Investigating the relationship between vancomycin-resistant Enterococcus control practices and the incidence of health care–associated Clostridioides difficile infections in Ontario

•Transmission modes are similar between C. difficile and VRE.•Discontinuation of VRE control practices was associated with a change in rate of health care-associated C difficile infection.•This study presents a potential unintended consequence associated with ending VRE control. We evaluated the impact of discontinuing vancomycin-resistant Enterococcus (VRE) screening and use of contact precautions on the incidence of health care–associated Clostridioides difficile infection (HA-CDI) in acute teaching hospitals in Ontario, Canada. Among hospitals that stopped VRE screening and contact precaution measures, there was a significant change in HA-CDI rates after the discontinuation of practices (incidence rate ratios, 1.11; 95% confidence interval, 1.01-1.22). No change in rate was observed among hospitals that continued VRE control practices. Screening and use of contact precautions for VRE may provide hospitals additional advantages for broadened HA-CDI control and prevention. We evaluated the impact of discontinuing vancomycin-resistant Enterococcus (VRE) screening and use of contact precautions on the incidence of health care–associated Clostridioides difficile infection (HA-CDI) in acute teaching hospitals in Ontario, Canada. Among hospitals that stopped VRE screening and contact precaution measures, there was a significant change in HA-CDI rates after the discontinuation of practices (incidence rate ratios, 1.11; 95% confidence interval, 1.01-1.22). No change in rate was observed among hospitals that continued VRE control practices. Screening and use of contact precautions for VRE may provide hospitals additional advantages for broadened HA-CDI control and prevention. The modes of transmission of Clostridioides difficile and vancomycin-resistant Enterococcus (VRE) in the health care setting are similar; patients become colonized or infected after exposure to fecally contaminated health care worker hands, shared patient equipment, or other environmental surfaces.1Ontario Agency for Health Protection and Promotion, Provincial Infectious Diseases Advisory CommitteeAnnex A–screening, testing and surveillance for antibiotic-resistant organisms (AROs). Annexed to: routine practices and additional precautions in all health care settings. Queen's Printer for Ontario, Toronto, ON2013Google Scholar, 2Gingras G Guertin MH Laprise JF Drolet M Brisson M Mathematical modeling of the transmission dynamics of Clostridium difficile infection and colonization in healthcare settings: a systematic review.PLoS One. 2016; 11e0163880Crossref PubMed Scopus (16) Google Scholar Accordingly, we hypothesized that hospitals that discontinued screening and use of contact precautions as part of their VRE control program may experience higher rates of health care–associated C difficile infection (HA-CDI) and outbreaks, as the inability to detect VRE transmission could delay the identification of infection control breaches. The objective was to determine the association between VRE screening and contact precaution measures and HA-CDI among acute teaching hospitals. We conducted an interrupted time series quasi-experiment to examine HA-CDI incidence among all acute teaching hospitals in Ontario, Canada that discontinued VRE screening and use of contact precautions for patients colonized or infected with VRE (ceased-screening cohort) between January 2011 and November 2018. Acute teaching hospitals that continued VRE screening and contact precautions throughout this period formed the control group (screening cohort). Data on hospital VRE control programs were obtained from annual surveys performed by Public Health Ontario.3Johnstone J Policarpio ME Lam F Adomako K Prematunge C Nadolny E et al.Rates of blood cultures positive for vancomycin-resistant Enterococcus in Ontario: a quasi-experimental study.CMAJ Open. 2017; 5: E273-E280Crossref Google Scholar Reported HA-CDI cases and outbreaks are defined as per the provincial case definitions set by the Ministry of Health and Long-Term Care.4Ministry of Health and Long-Term Care. Appendix B: provincial case definitions for disease of public health significance. Available from: http://www.health.gov.on.ca/en/pro/programs/publichealth/oph_standards/docs/cdi_cd.pdf. Accessed August 9, 2019.Google Scholar We analyzed monthly rates of HA-CDI attributable to the reporting facility, calculated as the number of HA-CDI cases divided by total patient days, using Ontario's