Received 4 July 2010; returned 10 August 2010; revised 27 September 2010; accepted 30 September 2010

J Antimicrob Chemother 2011; 66: 205 209 doi:10.1093/jac/dkq404 Advance Access publication 8 November 2010 Sustained reduction in antimicrobial use and decrease in methicillin-resistant Staphylococcus aureus and Clostridium difficile infections following implementation of an electronic medical record at a tertiary-care teaching hospital Paul P. Cook 1 *, Shemra Rizzo 2, Michael Gooch 3, Michelle Jordan 3, Xiangming Fang 4 and Suzanne Hudson 4 1 Division of Infectious Diseases, Department of Medicine, Brody School of Medicine at East Carolina University, Greenville, NC, USA; 2 Department of Biostatistics, University of California at Los Angeles, Los Angeles, CA, USA; 3 Pitt County Memorial Hospital, Greenville, NC, USA; 4 Department of Biostatistics, East Carolina University, Greenville, NC, USA *Corresponding author. Tel: +1-252-744-5700; Fax: +1-252-744-3472; E-mail: cookp@ecu.edu Received 4 July 2010; returned 10 August 2010; revised 27 September 2010; accepted 30 September 2010 Objectives: We evaluated the effect of implementation of an electronic medical record (EMR) on the use of antimicrobial agents and on the rates of infections with Clostridium difficile and methicillin-resistant Staphylococcus aureus (MRSA). Methods: This was a retrospective, observational study conducted between 1 January 2005 and 31 December 2009. Antimicrobial drug use, rates of nosocomial C. difficile infection (CDI) and MRSA infection, the number of medical charts reviewed and number of antimicrobial recommendations made and accepted were compared before and after implementing the EMR utilizing interrupted time-series analysis. Results: Compared with the 10 quarters prior to implementing the EMR, there was a 36.6% increase in the number of charts reviewed (P, 0.0001), a 98.1% increase in the number of antimicrobial recommendations made (P, 0.0001) and a 124% increase in the number of recommendations accepted (P, 0.0001). There was a 28.8% decrease in the use of 41 commonly used antibacterial agents (P, 0.0001). Nosocomial CDI decreased by 18.7% (P¼0.07) and nosocomial MRSA infections decreased by 45.2% (P, 0.0001) following implementation of the EMR. Conclusions: Adoption of an EMR facilitated a significant increase in chart reviews and antimicrobial recommendations, which resulted in a sustained decrease in antimicrobial use. There were decreased nosocomial infections with MRSA and a trend towards decreasing CDIs following implementation of the EMR. Keywords: antibiotic management, antibiotic resistance, nosocomial infections Introduction The use and overuse of antimicrobial agents has led to the development of antibiotic resistance in both the community and the hospital setting. Recommendations from various experts and organizations have promoted the prudent use of antimicrobial agents as a means of controlling resistance. 1 3 Recent guidelines from the Infectious Diseases Society of America (IDSA) promote the concept of an electronic medical record (EMR) with computerized physician order entry (CPOE) as a means of facilitating the goals of an antimicrobial stewardship programme (ASP). 3 Our hospital adopted an EMR with CPOE in July 2007. We examined the hospital use of antimicrobial agents before and after implementation of the EMR at our institution. We also determined the incidence of nosocomial infections with Clostridium difficile and methicillin-resistant Staphylococcus aureus (MRSA) during the same time period, as these infections are associated with antimicrobial use. 2,4 8 Methods Pitt County Memorial Hospital is an 861-bed, tertiary-care, teaching hospital affiliated with The Brody School of Medicine at East Carolina University. The hospital has a busy trauma unit, cardiothoracic surgery service, orthopaedic service, paediatric unit, oncology service and dialysis unit. Renal transplantations are performed at the hospital, but there are no bone marrow transplantation or burn units. Ethics approval was not required for this study because we used only routine data collected on # The Author 2010. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org 205

