancomycin-resistant enterococci. Clin Infect Dis 1996;23:1020-5. 15. Anglim AM, Klym B, Byers KB, Scheld WM, Farr BM. Effect of a vancomycin restriction policy on ordering practices during an outbreak of vancomycin-resistant Enterococcus faecium. Arch Intern Med 1997;157:1132-6. 16. Johnson SV, Hoey LL, Vance-Bryan K. Inappropriate vancomycin prescribing based on criteria from the Centers for Disease Control and Prevention. Pharmacotherapy 1995;15:570-85. 17. Rybak MJ, Palmer SM. Vancomycin-resistant enterococci. Pharmacotherapy 1995;15:684-5. 18. Ena J, Dick RW, Jones RN, Wenzel RP. The epidemiology of intravenous vancomycin usage in a university hospital. JAMA 1993;269:598-602.

April 1999

19. Evans ME, Kortas KJ. Vancomycin use in a university medical center: comparison with hospital infection control practices advisory committee guidelines. Infect Control Hosp Epidemiol 1996;17:356-9. 20. Misan GM, Martin ED, Smith ER, Somogyi AA, Bartholomeusz RC, Bochner F. Drug utilization review in a teaching hospital: experience with vancomycin. Eur J Clin Pharmacol 1990;39:457-61. 21. Nightingale J, Chaffee BW, Colvin CL, Alexander MR, Rice T, Ross MB. Retrospective evaluation of vancomycin use in a university hospital. Am J Health Syst Pharm 1987;44:1807-9. 22. Fletcher CV, Giese RM, Rodman JH. Pharmacist interventions to improve prescribing of vancomycin and tobramycin. Am J Health Syst Pharm 1986;43:2198-201. 23. Logsdon BA, Lee KR, Luedtke G, Barrett FF. Evaluation of vancomycin use in a pediatric teaching hospital based on CDC criteria. Infect Control Hosp Epidemiol 1997;18:780-2. 24. Watanakunakorn C. Prescribing pattern of vancomycin in a community teaching hospital with low prevalence of vancomycin-resistant enterococci. Infect Control Hosp Epidemiol 1997;18:767-9. 25. Salemi C, Becker L, Morrissey R, Warmington J. A clinical decision process model for evaluating vancomycin use with modified HICPAC guidelines. Clin Perform Qual Health Care 1998;6:12-6. 26. Morgan AS, Brennan PJ, Fishman NO. Impact of a vancomycin restriction policy on use and cost of vancomycin and incidence of vancomycin-resistant Enterococcus. Ann Pharmacother 1997;31:970-3. 27. Manipulating antibiotic use patterns may lower VRE [editorial]. Hosp Infect Control 1997;24(12):186-7. 28. Belliveau PB, Rothman AL, Maday CE. Limiting vancomycin use to combat vancomycin-resistant Enterococcus faecium. Am J Health Syst Pharm 1996;53:1570-5. 29. Lynch P, Cummings MJ, Roberts PL, Herriott MJ, Yates B, Stamm WE. Implementing and evaluating a system of generic infection precautions: body substance isolation. AJIC Am J Infect Control 1990;18:1-12. 30. Polk R. Vancomycin prescribing in the era of vancomycin-resistant enterococci [editorial]. Pharmacotherapy 1995;15:682-3. 31. Reichley RM, Miller JE, Sahm DF, Mundy LM, Fraser VJ, Kwhn MG, et al. Use of an administrative database to estimate vancomycin utilization and appropriateness. In: Programs and abstracts of the 34th Annual Meeting of the Infectious Diseases Society of America (abstract 41); September 18-20, 1996; New Orleans, La. 32. Rai M, Bobek MB, Capozzi D, Lee P, Goldman M, Longworth D, et al. Vancomycin use in a tertiary care referral center, compliance with the CDC recommendations for preventing the spread of vancomycin resistance. In: Programs and abstracts of the 34th Annual Meeting of the Infectious Diseases Society of America (abstract 3); September 18-20, 1996; New Orleans, La.