mandatory patient safety reporting database. Reported HA-CDI outbreaks were extracted from the integrated Public Health Information System. Outcome data from January 2011 to November 2018 were obtained. All statistical analyses were performed on SAS version 9.3 (SAS Institute, Cary, NC). We compared rates of HA-CDI before and after the change in VRE control measures in the ceased-screening cohort; the month in which VRE screening practices ended was used to define the pre- and postperiods. For the screening cohort, we used a hypothetical cessation date of December 2014, representing the midpoint of the study period. We first examined the effects on HA-CDI cases. Using a multivariable generalized estimating equation negative binomial model, we computed incidence rate ratios (IRR) to test for differences in HA-CDI trends (ie, slope value) before and after the VRE control change in the ceased-screening and screening cohorts. The model was adjusted for seasonality effects.5Furuya-Kanamori L McKenzie SJ Yakob L Clark J Paterson DL Riley TV et al.Clostridium difficile infection seasonality: patterns across hemispheres and continents–a systematic review.PloS One. 2015; 10e0120730Crossref PubMed Scopus (25) Google Scholar We hypothesized that HA-CDI rates may not change immediately but over successive months; sensitivity analyses for lagged effects were explored at 3- and 6-months periods. Second, to examine the effect on HA-CDI outbreaks, we compared the average rate of outbreaks per month before and after the change in VRE control measures. IRRs were computed with multivariable generalized estimating equation negative binomial regression for ceased-screening and screening cohorts. All 24 acute teaching hospitals in Ontario were included in this study: 11 hospitals continued screening practices and 13 hospitals ended screening practices in 2012 (4 in June and 7 in July) and 2015 (1 in April and 1 in June). Hospital characteristics between ceased-screening and screening hospitals were comparable for the types of services available (ie, transplant programs, dialysis centers, stem cell transplant programs, pediatric services) and bed sizes (Table 1).Table 1Comparison of hospital characteristics by vancomycin-resistant Enterococcus screening statusHospital characteristicsCeased-screening hospitals, n = 13Screening hospitals, n = 11P valueTypes of hospital services.61 Transplant programs4 (31%)3 (27%) Dialysis centers8 (62%)6 (55%) Stem cell transplant Programs4 (31%)2 (18%) Pediatrics services1 (8%)3 (27%)Bed size.66 Less than 1001 (8%)1 (9%) 100 to 2493 (23%)1 (9%) 250 or more9 (69%)9 (82%) Open table in a new tab Between 2011 and 2018, trends in annual HA-CDI rates decreased from 5.68 to 3.44 cases per 10,000 patient days in the ceased-screening cohort, and similarly from 4.37 to 3.82 cases per 10,000 patient days in the screening cohort. Among the ceased-screening cohort, there was a significant 11% increase in the slope value of HA-CDI rates after the change in VRE control programs (adjusted IRR, 1.11; 95% confidence interval [CI], 1.01-1.22). Among the screening cohort, no significant change was observed in HA-CDI rates after a hypothetical screening cessation (adjusted IRR, 1.03; 95% CI, 0.81-1.13). Regarding lagged effects, at 3- and 6-months periods, there was a significant increase in the slope of HA-CDI rates after the change in VRE control programs, which was not seen in the screening cohort (Table 2).Table 2Annual IRR of HA-CDI rates before and after vancomycin-resistant Enterococcus screening cessationAdjusted IRR*IRR adjusted for hospital clustering effects and seasonality of C difficile infections.Slope beforeSlope afterSlope change (95% CI)Slope change P valueCeased-screening cohort (n = 13) No lag0.840.931.11 (1.01-1.22).03 3 months lag0.840.941.11 (1.01-1.22).03 6 months lag0.840.931.11 (1.01-1.22).03Screening cohort (n = 11) No lag0.961.001.03 (0.94-1.13).49 3 months lag0.961.001.04 (0.95-1.14).43 6 months lag0.960.991.03 (0.94-1.12).57CI, confidence interval; HA-CDI, health care–associated Clostridioides difficile infections; IRR, incidence rate ratio. IRR adjusted for hospital clustering effects and seasonality of C difficile infections. Open table in a new tab CI, confidence interval; HA-CDI, health care–associated Clostridioides difficile infections; IRR, incidence rate ratio. During the study period, there were 68 HA-CDI outbreaks reported: 43 (63%) in the ceased-screening cohort and 25 (37%) in the screening cohort. No significant changes in HA-CDI outbreaks were observed after the discontinuation of VRE control practices in both the ceased-screening