Effectiveness of a multifaceted infection control policy in reducing vancomycin usage and vancomycin-resistant enterococci at a tertiary care cancer centre

Journal of Hospital Infection (2002) 51: 52±58 doi:10.1053/jhin.2002.1161, available online at http://www.idealibrary.com on

Effectiveness of a multifaceted infection control policy in reducing vancomycin usage and vancomycin-resistant enterococci at a tertiary care cancer centre Z. H. A. Shaikh*, C. A. Osting*, H. A. Hanna*, R. B. Arbuckley, J. J. Tarrandz and I. I. Raad* *Department of Infectious Diseases, Infection Control, and Employee Health, yDivision of Pharmacy and zDepartment of Microbiology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA Summary: We undertook a prospective cohort study to evaluate the role of a multifaceted infection control policy including the use of a `vancomycin order form,' in decreasing the transmission of vancomycin-resistant enterococci (VRE). In January 1997, a multifaceted infection-control policy was implemented amongst patients admitted to the M. D. Anderson Cancer Center in whom neutropenic fever developed or who were found to be colonized or infected with VRE. As part of this programme, we initiated the use of a vancomycin order form to reduce the use of empirical vancomycin. The total incidence of VRE infections declined from 0.437/1000 patient days in 1996±97 to 0.229/1000 patient days in 1998±99 (P ˆ 0.008). The VRE bloodstream infections declined from 0.338/1000 patient days in 1996±97 to 0.181/1000 patient days in 1998±99 (P ˆ 0.027). Empiric vancomycin use decreased from 416 g/1000 patient days in 1996±97 to 208 g/1000 patient days in 1998±99 (P < 0.001), resulting in a decreased vancomycin cost from $2561/1000 patient days in 1996±97 to $1195/1000 patient days in 1997±98 (P < 0.001). We conclude that a multifaceted infection control policy incorporating the use of a vancomycin order form can effectively decrease the use of empirical vancomycin and can play a role in limiting the spread of VRE in an endemic setting. & 2002 The Hospital Infection Society

Keywords: Infection control; vancomycin-resistant enterococci; vancomycin; antimicrobial restriction program.

Introduction Vancomycin-resistant enterococci (VRE) have been established as significant pathogens in the majority of This manuscript was presented in part as an abstract at the 10th Annual Meeting of the Society for Healthcare Epidemiology of America, Toronto, Ontario, April 1±3, 2001. Received 12 June 2001; revised manuscript accepted 11 December 2001. Author for correspondence: Dr I. I. Raad, Department of Infectious Diseases, Infection Control and Employee Health, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, Texas 77030, USA. Fax: ‡1 713 792 8233; E-mail: [email protected]

0195-6701/02/010052 1 07 $35.00/0

hospitals in the United States and more recently, VRE has been ranked as the third most common cause of nosocomial infections.1 In an older report from hospitals participating in the National Nosocomial Infection Surveillance System, enterococcus was ranked as the second most common nosocomial pathogen.2 The report also demonstrated an overall increase in the incidence of VRE from 0.3% to 7.9% between 1989 and 1993. The first case of VRE was identified at the M. D. Anderson Cancer Center in February 1994; however, there did not appear to be a significant threat of VRE until an increase in the number of & 2002 The Hospital Infection Society

Infection control and VRE

sporadic cases began in April 1996, followed by an outbreak in December 1996. This epidemic, involving 17 patients, was controlled by aggressive infection control measures.3 VRE however, continued to remain an endemic problem at our institution. Whilst there are various reasons for the persistence of VRE, indiscriminate use of vancomycin was one of the most challenging to resolve. We describe here the successful implementation of a multifaceted infection-control policy that included the use of a `vancomycin order form,' to reduce the use of empirical vancomycin and control the spread of VRE at the M. D. Anderson Cancer Center. The program was developed to accomplish the long-term goal of decreasing the emergence of VRE, preventing its associated morbidity and mortality. In this process, cost savings due to decreased vancomycin usage was perceived as a substantial secondary benefit. Methods Hospital setting The University of Texas M. D. Anderson Cancer Center is a 417-bed tertiary care cancer centre. Each year, about 40 000 new patients are seen at this institution and 600 blood and marrow transplants are performed. There are 16 medical intensive care unit (MICU) beds and 20 surgical ICU beds. VRE surveillance Rectal swabs were obtained from all patients on the leukaemia and bone marrow transplant (BMT) services in January 1997 in an attempt to identify all the patients colonized with VRE. Subsequently, rectal surveillance cultures were performed on all patients upon admission to the leukaemia service, the BMT service and MICU, and at weekly intervals thereafter. Microbiological methods Rectal swabs were inoculated on Campylobacter agar plates with 10% sheep blood and five antimicrobials including nominal vancomycin concentration 10 mg/ mLÐapparent available concentration 6 mg/mL (BBL, Cockeysville, MD, USA), and incubated at 35 C for 24±48 h. Isolates from colonies with typical morphology that were catalase-negative and pyrrolidonyl-arylamidase positive were then

