Abstract
Background
Vancomycin-resistant Enterococcus (VRE) is a leading cause of healthcare-associated infections, and asymptomatic colonization precedes infection. VRE continues to spread despite widespread application of pathogen-specific control guidelines. A better understanding of the risk factors for transmission is needed.
Methods
A retrospective matched case-control study was performed from June 2013 through December 2016 in a single institution. Patients in 6 intensive care units, 1 hematology and oncology unit, and 1 bone marrow transplant unit were screened by means of rectal swab sampling on admission and weekly thereafter. Case patients had a negative swab sample followed by a positive sample >3 days after admission. Controls were closely matched to case patients based on time from admission to the second swab sample, unit in which the second sample was obtained, and date of admission. Comorbidity data, procedures, healthcare settings and exposures, culture data, and duration of antibiotic and proton pump inhibitor (PPI) therapy were abstracted from the electronic medical record. A multivariable risk factor model for conversion was generated using purposeful selection.
Results
A total of 551 case patients were matched with controls. The largest modifiable effects on VRE acquisition were ≥1 day of vancomycin therapy (odd ratio, 1.98; P < .001), ≥1 day of aerobic antibiotic therapy (1.90; P < .001), and a dose-dependent effect of PPI therapy (odds ratio per day of therapy, 1.09; P < .001). Colonization pressures from patients identified to be carriers and placed in contact precautions did not confer increased risk.
Conclusions
Decreasing PPI use and preventing the inappropriate initiation of antibiotic therapy are modifiable targets to decrease VRE transmission in the hospital.
Keywords: hospital acquisition, proton pump inhibitor, risk factors, vancomycin resistant Enterococcus
Decreasing proton pump inhibitor use and preventing inappropriate antibiotic initiation are modifiable targets to decrease in-hospital vancomycin-resistant Enterococcus (VRE) transmission. Colonization pressures from patients identified as carriers and placed in contact precautions does not increase the odds of VRE acquisition.
Vancomycin-resistant Enterococcus (VRE) is a leading cause of nosocomial infections and results in higher morbidity and mortality rates than vancomycin-susceptible enterococci [1, 2]. First identified in 1988, VRE rapidly spread and now comprises >25% of all enterococcal bloodstream infections in the United States [3, 4]. It is a hospital-acquired pathogen, and the significant reservoirs for transmission are acute and long-term healthcare facilities [5, 6]. Molecular epidemiologic analyses have shown that cross-transmission between hospitalized patients is the primary means of VRE spread [7].
Colonization with VRE generally precedes infection [8, 9]. There may be 10 times more colonized than infected patients [10]. Risk factors for VRE acquisition include colonization pressure, a measure of the exposure to patients colonized with VRE during a given period [11–13], host debility and comorbid conditions [7, 14], and antibiotic exposure. In particular, vancomycin, cephalosporins, and drugs with activity against anaerobes have all been shown to increase the risk for VRE acquisition [7, 14–16].
In the past 20 years, there has been a substantial effort to minimize the spread of VRE by targeting known risk factors for VRE acquisition [17, 18]. Despite the application of these extensive control measures, VRE remains endemic in many hospitals [19], emphasizing the need to identify new modifiable risk factors for VRE acquisition. It is unclear whether the previously identified risk factors remain important in the setting of current enhanced infection prevention and antibiotic stewardship practices. The current study examined whether colonization pressure remains a significant risk factor for VRE spread and reassessed modifiable risk factors.
METHODS
Study Setting and Design
A retrospective case-control investigation of risk factors was conducted at the University of Michigan Healthcare system from 1 June 2013 through 31 December 2016. The University of Michigan Healthcare system consists of approximately 1000 inpatient beds and serves as a tertiary referral hospital for southeastern Michigan. The institutional review board at the University of Michigan approved the study before its initiation.