Cook et al. hospital wards. This was an observational study from 1 January 2005 until 31 December 2009. Our ASP was established in 2001 and has been described previously. 9 Antimicrobial drug use was measured in defined daily doses per 1000 patient days (DDDs/1000 PDs) using the standards of the WHO (www.whocc.no/atcddd/). Beginning in July 2007, Pitt County Memorial Hospital implemented an EMR (EPIC, Madison, WI, USA) with CPOE. Recommended dosing was built into the order screens for antibiotics following national guidelines. Reports for the ASP pharmacists were built into the EMR and were reviewed daily. The programme operated 5 days a week. The report listed all patients receiving antimicrobial agents continuously for at least 48 h. Information in the report included patient location, patient weight, allergies and ordering physician. The EMR allowed the ASP pharmacists to open an intervention on a patient and allowed documentation within the report on what recommendation was made. If a patient had cultures that had not been finalized, the intervention could be left open and a report run later listing the open interventions for follow-up. If a patient s initial antibiotic regimen was deemed appropriate, but follow-up was desired to ensure the patient received the recommended time course of therapy, the intervention could also be left open for that purpose. Information entered into the interventional database stayed with the patient s intervention record and could be seen if that patient came up for review at a later date. Reports were also available to calculate antibiotic usage overall, by unit and by service, allowing targeted education efforts. The ASP pharmacists entered their recommendations as a progress note and then entered an order for antimicrobial monitoring. This triggered a pop-up screen whenever a physician, nurse practitioner or physician s assistant entered that particular patient s record, directing them to the recommendation in the progress note. The physician had to acknowledge seeing the record at the time. It was possible for the pharmacists to determine whether the pop-up had been triggered and who had seen it. The antimicrobial management programme reviewed 49 antimicrobial agents used at the hospital (Table 1). Broad-spectrum antibiotics were divided into seven classes. Extended-spectrum penicillins included ampicillin, amoxicillin, ampicillin/sulbactam, piperacillin and piperacillin/ tazobactam. Cephalosporins included cefazolin, cefadroxil, cefalexin, cefuroxime, cefotetan, ceftriaxone, ceftazidime, cefotaxime and cefepime. The quinolone class included moxifloxacin and ciprofloxacin. Table 1. List of antibacterial and antifungal agents that were monitored as part of the ASP Antibacterial agents amikacin, amoxicillin, amoxicillin/clavulanate, ampicillin, ampicillin/sulbactam, azithromycin, aztreonam, cefaclor, cefadroxil, cefazolin, cefepime, cefotaxime, cefoxitin, cefotetan, ceftazidime, ceftriaxone, cefuroxime, cefalexin, ciprofloxacin, clarithromycin, clindamycin, daptomycin, doripenem, doxycycline, ertapenem, erythromycin, gentamicin, imipenem, linezolid, meropenem, metronidazole, moxifloxacin, nafcillin, piperacillin, piperacillin/tazobactam, sulfamethoxazole/trimethoprim, telithromycin, tetracycline, tigecycline, tobramycin, vancomycin Antifungal agents amphotericin B, anidulafungin, caspofungin, fluconazole, itraconazole, micafungin, posaconazole, voriconazole Imipenem was replaced by meropenem in the hospital formulary in 2006; meropenem was replaced by doripenem in 2008. Carbapenems included ertapenem, doripenem, imipenem and meropenem. Macrolides included azithromycin, erythromycin, clarithromycin and telithromycin. Aminoglycosides included amikacin, gentamicin and tobramycin. Tetracyclines/glycylcyclines included tetracycline, doxycycline and tigecycline. Clindamycin, aztreonam, vancomycin, nafcillin, linezolid, daptomycin, metronidazole and trimethoprim/sulfamethoxazole were examined separately. Stool samples were tested for the presence of C. difficile toxin using a standard cell culture cytotoxicity assay with MRC-5 lung fibroblast cells and antitoxin from TechLab (Blacksburg, VA, USA). Clinical cultures for S. aureus included blood, CSF, wounds, respiratory specimens, pleural fluid, tissue specimens and urine. Surveillance cultures were not included. In vitro bacterial susceptibilities were determined using the MicroScan system (Dade Behring). The nosocomial C. difficile and MRSA datasets were created by querying MedMined (CareFusion, Birmingham, AL, USA). In an effort to limit the analysis to nosocomial infections, only cultures or stool specimens from patients who were in hospital more than 3 days following admission were included for evaluation. Duplicate positive test results from the same patient within a 3 month period were counted only once. Rates of nosocomial C. difficile and MRSA infections were expressed as cases per 10000 patient hospital days. In evaluating the effect of implementation of EMR on the use of antibiotic agents and on the rates