53

suspended in trypticase soy broth (BBL) to 0.5 McFarland and subsequently inoculated to vancomycin screen plates, vancomycin 6 mg/mL (BBL) and blood agar plates (BBL) with ampicillin disc. Growth on vancomycin plates and resistance to ampicillin indicated isolation of a VRE and was reported as such. Isolates collected during the period of VRE outbreak were compared for clonal relatedness by pulsed-field gel electrophoresis (PFGE) using SmaI restriction endonuclease digestion of chromosomal DNA.4 Infection control policy In an effort to control the VRE outbreak at our institution, a number of infection control measures were initiated, including hand washing; contact isolation of patients with VRE; disposable, dedicated-use equipment (i.e., stethoscope) when possible; thorough cleaning of non-disposable equipment (i.e., portable X-ray machine) with a germicidal solution; cohorting of patients with VRE with nursing personnel/respiratory technicians; and terminal environmental cleaning of rooms upon discharge of patients. In an effort to decontaminate the MICU, it was closed to new admissions on January 10±12 and 25±26, 1997, and all surfaces were wiped with Wex-Cide1 or Spectrum1 germicidal solution. Weekly VRE surveillance was implemented for all high-risk patients. Patients identified as colonized or infected with VRE during current or prior hospitalization remained in contact isolation with staff and visitors required to use gowns and gloves until three consecutive weekly rectal surveillance cultures were found to be negative for VRE. Highrisk patients with unknown VRE status on admission were placed in contact isolation until a negative VRE screen was documented. These measures were streamlined in the guidelines issued to the leukaemia and BMT services and to the MICU on 24 January 1997 (Figure 1). The infection control staff closely monitored the isolation practices and educated staff and visitors regarding the importance of isolation precautions. The routine use of isolator tubes for quantitative blood cultures was encouraged on the leukaemia and BMT services in an attempt to differentiate between contamination and infection. In addition, central venous catheters (CVC) impregnated with minocycline and rifampin were made available in the MICU where two cases of CVC-associated VRE bacteraemia had occurred.

54

Z. H. A. Shaikh et al.

GENERAL PREVENTION MEASURES – Prudent vancomycin use – Education of hospital staff regarding VRE and the importance of adequate handwashing (>15 s) – Handwashing with chlorhexidine gluconate – Early detection and reporting of VRE by microbiology laboratory – Dedicated equipment (stethoscopes, etc.) for VRE patients. – Decontamination of examination material/equipment with alcohol prep after each patient contact – Education of patients and visitors regarding infection control measures – Quantitative blood cultures using isolator tubes

Targeted VRE surveillance of leukaemia, BMT and critically ill patients

Weekly stool/perirectal cultures

Positive for VRE

Negative for VRE

SPECIFIC PREVENTION MEASURES – Gloves and gown when entering room – Cohorting of patients, if possible – Daily cleaning of frequently touched environmental surfaces (telephone, bed rails, stethoscopes, etc.) with phenolic disinfectant and allowing time to dry – Discontinue isolation after three consecutive weekly cultures become negative for VRE, but continued surveillance

Continue weekly surveillance –if possible

POST DISCHARGE – Disinfect medical devices and environmental surfaces, including bed rails, telephones, stethoscopes, etc. – Contact isolation at each healthcare visit until three VRE cultures are negative Figure 1 Guidelines for vancomycin-resistant enterococci (VRE) surveillance at the M. D. Anderson Cancer Center.

Vancomycin restriction programme Initial guidelines for the use of vancomycin at the M. D. Anderson Cancer Center were established on 15 January 1993 and were similar to the guidelines published by the Hospital Infection Control Practices Advisory Committee (HICPAC) in 1995.5 Unfortunately, these guidelines were only partially

implemented with regards to the use of oral vancomycin and were never implemented for the use of intravenous vancomycin. Although the recommendations also included establishment of the vancomycin order form for the use of empirical vancomycin, they were not placed into effect at that time. Following the VRE outbreak in the MICU in December 1996, a multifaceted infection-control