Patients and Variables
The infection control practice throughout the study period was to perform routine surveillance for VRE on 8 adult units, including intensive care units, the hematology and oncology ward, and the bone marrow transplant ward. All patients were routinely screened on admission and weekly thereafter with rectal swab samples that were then applied to Bio-Rad VRESelect chromogenic medium, used to detect VRE. Case patients were those with an initial negative swab sample followed by a positive sample. Additional requirements were that the positive sample was collected 3 days after admission, and that case patients had no prior isolation of VRE from any previous screening swab samples or sterile site cultures. Thus, each case patient was unique, but the patient’s first admission was not necessarily the admission that was used.
Controls were patients with an initial negative swab sample, followed by ≥1 negative samples. Their results remained negative on serial VRE screening and clinical culturing, with no isolation of VRE in any culture before inclusion in the study. Case patients were matched to controls at a 1:1 ratio. Each control was used only once. Matching parameters were (1) time at risk (within ±5%). (2) the unit from which the first positive VRE was recovered for case patients or the matched index swab samples for controls, and (3) calendar year (±365 days). Time at risk was defined for case patients as time elapsed between admission and positive VRE screening result. For controls, it was the time between admission and the index swab sample within the same time frame (±5% of the time at risk for the case patient).
Parameters retrieved from patient records included (1) demographics variables, (2) underlying conditions and comorbid conditions, (3) recent healthcare-associated exposures within 90 days before admission, (4) invasive procedures during the time at risk, (5) indwelling devices inserted or present during the time at risk, (6) isolation of Clostridium difficile during the time at risk, and (7) days of therapy for individual antimicrobials and proton pump inhibitor (PPI) during the time at risk.
Colonization pressure was analyzed with 2 different metrics. The first was defined as the number of patient-days of exposure for patients known to be VRE positive. For patient i in the study, this was calculated as the sum Σjtij over all j other patients in the hospital, where tij was the time, in days, that patients i and j were in the same unit at the same time, and for which patient i was not yet VRE positive and patient j was VRE positive.
A second measure of colonization pressure, average prevalence during time at risk, was also calculated to compare our results with those of previous studies. Colonization pressure as average prevalence, for an individual patient i was calculated as Σjtij/Σjcij , where tij is as above and cij is the time, in days, that patients i and j were in the same unit at the same time, and for which patient i was not yet VRE positive but patient j was either positive or negative.
Statistical Analysis
Bivariable analysis was performed with conditional logistic regression, using the clogit package in R software (version 3.4.1) [20]. A conditional multivariable logistic regression model was built using purposeful selection in SAS Studio online, version 3.71 (Enterprise Edition). Purposeful selection is an iterative process of variable selection, in which covariates are removed from the model if they are nonsignificant and not confounders [21–23]. Significance was evaluated at the 0.1 α level, and confounding was defined as a change in any remaining parameter estimate >20%, compared with the full model.
Variables not selected for the original multivariable model were then added back one at a time, to identify those that made an important contribution in the presence of other variables. Any variables that were significant at the 0.1 α level were put back in the model, and the model was iteratively reduced as before, but only for the variables that were additionally added. Multiple correlations were calculated in R software using the lm package. In multiple correlation figures, squares are scaled based on significance (P < .05; false discovery rate corrected).
In multivariable analysis, antibiotic use was entered as a different type of dichotomous variables in 4 separate models: (1) as individual antibiotics, (2) by antibiotic class, (3) by categorizing antibiotics into either aerobic or anaerobic or vancomycin, and (4) by classifying antibiotics as either vancomycin or any other antibiotic. Vancomycin was considered independently of other classes regardless of categorization, owing to its unique importance to VRE. Colonization pressure was defined according the duration of exposure, in patient-days. The criteria for choosing the final model were the highest Cox and Snell R2 value, the lowest Akaike information criterion, and the lowest correlation between variables. Based on these criteria, categorizing antibiotics by use of aerobic or anaerobic or vancomycin was chosen for the final model.
RESULTS
There were 22 572 patients with ≥1 screening swab sample over the 3.5-year study period, and 688 with an initial negative sample followed by a positive sample were identified as potential case patients. There were 6286 patients (with 7174 admissions) who could have served as controls. Of the potential case patients, 90 were excluded because they had prior isolation of VRE from sterile site cultures, 26 converted from VRE negative to positive in <72 hours, and 21 had no controls that met matching criteria. Thus, 551 case patients were included, and 551 matched controls were identified. The distribution of time to VRE conversion among case patients is displayed in Figure 1.