of infections with C. difficile and MRSA, we adopted interrupted time-series models to control for the serial correlation among the data. It turned out that an interrupted AR(1) model (where AR stands for autoregressive) was able to sufficiently account for the autocorrelation and thus was selected for our final analysis. The model can be formulated as follows: Y t O/ Y t 1 = b 0 + b 1 Intervention + e t where Y t is the response value at time t, Ø is the lag 1 coefficient of the AR(1) model, e t is the measurement error and Intervention is an indicator variable whose value is 0 for timepoints before EMR was implemented and 1 for timepoints after EMR was implemented. Due to the fact that our time-series data are fairly stationary both before and after the introduction of EMR, we did not include a linear trend in the model above. Thus b 0 is (12Ø) times the mean response before EMR was implemented and b 1 is the change in (12Ø) times the mean response due to the implementation of EMR. As a result, testing on the effect of EMR is equivalent to testing on the significance of b 1. This interrupted AR(1) model was fitted using the ARIMA procedure in SAS 9.2. This statistical method was used to compare antibiotic use, number of charts reviewed, number of recommendations made, number of recommendations accepted, demographic data and rates of C. difficile infection (CDI) and MRSA infection with P,0.05 as the level of significance. Linear regression was used to assess hand washing compliance rates over the 5 year period of the study. Spearman s rank correlation coefficient was used to correlate antimicrobial use with rates of CDI and MRSA infection. Results Effect of EMR on chart reviews There were statistically significant increases in the number of charts reviewed, number of antimicrobial recommendations made and number of accepted recommendations following implementation of the EMR (Table 2). Compared with the 10 quarters prior to implementing the EMR, there was a 36.6% increase in the number of charts reviewed (P, 0.0001), a 98.1% increase in the number of antimicrobial recommendations made (P,0.0001) and a 124% increase in the number of recommendations accepted (P, 0.0001). Recommendations for antimicrobial changes were made on 23.6% of the charts 206

Effect of EMR on antimicrobial use, MRSA infections and CDIs JAC reviewed prior to implementation of the EMR; this number increased to 34.3% following implementation of the EMR. The percentage of recommendations that were accepted increased from 81.2% prior to implementation of the EMR to 91.7% following implementation of the EMR. The time to implementation of ASP recommendations decreased from 23 h prior to the EMR to 8 h after the EMR was implemented (P¼0.0001). Antibacterial drug use For the 41 antibacterial drugs monitored, there was a 28.8% decrease in use following implementation of the EMR (Figure 1 and Table 3). There were significant decreases in use of Table 2. Mean number of charts reviewed, number of recommendations made and number of recommendations accepted per quarter before and after implementation of the EMR No. of charts reviewed/ quarter No. of charts recommendations/ quarter No. of charts recommendations accepted/quarter #/10 000 PDs 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 Pre-EMR Post-EMR Percentage change MRSA infections/10 000 PDs CDIs/10 000 PDs DDDs/1000 PDs 2005-1st 2005-2nd 2005-3rd 2005-4th 2006-1st 2006-2nd 2006-3rd 2006-4th 2007-1st 2007-2nd 2007-3rd 2007-4th 2008-1st 2008-2nd 2008-3rd 2008-4th 2009-1st 2009-2nd 2009-3rd 2009-4th 900.0 800.0 700.0 600.0 500.0 400.0 300.0 200.0 100.0 Figure 1. Antibacterial drug use, nosocomial MRSA infection rate and nosocomial CDI rate plotted over the study time period. The time units are quarters. The EMR with CPOE was implemented at the beginning of the third quarter of 2007. Antibacterial drug use is measured in DDDs/ 1000 PDs. The rates of MRSA infection and CDI are expressed as the number of cases per 10 000 patient days (#/10 000 PDs). 