Infection control and VRE

approach was undertaken, and the concept of having a pre-printed vancomycin order form that included a checklist of approved indications for the use of empirical vancomycin was approved in January 1997. In November 1997, an antibiotic taskforce was setup to reinforce and implement these policies. The objective of the vancomycin-restriction programme was to limit use of empirical vancomycin therapy in febrile neutropenic patients to one of the following four situations: (1) history of prior colonization with methicillin-resistant Staphylococcus aureus or penicillin-resistant Streptococci; (2) severe (grade 3 or 4) mucositis with or without hypotension (BP < 90); (3) clinical signs of CVC site infection (erythema, discharge); or (4) postneurosurgical meningitis with no use of vancomycin in the previous 30 days. Any use of empirical vancomycin longer than 72 h either required the approval of the attending infectious diseases physician or needed to be substantiated with a blood culture positive result for an organism susceptible to vancomycin alone. The attending physician was required to complete a vancomycin order form if neutropaenic fever developed and one of the four indications listed above were satisfied and if the patient had not received vancomycin in the previous 30 days. Prior to the release of vancomycin to the nursing personnel, a pharmacist was required to make sure that the patient qualified for vancomycin, based on information supplied on the form. Programme evaluation We evaluated the effectiveness of the infection control measures in decreasing the incidence of VRE and compared the total number of VRE infections, as well as the number of VRE bloodstream infections (BSI) for the fiscal period 1996±97 to fiscal period 1998±99. We also evaluated the impact of the vancomycin order form in reducing the use of empirical intravenous vancomycin in the setting of febrile neutropenia by comparing the cost and total amount of vancomycin used for the two fiscal-year periods, 1996±97 and 1998±99. Definitions Fever was defined as temperature >38.3 C. Neutropenia was defined as absolute neutrophil count <1000/mm3. Total VRE infections and

55

VRE BSI was defined by standard criteria of the Centers for Disease Control and Prevention (CDC).6 Statistical analysis Data were analysed using Epi Info 2000 software, version 2000 (CDC, Atlanta, GA, USA). Patient characteristics were compared using the Chi-square test for categorical data and the Student's t-test for continuous variables. Significance was defined as a two-tailed P value of 0.05. Results The infection control measures that were initiated in January 1997 following the VRE outbreak were strictly enforced by February 1997. Most of these measures, including VRE surveillance of high-risk patients continued, even after the outbreak was brought under control, except for cohorting of VRE patients and monitoring of the infection control/ isolation practices. By May 1997, contact isolation measures were eased with respect to newly admitted patients awaiting VRE screening results. VRE surveillance In January 1997, patients on the leukaemia (N ˆ 39) and BMT (N ˆ 34) services were screened to determine the prevalence of VRE colonization/infection. Among these 73 patients, five (7%) were found to be positive. Of these, four (10%) were among the 39 leukaemia patients and one (3%) was among the 34 BMT patients. The VRE isolates tested during the VRE outbreak between December 1996 and February 1997 were all found to be clonally related by PFGE (data not shown), and traced back to a patient admitted with VRE bacteraemia in September 1996. Since then, multiple strains of VRE have established endemicity at our institution pointing to external sources for VRE, as has been demonstrated in another study.7 VRE infections As illustrated in Figure 2, the number of total VRE infections declined from 0.437/1000 patient days in 1996±97 to 0.229/1000 patient days in 1998±99 [relative risk ratio, 1.91 (95% CI, 1.20±3.06); P ˆ 0.008], whereas the VRE BSI declined from 0.338/1000 patient days in 1996±97 to 0.181/1000

56

Z. H. A. Shaikh et al.

patient days in 1998±99 [relative risk ratio, 1.86 (95% CI, 1.10±3.16); P ˆ 0.027]. Vancomycin usage The use of the vancomycin order form resulted in a decrease in the use of empirical intravenous vancomycin from 416 g/1000 patient days in 1996±97, to 208 g/1000 patient days in 1998±99 (P < 0.001) (Figure 3). This resulted in a substantial decrease (>50%) in the cost of vancomycin from

1 0.9

0.7 0.6

8

1

9 0. 18

0.

22

0.

27

28

8 33 0.

0.3

0.

0.4

6

43

7

0.5

0.

Per 1000 Pt days

0.8

0.2 0.1 0

FY 96/97

FY 97/98

FY 98/99

Figure 2 Incidence rates of nosocomial vancomycin-resistant enterococci (VRE) infections by fiscal year. &, total infections; , bloodstream infections.

125 000

60 000

40 000 115 000 20 000

Vancomycin (g)

Patient days

120 000

110 000

8 97 –9 FY

96 –9 FY

95 –9 FY

7

0

6

105 000

Figure 3 Vancomycin useage by fiscal year. Ð&Ð, patient days; --^ - - - -, vancomycin (g).