Figure 1.
Distribution of time at risk for case patients and controls, with time at risk defined as the time from admission until time to positive vancomycin-resistant Enterococcus screening results for case patients, and time from admission to index swab sample for controls.
Full results of the bivariable logistic regression are included in the Supplementary Materials (Supplementary Table 1), with the notable results shown in Table 1. There were no significant differences in age, sex, race, or ethnicity between case patients and controls.
Table 1.
Notable Findings of Univariable Analysis and Complete Final Multivariable Model
| Variable | Patients, No. (%)a | OR (95% CI) | ||
|---|---|---|---|---|
| Control (n = 551) | Case Patients (n = 551) | Univariable Analysis |
Multivariable Analysisb | |
| Age, mean, y | 58 | 59 | 1.00 (.99–1.01) | … |
| Nonwhite race | 107 (19) | 96 (17) | 0.87 (.65–1.20) | … |
| Time at risk, mean, d | 14.9 | 14.9 | 1.25 (.93–1.68) | … |
| Colonization pressure, mean, patient-days of exposure | 46 | 48 | 1.00 (.99–1.01) | … |
| Total duration of antibiotic treatment, mean, d | 19.3 | 20.4 | 1.00 (.99–1.01) | … |
| Any antibiotic use | 495 (89) | 536 (97) | 4.72 (2.47–9.05) | … |
| Any PPI use | 394 (72) | 445 (81) | 1.75 (1.30–2.36) | … |
| Final multivariable model | ||||
| Vancomycin | 358 (65) | 421 (76) | 1.94 (1.44–2.60) | 1.98 (1.39–2.82) |
| Aerobic antibiotic | 433 (79) | 493 (89) | 2.66 (1.81–3.91) | 1.90 (1.21–2.93) |
| Duration of PPI use, mean, d | 8.6 | 10.5 | 1.06 (1.03–1.08) | 1.09 (1.06–1.13) |
| Total parenteral nutrition | 173 (31) | 201 (36) | 1.26 (.98–1.62) | 1.56 (1.16–2.09) |
| Congestive heart failure | 79 (14) | 105 (19) | 1.46 (1.04–2.05) | 1.54 (1.04–2.28) |
| Pulmonary hypertension | 53 (10) | 75 (14) | 4.72 (2.47–9.05) | 3.50 (1.54–7.95) |
| Chronic renal failure | 76 (14) | 93 (17) | 1.27 (.91–1.77) | 1.79 (1.15–2.78) |
| Hypertension | 45 (8) | 32 (6) | 0.68 (.42–1.10) | 0.32 (.17–.62) |
| Hemiplegia | 64 (12) | 85 (15) | 1.40 (.98–2.00) | 2.54 (1.40–4.68) |
| Vascular surgery | 48 (9) | 36 (7) | 0.71 (.44–1.14) | 0.57 (.34–.98) |
| Central catheter placement | 395 (72) | 348 (63) | 0.66 (.51–.86) | 0.99 (.61–1.38) |
| Transfer outside hospital | 279 (51) | 247 (45) | 0.77 (.61–.99) | 1.93 (1.15–3.17) |
| Male sex | 325 (59) | 307 (56) | 0.86 (.68–1.11) | 1.09 (.75–1.60) |
| Interaction terms | ||||
| Pulmonary hypertension and vancomycin | … | … | … | 0.27 (.10–.71) |
| Duration of PPI use and sex | … | … | … | 0.96 (.93–.98) |
| Central catheter and transfer outside hospital | … | … | … | 0.36 (.20–.64) |
| Transfer outside hospital and hemiplegia | … | … | … | 0.41 (.18–.94) |
Abbreviations: CI, confidence interval; OR, odds ratio; PPI, proton pump inhibitor.
aData represent no. (%) of patients unless otherwise specified.
bResults shown for the variables included in final multivariable analysis. For full analysis of all factors displayed, see Supplementary Table 1. Values were as follows for the multivariable analysis: pseudo R2, 0.11 (maximum, 0.5); concordance, 0.66; and Akaike information criterion, 666.3.