0.0 P value 1527 2085 +36.6,0.0001 361 715 +98.1,0.0001 293 656 +124,0.0001 P values are based on interrupted time-series analysis. DDDs/1000 PDs Table 3. Antibacterial agent use before and after implementing the EMR Antimicrobial agent(s) extended-spectrum penicillins, cephalosporins, clindamycin, quinolones, macrolides, metronidazole, vancomycin and aminoglycosides. The use of cefazolin, the drug of choice for surgical prophylaxis, decreased by 53% (81 DDDs/1000 PDs to 38 DDDs/ 1000 PDs) following implementation of the EMR (P,0.0001). There were no significant changes in use of nafcillin, daptomycin, tetracyclines/glycylcyclines, trimethoprim/sulfamethoxazole or linezolid. Carbapenem use increased by 51.3% (P, 0.0029). Antifungal drug use Pre-EMR, DDDs/ 1000 PDs Post-EMR, DDDs/ 1000 PDs Percentage change P value Extended-spectrum 167.3 115.7 230.8 0.0017 penicillins Cephalosporins 206.9 141.2 231.7,0.0001 Quinolones 75.2 32.6 256.6,0.0001 Clindamycin 5.3 3.2 240.0 0.001 Tetracyclines/ 11.9 14.3 +20.5 0.33 glycylcyclines Carbapenems 19.4 29.4 +51.3 0.003 Macrolides 78.7 54.6 230.6 0.003 Daptomycin 6.4 7.4 +15.2 0.85 Linezolid 12.6 12.2 23.0 0.71 Nafcillin 11.2 11.6 +3.1 0.83 Vancomycin 78.4 56.5 228.1,0.0001 Metronidazole 30.8 17.6 242.9,0.0001 Trimethoprim/ 34.7 41.2 +18.7 0.43 sulfamethoxazole Aminoglycosides 30.2 14.7 251.3,0.0001 All antibacterial agents 775.3 552.2 228.8,0.0001 Usage is expressed in DDDs/1000 PDs. Values are means. Pre-EMR and post-emr represent drug use for the 10 quarters prior to and the 10 quarters following implementation of the EMR, respectively. P values are based on interrupted time-series analysis. Total antifungal use decreased by 17.8% (P¼0.01) following implementation of the EMR. MRSA rate There were 807 nosocomial MRSA infections during the study period. Nosocomial MRSA infections decreased by 45.2% (P, 0.0001) following implementation of the EMR (Figure 1). There was a correlation of MRSA infection with total antibacterial agent use; the correlation was strongest with use of quinolones, followed by cephalosporins, then extended-spectrum penicillins (Table 4). Because of the known association of MRSA infections with poor hand hygiene, we examined hand hygiene compliance rates over the 5 year period of the study. The monthly mean compliance rate was 83.7%. Over the course of the study 207

Cook et al. Table 4. Correlation of antibacterial class use with MRSA infections (r MRSA) and CDIs (r CDI) Antimicrobial agents r MRSA r CDI Extended-spectrum penicillins 0.65 (P¼0.002) 0.35 (P¼0.13) Cephalosporins 0.66 (P¼0.001) 0.61 (P¼0.004) Quinolones 0.70 (P¼0.0006) 0.43 (P¼0.055) Total antibacterial agents 0.62 (P¼0.003) 0.38 (P¼0.10) For MRSA infections, the correlation was highest with quinolone use. For CDIs, the correlation was highest with cephalosporin use. period, linear regression analysis revealed a slight, but statistically significant (P¼0.02), decrease in hand hygiene compliance hospital-wide. CDI rate There were a total of 394 nosocomial C. difficile toxin-positive cases during the study period. There were no significant differences in the average length of stay of all patients before and after implementation of the EMR and there were no significant differences in the ages of patients with CDI in the period before and after implementation of the EMR (data not shown). Similarly, there were no differences in the two groups with regard to time in hospital prior to the first positive C. difficile toxin-positive stool (data not shown). The hospital rates for CDI per quarter are shown in Figure 1. There was an 18.7% decrease in the rate of CDI following implementation of the EMR (P¼0.07). The mean rate of nosocomial CDI prior to implementation of the EMR was 3.9 cases/10000 PDs; after the EMR was implemented, the mean rate of nosocomial CDI was 3.2 cases/10000 PDs. The rate of nosocomial CDI correlated best with the use of cephalosporins and, to a lesser extent, quinolones (Table 4). Discussion The IDSA has recommended that hospitals adopt an ASP as a means of controlling unnecessary and inappropriate antibiotic use within hospitals. 3 That same organization has recommended that an EMR be adopted by institutions to facilitate the process of antimicrobial stewardship. 3 There are very few data regarding the effect of computerized programmes on antibiotic use and outcomes. 10,11 Our hospital instituted an ASP in January 2001 and we previously reported significant reductions in antimicrobial use following implementation of that programme. 9 In July 2007 our institution implemented an EMR with CPOE. Our programme involved feedback to the provider once microbiology cultures and susceptibility data were available. The present study demonstrates significant reductions in antimicrobial use following the implementation of this programme. Our data show that the EMR was a tool that allowed our pharmacists (M. G. and M. J.) to review more charts and to make more recommendations. Following implementation of the EMR, the majority of ASP recommendations were accepted within a few hours. Also, the antibiotic order sets that were built into the EMR did not allow inappropriately high doses of antimicrobial agents, as sometimes occurred before the EMR was