$2561/1000 patient days in 1996±97 to $1195/1000 patient days in 1997±98 (P < 0.001).

Discussion VRE appears to have become an endemic problem in the majority of hospitals in the United States, especially in the intensive-care setting. Even the use of enhanced infection-control measures have at times been unsuccessful in eradicating the VRE transmission,8,9 which has led to a shift in focus from eradicating VRE to slowing the spread of VRE colonization and infection.10 Efforts are now being directed to combining infection-control measures with restriction of antimicrobials. This approach has been substantiated at a Veterans Administration hospital, where infection-control measures alone were unsuccessful in decreasing the prevalence of VRE. When infection control was combined with an antimicrobial restriction programme, there was a decrease in VRE colonization in the endemic setting.11 The efficacy of infection-control measures (such as the cohorting of patients with VRE and nursing personnel) and antibiotic restriction, has also been demonstrated in a mathematical model looking at the transmission dynamics of VRE in the MICU.12 Vancomycin is the second most commonly prescribed antibiotic in the intensive-care setting in hospitals in the United States.13 Inappropriate use of vancomycin is well recognized, especially for empirical therapy.14,15 Previous studies have demonstrated the effectiveness of a vancomycin order form,16,17 a vancomycin continuation form,18 and vancomycin restriction policies19,20 in reducing inappropriate vancomycin use. However, decrease in vancomycin usage due to vancomycin restriction should not be assumed to translate into reduction of VRE incidence. Although a temporal relationship may be intuitive, vancomycin use was not found to be significantly associated with the outbreak of VRE in a tertiary care centre.21 Furthermore, two studies failed to demonstrate a reduction in VRE colonization by combining vancomycin restriction with the use of barrier precautions.22,23 In an effort to contain the spread of VRE infection in our institution, we initiated a multifaceted infection control policy that focused on strengthening the infection control measures. Similar comprehensive infection control measures implemented in a community-wide effort in the Siouxland region of

Infection control and VRE

Iowa, Nebraska and South Dakota were shown to effectively reduce or eliminate the transmission of VRE in acute and long-term healthcare facilities.24 We also incorporated a vancomycin order form as a means to reduce the use of empirical vancomycin. The vancomycin-restriction programme was postulated as a self-pilot programme without any enforcement mechanism. Due to staffing issues, it was considered impractical to review patients' charts to verify the accuracy of this information. Even if the information were found to be incorrect, there remained the issue of who would confront the attending physician. The role of the antibiotic task force was to ensure that vancomycin was not continued beyond 72 h without a positive blood culture substantiating a Gram-positive organism susceptible only to vancomycin or without approval of the infectious disease physician. The programme also assumed that the team co-ordinating patient care, including attending physicians, residents/fellows, the clinical nurse specialist and the pharmacist, would be knowledgeable as to institutional policies and CDC/HICPAC guidelines. Although the recommendations for the prevention of the spread of vancomycin-resistant organisms were initially circulated to all departments, the programme did not take into account the constant turnover of interns and fellows, clinical nurse specialists, or pharmacists, who, in most cases, would be the ones to fill out the vancomycin order form. There are several limitations in our study. First and foremost, because a number of measures were all initiated within the same time period, it is difficult to ascribe the success in reducing VRE to any one of these measures. It is likely that all our efforts contributed to the fulfilment of our study goal. Second, because the study was conducted over a period of time, there might be some underlying changes (i.e., baseline patient characteristics, increasing staff awareness of the VRE problem and a decrease in use of other antimicrobials) that could have influenced the results. Third, our study did not measure the compliance rate of the various infection control measures, which makes it difficult to judge their effectiveness and their role in containing the spread of VRE. Finally, this study was conducted at a single hospital providing specialized care to cancer patients, which may potentially limit the generalization of the results to other settings. Although, several factors may have contributed to the success of our multifaceted infection-control

57

policy, we believe that the inter-disciplinary effort in controlling the VRE outbreak played the most important role. In conclusion, it appears that a multifaceted infection-control policy incorporating the use of a vancomycin order form can effectively decrease the use of empirical vancomycin and can play a role in limiting the spread of VRE in an endemic setting.