In the bivariable analysis, antibiotic use among all patients was widespread, with 495 of 551 controls (90%) and 536 of 551 case patients (97%) receiving a dose of antibiotic, a significant difference (P < .01; Table 1). There was no significant difference between case patients and controls in the duration of therapy, measured in days, for any individual antibiotic (Supplementary Table 1) or for the total duration of antibiotic treatment during the time at risk (mean, 20.4 days for case patients vs 19.4 days for controls; P = .15) (Table 1 and Figure 2). PPI use was more common in case patients than in controls, when analyzed as a dichotomous variable; 394 of 551 controls (72%) and 445 of 551 case patients (81%) received ≥1 day of PPI therapy during the time at risk (P < .01) (see Table 1). In contrast to antibiotic use, case patients had a longer duration of PPI therapy (mean, 10.5 days for case patients vs 8.5 days for controls; P < .01).
Figure 2.
Total duration of antibiotic therapy. The mean was 19.3 days for controls and 20.4 days for case patients (P = .14 in bivariable analysis).
In the final multivariable model (Table 2), no candidate covariates were found to be confounders that changed the magnitude of the ratio coefficients by >20%. Notable modifiable risk factors that achieved significance were use of vancomycin (odds ratio, 1.98; 95% confidence interval, 1.39–2.82) or an aerobic antibiotic (1.90; 1.21–2.93]), total parenteral nutrition (1.56; 1.16–2.09), and duration of PPI therapy. Each day of PPI therapy increased the risk of VRE acquisition by 9% (odds ratio, 1.09; 1.06–1.13]). There was no significant difference in colonization pressure in bivariable analysis (Figure 3), and colonization pressure did not achieve significance in multivariable analysis and was removed from the final multivariable model. We noted that removing interaction terms included in the final multivariable model produced no qualitative difference in significant individual risk factors for VRE acquisition.
Figure 3.
Colonization pressure measured as patient-days of exposure. The mean was 46 patient-days for controls and 48 patient-days for case patients (P = .08 in bivariable analysis).
DISCUSSION
This study is the first to demonstrate the use of PPI therapy as a significant risk factor for VRE acquisition. It also showed that a single dose of antibiotic predisposes patients to VRE acquisition, but the duration of antibiotic treatment is not a significant risk factor. In contrast to previous findings, colonization pressure was not a significant risk factor for individual patients.
PPI therapy significantly increased the odds of acquiring VRE, and the effect was dose dependent. In the multivariable analysis, the odds of VRE acquisition for patients with prior PPI exposure is 9% more than that in a similar patient without exposure for each day of PPI treatment. This finding has biologic plausibility. The stomach acid barrier kills the vegetative forms of several enteric pathogens [24] and plays a vital role as a barrier to ingested bacteria. A mouse model of VRE acquisition showed that gastric acid suppression by PPI treatment facilitates colonization [25]. PPIs are often overprescribed without a clear clinical indication [26]. The findings of the current study suggest that reducing inappropriate PPI use could be a novel infection control strategy to limit the spread of VRE. A similar association has been proposed with C. difficile infection, but the results have been conflicting [27, 28]. The conflicting evidence in C. difficile infection highlights the need for additional studies on the association between PPI use and VRE acquisition. The final model had a significant interaction term between PPI use and sex, because the effect of PPI therapy was diminished in male patients. More study of the interaction between sex, PPI therapy, and the risk of VRE acquisition is needed.
Vancomycin or aerobic antibiotic therapy given for ≥1 day increased the odds of acquisition, but the duration of therapy had no impact. Antimicrobial stewardship programs have recently emphasized deescalation after antibiotic treatment is already initiated [29]. This study suggests that efforts to contain VRE acquisition should focus on unnecessary initiation of antibiotics rather than deescalation. Previous studies found that the use of anaerobic antibiotics is a risk factor for VRE colonization [30–32], but the current study did not replicate this finding. This discrepancy may be related to the high degree of overlap between vancomycin, anaerobic, and aerobic antibiotic use in the current study. Only a few patients received anaerobic antibiotics alone. Aerobic antibiotics and vancomycin accounted for the majority of antibiotic use.