implemented. By reviewing more charts and having their recommendations accepted sooner, the pharmacists were able to discontinue unnecessary antimicrobial agents in more patients, thus accounting for the large decrease in antimicrobial use. There were large decreases in the use of cephalosporins, extended-spectrum penicillins and quinolones, agents that are 7,8,12 16 highly associated with both MRSA infections and CDIs. Use of cefazolin, the drug of choice for surgical prophylaxis, decreased by 53% following implementation of the EMR. We suspect that the EMR facilitated the appropriate dose and duration of antibiotic use for surgical cases, and eliminated prolonged courses of unnecessary antibiotic prophylaxis in this patient population. With regard to quinolones, ciprofloxacin use was restricted at our institution beginning in July 2007. Quinolones accounted for approximately 15% of the total antibacterial agent use prior to the implementation of the EMR, but that use decreased to 6.5% following implementation of the programme. Our data show a high correlation of quinolone use with MRSA infection, and, to a lesser extent, CDI. Furthermore, our results demonstrate a high correlation of cephalosporin use with both C. difficile and MRSA infections. We speculate that reduced cephalosporin and quinolone use contributed to the decreased nosocomial infection rates of both C. difficile and MRSA. We suspect that the increase in carbapenem use following implementation of the EMR was a result of our efforts to reduce the use of both quinolones and antipseudomonal penicillins (primarily piperacillin/tazobactam) at the hospital. In other words, carbapenems became an alternative agent to quinolones and piperacillin/tazobactam. Despite the large increase in carbapenem use, the total use of these drugs remained less than the use of both the extended-spectrum penicillins and the quinolones (Table 3). There are several limitations associated with an aggregate data study such as this. The duration of antibiotic use for individual patients was not assessed. Also, it is not clear whether all of the C. difficile or MRSA infections began in hospital or prior to hospital admission. We tried to account for this possibility by only including specimens positive for C. difficile toxin or MRSA from patients who had been in hospital for more than 3 days. Antibiotics are only one of many risk factors for the development of CDI or MRSA infection. 17 22 Factors such as renal insufficiency, malignancy, gastrointestinal surgery, use of proton pump inhibitors, previous antibiotic exposures and enteral feeding were not evaluated in this study. Also, nosocomial spread of C. difficile or MRSA was not assessed. Nevertheless, our assessment of compliance with hand washing demonstrated a slight decrease in hand hygiene compliance over the course of the study period, suggesting that the improvements in CDI and MRSA rates were not completely related to improvements in infection control policies. Finally, our hospital began a programme to identify all incoming patients with MRSA colonization of the nares in February 2007. As part of this programme, all patients who screened positive for MRSA were placed in private rooms with strict contact isolation infection control measures. There is no question that infection control policies play a major role in the rates of nosocomial MRSA as well as CDIs. Our data support the concept that programmes that promote both infection control surveillance and antimicrobial stewardship are more likely to be effective than either measure alone. 4 208

Effect of EMR on antimicrobial use, MRSA infections and CDIs JAC Infections with MRSA and C. difficile are associated with an increase in hospital length of stay, high mortality and increased hospital costs. 23 25 Our data support the use of an EMR with CPOE as a means of reducing unnecessary antimicrobial use. Reduction in antimicrobial drug use facilitated by the EMR can indirectly lead to decreased rates of infection with both MRSA and C. difficile. Acknowledgements Portions of this work were presented at the IDSA Meeting, Philadelphia, PA, 2009. We thank Delores Nobles and Kathy Cochran for help with the MedMined queries and hand hygiene data and Joy Barwick for her assistance with the C. difficile data. Funding No specific funding. Data for this manuscript were collected as part of routine surveillance of the hospital antibiotic management programme. Transparency declarations P. P. C. is a member of the speakers bureau of Pfizer, Astellas and Merck. P. P. C. has received research funding from GlaxoSmithKline, Merck, Gilead, Pfizer and Bristol-Myers Squibb. All other authors: none to declare. References 1 MacDougall C, Polk RE. Antimicrobial stewardship programs in healthcare systems. Clin Microbiol Rev 2005; 18: 638 56. 