References 1. Jones RN, Marshall SA, Pfaller MA et al. Nosocomial enterococcal blood stream infections in the SCOPE Program: Antimicrobial resistance, species occurrence, molecular testing results, and laboratory testing accuracy. SCOPE Hospital Study Group. Diagn Microbiol Infect Dis 1997; 29: 95±102. 2. Centers for Disease Control and Prevention. Nosocomial enterococci resistant to vancomycinÐUnited States, 1989±1993. Morb Mortal Wkly Rep CDC Surveillance Summary 1993; 42: 597±599. 3. Hanna H, Umphrey J, Tarrand J, Mendoza M, Raad I. Management of an outbreak of vancomycin-resistant enterococci in the medical intensive care unit of a cancer center. Infect Control Hosp Epidemiol 2001; 22: 217±219. 4. Murray BE, Singh KV, Heath JD, Sharma BR, Weistock GM. Comparison of genomic DNAs of different enterococcal isolates using restriction endonucleases with infrequent recognition sites. J Clin Microbiol 1990; 28: 2059±2063. 5. Hospital Infection Control Practices Advisory Committee. Recommendations for preventing the spread of vancomycin resistance. Infect Control Hosp Epidemiol 1995; 16: 105±113. 6. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988; 16: 128±140. 7. Bonten MJM, Slaughter S, Hayden MK, Nathan C, van Voorhis J, Weinstein RA. External sources of vancomycin-resistant enterococci for intensive care units. Crit Care Med 1998; 12: 2001±2004. 8. Lai KK, Kelley AL, Melvin ZS, Belliveau PP, Fontecchio SA. Failure to eradicate vancomycinresistant enterococci in a university hospital and the cost of barrier precautions. Infect Control Hosp Epidemiol 1998; 19: 647±652. 9. Montecalvo MA, Jarvis WR, Uman J et al. Infectioncontrol measures reduce transmission of vancomycinresistant enterococci in an endemic setting. Ann Intern Med 1999; 17: 269±272. 10. Hayden MK. Insights into the epidemiology and control of infection with vancomycin-resistant enterococci. Clin Infect Dis 2000; 31: 1058±1065. 11. Quale J, Landman D, Saurina G, Atwood E, DiTore V, Patel K. Manipulation of a hospital antimicrobial formulary to control an outbreak of vancomycinresistant enterococci. Clin Infect Dis 1996; 23: 1020±1025.

58 12. Austin DJ, Bonten MJ, Weinstein RA, Slaughter S, Anderson RM. Vancomycin-resistant enterococci in intensive-care hospital settings: transmission dynamics, persistence, and the impact of infection control programs. Proc Natl Acad Sci USA 1999; 96: 6908±6913. 13. Gaynes R. The impact of antimicrobial use on the emergence of antimicrobial-resistant bacteria in hospitals. Infect Dis Clin Am 1997; 11: 757±765. 14. Ena J, Dick RW, Jones RN, Wenzel RP. The epidemiology of intravenous vancomycin usage in a university hospital: A 10-year study. JAMA 1993; 269: 598±602. 15. Roghmann M-C, Perdue BE, Polish L. Vancomycin use in a hospital with vancomycin restriction. Infect Control Hosp Epidemiol 1999; 20: 60±63. 16. Adachi W, Bolding F, Armstrong R. Experience with vancomycin education and order sheet to limit vancomycin use. Hospital Pharmacy 1997; 32: 1370±1373. 17. Burke CE, Piper J, Holloway W. Order form for restricting vancomycin prescribing. Am J Health Syst Pharm 1997; 54: 1893±1897. 18. Evans ME, Millheim ET, Rapp RP. Vancomycin use in a university medical center: Effect of a vancomycin continuation form. Infect Control Hosp Epidemiol 1999; 20: 417±420.

Z. H. A. Shaikh et al. 19. Morgan AS, Brennan PJ, Fishman NO. Impact of a vancomycin restriction policy on use and cost of vancomycin and incidence of vancomycin-resistant entterococcus. Ann Pharmacother 1997; 31: 970±973. 20. Anglim AM, Klym B, Byers KE, 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±1136. 21. Quale J, Landsman D, Atwood E et al. Experience with a hospital-wide outbreak of vancomycinresistant enterococci. Am J Infect Control 1996; 24: 372±379. 22. Wells CL, Juni BA, Cameron SB et al. Stool carriage, clinical isolation, and mortality during an outbreak of vancomycin-resistant enterococci in hospitalized medical and/or surgical patients. Clin Infect Dis 1995; 21: 45±50. 23. Morris JG Jr, Shay DK, Hebden JN et al. Enterococci resistant to multiple antimicrobial agents, including vancomycin: establishment of endemicity in a university medical center. Ann Intern Med 1995; 123: 250±259. 24. Ostrowsky BE, Trick WE, Sohn AH et al. Control of vancomycin-resistant Enterococcus in health care facilities in a region. N Engl J Med 2001; 344: 1427±1433.