In contrast to previous studies, colonization pressure was not identified as a risk factor in either the bivariable or the multivariable analysis. Our data suggest that 20 years after widespread implementation of infection control practices targeted at VRE acquisition, individual patient risk factors play a larger role than environmental exposure to VRE. This suggests that future interventions directed at controlling the acquisition and spread of VRE should target the individual patient and not the hospital population as a whole—for instance, emphasizing the appropriate use of antibiotics and PPI therapy for an individual patient. We do note that prior studies of VRE acquisition had higher overall colonization pressure. The mean point prevalence of VRE in the current study was 17%, compared with a range of 25%–38% in other studies [11, 12, 33–37]. Our study was conducted in a different era of infection control for the VRE epidemic, an era with more widespread adoption of infection control measures. Future study of the impact of infection prevention strategies, comparing strategies aimed at the population level and individual level, may be helpful.
Our final multivariable model contained several significant interaction terms. These terms were included to account for complex covariation in observational data, although some were not readily interpretable. A sensitivity analysis, in which the interaction terms included in the final multivariable model were each individually removed, produced no qualitative difference in the effect of antibiotic use, PPI use, and colonization pressure on VRE acquisition found in our model.
The weaknesses of the current study include its retrospective, single-center design and lack of access to physiologic data to categorize the severity of illness. A major strength of the study is a data-driven, purposeful selection approach to model building. This algorithmic approach reduced selection bias for known risk factors and allowed for discovery of new findings. Other strengths of the study are its large sample size, larger than any previous single-center case-control study of VRE acquisition [7, 10, 38], as well as its rigorous matching protocol. It is also among the few studies performed in the context of the modern aggressive infection control protocols targeted at VRE.
It has been >2 decades since the first studies of VRE colonization, and in that time infection control practices have changed dramatically. In this setting, colonization pressure was not a risk factor for VRE acquisition for the individual patient. PPI use is newly identified as a modifiable target to decrease VRE transmission in the hospital. Our findings also suggest that stewardship efforts directed at stopping the spread of VRE should focus on avoiding the inappropriate initiation of antibiotics rather than on antibiotic deescalation.
Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Acknowledgments
Financial support.This is work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (grant K08 AI119182 to R. J. W.) and by the National Science Foundation Graduate Research Fellowship (graduate research fellowship DGE-1256260 to J. A. M.).
Potential conflicts of interest.All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1. DiazGranados CA, Zimmer SM, Klein M, Jernigan JA. Comparison of mortality associated with vancomycin-resistant and vancomycin-susceptible enterococcal bloodstream infections: a meta-analysis. Clin Infect Dis 2005; 41:327–33. [DOI] [PubMed] [Google Scholar]
- 2. Edmond MB, Ober JF, Dawson JD, et al. Vancomycin-resistant enterococcal bacteremia: natural history and attributable mortality. Clin Infect Dis 1996; 23:1234–9. [DOI] [PubMed] [Google Scholar]
- 3. Deshpande LM, Fritsche TR, Moet GJ, et al. Antimicrobial resistance and molecular epidemiology of vancomycin-resistant enterococci from North America and Europe: a report from the SENTRY antimicrobial surveillance program. Diagn Microbiol Infect Dis 2007; 58:163–70. [DOI] [PubMed] [Google Scholar]