2 Gould IM. Antibiotic policies to control hospital-acquired infection. J Antimicrob Chemother 2008; 61: 763 65. 3 Dellit TH, Owens RC, McGowan JE Jr et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44: 159 77. 4 Aldeyab MA, Monnet DL, Lopez-Lozano JM et al. Modelling the impact of antibiotic use and infection control practices on the incidence of hospital-acquired methicillin-resistant Staphylococcus aureus: a time-series analysis. J Antimicrob Chemother 2008; 62: 593 600. 5 Vernaz N, Sax H, Pittet D et al. Temporal effects of antibiotic use and hand rub consumption on the incidence of MRSA and Clostridium difficile. J Antimicrob Chemother 2008; 62: 601 7. 6 Muller A, Mauny F, Talon D et al. Effect of individual- and group-level antibiotic exposure on MRSA isolation: a multilevel analysis. J Antimicrob Chemother 2006; 58: 878 81. 7 Dancer SJ. The effect of antibiotics on methicillin-resistant Staphylococcus aureus. J Antimicrob Chemother 2008; 61: 246 53. 8 Owens RC, Donskey CJ, Gaynes RP et al. Antimicrobial-associated risk factors for Clostridium difficile infection. Clin Infect Dis 2008; 46 Suppl 1: S19 31. 9 Cook PP, Catrou PG, Christie JD et al. Reduction in broad-spectrum antimicrobial use associated with no improvement in hospital antibiogram. J Antimicrob Chemother 2004; 53: 853 9. 10 Buising KL, Thursky KA, Robertson MB et al. Electronic antibiotic stewardship reduced consumption of broad-spectrum antibiotics using a computerized antimicrobial approval system in a hospital setting. J Antimicrob Chemother 2008; 62: 608 16. 11 Lesprit P, Duong T, Girou E et al. Impact of a computer-generated alert system prompting review of antibiotic use in hospitals. J Antimicrob Chemother 2009; 63: 1058 63. 12 Weber SG, Gold HS, Hooper DC et al. Fluoroquinolones and the risk for methicillin-resistant Staphylococcus aureus in hospitalized patients. Emerg Infect Dis 2003; 9: 1415 22. 13 Charbonneau P, Parienti J, Thibon P et al. Fluoroquinolone use and methicillin-resistant Staphylococcus aureus isolation rates in hospitalized patients: a quasi experimental study. Clin Infect Dis 2006; 42: 778 84. 14 Pepin J, Saheb N, Coulombe MA et al. Emergence of fluoroquinolones as the predominant risk factor for Clostridium difficile-associated diarrhea: a cohort study during an epidemic in Quebec. Clin Infect Dis 2005; 41: 1254 60. 15 Cook PP, Catrou P, Gooch M et al. Effect of reduction in ciprofloxacin use on prevalence of methicillin-resistant Staphylococcus aureus rates within individual units of a tertiary care hospital. J Hosp Infect 2006; 64: 348 51. 16 Muto CA, Pokrywka M, Shutt K et al. A large outbreak of Clostridium difficile-associated disease with an unexpected proportion of deaths and colectomies at a teaching hospital following increased fluoroquinolone use. Infect Control Hosp Epidemiol 2005; 26: 273 80. 17 Dial S, Alrasadi K, Manoukian C et al. Risk of Clostridium difficile diarrhea among hospital inpatients prescribed proton pump inhibitors: cohort and case-control studies. CMAJ 2004; 171: 33 8. 18 Bartlett JG. Antibiotic-associated diarrhea. N Engl J Med 2002; 346: 334 9. 19 Kyne L, Sougioultzis S, McFarland LV et al. Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea. Infect Control Hosp Epidemiol 2002; 23: 653 9. 20 Cunningham R, Dale B, Undy B et al. Proton pump inhibitors as a risk factor for Clostridium difficile diarrhoea. J Hosp Infect 2003; 54: 243 5. 21 Merrer J, Santoli F, Appere de Vecchi C et al. Colonization pressure and risk of acquisition of methicillin-resistant Staphylococcus aureus in a medical intensive care unit. Infect Control Hosp Epidemiol 2000; 21: 718 23. 22 Muto CA, Jernigan JA, Ostrowsky BE et al. SHEA guideline for preventing nosocomial transmission of multidrug-resistant strains of Staphylococcus aureus and Enterococcus. Infect Control Hosp Epidemiol 2003; 24: 362 86. 23 Kyne L, Hamel MB, Polavaram R et al. Health care costs and mortality associated with nosocomial diarrhea due to Clostridium difficile. Clin Infect Dis 2002; 34: 346 53. 24 O Brien JA, Lahue BJ, Caro JJ et al. The emerging infectious challenge of Clostridium difficile-associated disease in Massachusetts hospitals: clinical and economic consequences. Infect Control Hosp Epidemiol 2007; 28: 1219 27. 25 Dubberke ER, Reske KA, Olsen MA et al. Short and long term attributable cost of Clostridium difficile-associated disease in non-surgical patients. Clin Infect Dis 2008; 46: 497 50. 209