cohort (IRR, 0.82; 95% CI, 0.32-2.09) and the screening cohort (IRR, 1.26; 95% CI, 0.57-2.85). In this study, although overall trends of HA-CDI rates showed a decline over time, we observed an attenuated decline in HA-CDI rates that was significantly associated with the discontinuation of targeted VRE control programs among acute teaching hospitals. Conversely, no change was observed in HA-CDI rates in hospitals that continued to screen and use contact precautions for VRE. These results support the hypothesis that the absence of VRE screening programs may indirectly influence higher than expected HA-CDI rates because of the inability to detect VRE transmission and delayed identification of infection control breaches, although these effects were masked by an overall declining trend in HA-CDI rates in our study. Therefore, VRE can be thought of as a so-called “indicator organism” for infection control lapses. In a study by our group, the discontinuation of VRE control practices has been associated with higher VRE bloodstream infection rates in the same sample of hospitals.3Johnstone J Policarpio ME Lam F Adomako K Prematunge C Nadolny E et al.Rates of blood cultures positive for vancomycin-resistant Enterococcus in Ontario: a quasi-experimental study.CMAJ Open. 2017; 5: E273-E280Crossref Google Scholar Given the similar modes of transmission between VRE and C difficile, this study suggests there may be unintended infection control consequences that extend beyond just VRE when hospitals cease screening and contact precautions for VRE. Our study had some limitations that may have influenced the results. Because of the observational nature of the study, we could not control for extraneous factors that may influence HA-CDI rates. First, routine and discharge cleaning practices have been shown to be effective in reducing VRE infections; however, data on hospital adherence to these practices and other VRE control measures (ie, antimicrobial stewardship programs) were not available.6Mitchell BG Hall L White N Barnett AG Halton K Paterson DL et al.An environmental cleaning bundle and health-care-associated infections in hospitals (REACH): a multicentre, randomised trial.Lancet Infect Dis. 2019; 19: 410-418Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar Additionally, limitations related to the small sample size of 24 hospitals may explain why no change was detected in the outbreak analysis, despite seeing a difference in HA-CDI rates in the ceased-screening cohort. Next, we observed an overall declining rate of HA-CDI during the study period; this may be attributed to the introduction of mandated patient safety indicator reporting or evolving epidemiology from virulent NAP1 to other strain types such as NAP4 and NAP11.7Daneman N Stukel TA Ma X Vermeulen M Guttmann A Reduction in Clostridium difficile infection rates after mandatory hospital public reporting: findings from a longitudinal cohort study in Canada.PLoS Med. 2012; 9e1001268Crossref PubMed Scopus (43) Google Scholar, 8Katz KC Golding GR Choi KB Pelude L Amaratunga KR Taljaard M et al.The evolving epidemiology of Clostridium difficile infection in Canadian hospitals during a postepidemic period (2009-2015).CMAJ. 2018; 190: E758-E765Crossref PubMed Scopus (26) Google Scholar Notably, we adapted the period of analysis to circumvent potential effects from changes to the case definitions and in laboratory testing methods (increasing sensitivity) in 2010, although some changes may have occurred during the study period.9Leber AL Hindler JF Kato EO Bruckner DA Pegues DA Laboratory-based surveillance for vancomycin-resistant enterococci: utility of screening stool specimens submitted for Clostridium difficile toxin assay.Infect Control Hosp Epidemiol. 2001; 22: 160-164Crossref PubMed Scopus (19) Google Scholar, 10Public Health Ontario LaboratoryClostridium difficile diagnostic testing-update.2016https://www.publichealthontario.ca/en/eRepository/LAB-SD-002-004_C_difficile_Final_V4.pdfGoogle Scholar Finally, the generalizability of the findings are limited to only acute teaching hospitals, as we were unable to include community hospitals due to low incidence of HA-CDI and low number of community hospitals that discontinued VRE screening and use of contact precautions. Our study found a significant association between discontinued targeted VRE control programs and a change in rates of HA-CDI. Because of the shared modes of transmission between VRE and C difficile, this study suggests that there may be additional benefits for VRE screening and contact precaution measures on HA-CDI control and prevention.