- 4. Reik R, Tenover FC, Klein E, McDonald LC. The burden of vancomycin-resistant enterococcal infections in US hospitals, 2003 to 2004. Diagnostic Microb Infect Dis 2004; 62:81–5. [DOI] [PubMed] [Google Scholar]
- 5. Hayden MK. Insights into the epidemiology and control of infection with vancomycin-resistant enterococci. Clin Infect Dis 2000; 31:1058–65. [DOI] [PubMed] [Google Scholar]
- 6. Chavers LS, Moser SA, Benjamin WH, et al. Vancomycin-resistant enterococci: 15 years and counting. J Hosp Infect 2003; 53:159–71. [DOI] [PubMed] [Google Scholar]
- 7. Bonten MJ, Hayden MK, Nathan C, et al. Epidemiology of colonisation of patients and environment with vancomycin-resistant enterococci. Lancet 1996; 348:1615–9. [DOI] [PubMed] [Google Scholar]
- 8. Alevizakos M, Gaitanidis A, Nasioudis D, et al. Colonization with vancomycin-resistant enterococci and risk for bloodstream infection among patients with malignancy: a systematic review and meta-analysis. Open Forum Infect Dis 2016; 4:ofw246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Zirakzadeh A, Gastineau DA, Mandrekar JN, et al. Vancomycin-resistant enterococcal colonization appears associated with increased mortality among allogeneic hematopoietic stem cell transplant recipients. Bone Marrow Transplant 2008; 41:385–92. [DOI] [PubMed] [Google Scholar]
- 10. Slaughter S, Hayden MK, Nathan C, et al. A comparison of the effect of universal use of gloves and gowns with that of glove use alone on acquisition of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med 1996; 125:448–56. [DOI] [PubMed] [Google Scholar]
- 11. Bonten MJM, Slaughter S, Ambergen AW, Hayden MK. The role of “colonization pressure” in the spread of vancomycin-resistant enterococci. Arch Intern Med 2017; 158:1127–32. [DOI] [PubMed] [Google Scholar]
- 12. Drees M, Snydman DR, Schmid CH, et al. Prior environmental contamination increases the risk of acquisition of vancomycin-resistant enterococci. Clin Infect Dis 2008; 46:678–85. [DOI] [PubMed] [Google Scholar]
- 13. Puzniak LA, Leet T, Mayfield J, et al. To gown or not to gown: the effect on acquisition of vancomycin-resistant enterococci. Clin Infect Dis 2002; 35:18–25. [DOI] [PubMed] [Google Scholar]
- 14. Safdar N, Maki DG. The commonality of risk factors for nosocomial colonization and infection with antimicrobial-resistant Staphylococcus aureus, Enterococcus, gram-negative bacilli, Clostridium difficile, and Candida. Ann Intern Med 2002; 136:834–44. [DOI] [PubMed] [Google Scholar]
- 15. Beezhold DW, Slaughter S, Hayden MK, et al. Skin colonization with vancomycin-resistant enterococci among hospitalized patients with bacteremia. Clin Infect Dis 1997; 24:704–6. [DOI] [PubMed] [Google Scholar]
- 16. Tornieporth NG, Roberts RB, John J, et al. 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. [DOI] [PubMed] [Google Scholar]
- 17. Centers for Disease Control and Prevention. Antibiotic resistance threats in the United States, 2013. Atlanta, GA: Centers for Disease Control and Prevention, 2013. [Google Scholar]
- 18. Fishman N; Infectious Diseases Society of America, Pediatric Infectious Diseases Society Policy statement on antimicrobial stewardship by the Society for Healthcare Epidemiology of America (SHEA), the Infectious Diseases Society of America (IDSA), and the Pediatric Infectious Diseases Society (PIDS). Infect Control Hosp Epidemiol 2012; 33:322–7. [DOI] [PubMed] [Google Scholar]
- 19. Sievert DM, Ricks P, Edwards JR, et al. Antimicrobial-resistant pathogens associated with healthcare-associated infections summary of data reported to the National Healthcare Safety Network at the Centers for Disease Control and Prevention, 2009–2010. Infect Control Hosp Epidemiol 2013; 34:1–14. [DOI] [PubMed] [Google Scholar]
- 20. R Development Core Team R. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2011. [Google Scholar]
- 21. Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med 2008; 3:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Chen Y, Millar JA. Machine learning techniques in cancer prognostic modeling and performance assessment. In: Frontiers of biostatistical methods and applications in clinical oncology. Singapore: Springer Singapore, 2017:193–230. [Google Scholar]
- 23. Hosmer DW, Lemeshow S.. Applied logistic regression. Hoboken, NJ: Wiley, 2000. [Google Scholar]
- 24. Rao A, Jump RLP, Pultz NJ, et al. In vitro killing of nosocomial pathogens by acid and acidified nitrite. Antimicrob Agents Chemother 2006; 50:3901–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Stiefel U, Rao A, Pultz MJ, et al. Suppression of gastric acid production by proton pump inhibitor treatment facilitates colonization of the large intestine by vancomycin-resistant Enterococcus spp. and Klebsiella pneumoniae in clindamycin-treated mice. Antimicrob Agents Chemother 2006; 50:3905–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kelly OB, Dillane C, Patchett SE, et al. The inappropriate prescription of oral proton pump inhibitors in the hospital setting: a prospective cross-sectional study. Dig Dis Sci 2015; 60:2280–6. [DOI] [PubMed] [Google Scholar]
- 27. Deshpande A, Pasupuleti V, Thota P, et al. Risk factors for recurrent Clostridium difficile infection: a systematic review and meta-analysis. Infect Control Hosp Epidemiol 2015; 36:452–60. [DOI] [PubMed] [Google Scholar]
- 28. Arriola V, Tischendorf J, Musuuza J, et al. Assessing the risk of hospital-acquired Clostridium difficile infection with proton pump inhibitor use: a meta-analysis. Infect Control Hosp Epidemiol 2016; 37:1408–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Silva BN, Andriolo RB, Atallah ÁN, Salomão R. De-escalation of antimicrobial treatment for adults with sepsis, severe sepsis or septic shock. Cochrane Database Syst Rev 2013. doi:10.1002/14651858.CD007934.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Donskey CJ, Chowdhry TK, Hecker MT, et al. Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N Engl J Med 2000; 343:1925–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Paterson DL, Muto CA, Ndirangu M, et al. Acquisition of rectal colonization by vancomycin-resistant Enterococcus among intensive care unit patients treated with piperacillin-tazobactam versus those receiving cefepime-containing antibiotic regimens. Antimicrob Agents Chemother 2008; 52:465–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Stiefel U, Pultz NJ, Helfand MS, Donskey CJ. Increased susceptibility to vancomycin-resistant Enterococcus intestinal colonization persists after completion of anti-anaerobic antibiotic treatment in mice. Infect Control Hosp Epidemiol 2004; 25:373–9. [DOI] [PubMed] [Google Scholar]
- 33. Puzniak LA, Mayfield J, Leet T, et al. Acquisition of vancomycin-resistant enterococci during scheduled antimicrobial rotation in an intensive care unit. Clin Infect Dis 2001; 33:151–7. [DOI] [PubMed] [Google Scholar]
- 34. Puzniak LA, Leet T, Mayfield J, et al. To gown or not to gown: the effect on acquisition of vancomycin-resistant enterococci. Clin Infect Dis 2002; 35:18–25. [DOI] [PubMed] [Google Scholar]
- 35. Winston LG, Charlebois ED, Pang S, et al. Impact of a formulary switch from ticarcillin-clavulanate to piperacillin-tazobactam on colonization with vancomycin-resistant enterococci. Am J Infect Control 2004; 32:462–9. [DOI] [PubMed] [Google Scholar]
- 36. Martínez JA, Ruthazer R, Hansjosten K, et al. Role of environmental contamination as a risk factor for acquisition of vancomycin-resistant enterococci in patients treated in a medical intensive care unit. Arch Intern Med 2003; 163:1905–12. [DOI] [PubMed] [Google Scholar]
- 37. Michigan Medicine Infection Prevention. Michigan medicine hand hygiene compliance data 2018. Available at: http://www.med.umich.edu/i/ice/data/hand_hygiene_data.html. Accessed 12 December 2018.
- 38. Rao GG, Ojo F, Kolokithas D. Vancomycin-resistant gram-positive cocci: risk factors for faecal carriage. J Hosp Infect 1997; 35:63–9. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.