2369. Prescribing Choices and C. difficile Infection Risk: A Longitudinal Cohort Study of Nursing Home Residents in Ontario, Canada

Abstract

BackgroundAntibiotics are the primary modifiable risk factor for C. difficile infection (CDI). However, the comparative risks of different prescribing choices are not known. Our objective was to quantify the benefits of: (1) preventing antibiotic initiation, (2) substituting high-risk antibiotics for low-risk antibiotics, (3) reducing antibiotic duration.MethodsWe conducted a cohort study of residents of over 600 Ontario nursing homes in the 2014 to 2017 period. All non-acute care hospital antibiotic exposures were identified from the Ontario Drug Benefit database, while CDI was identified using outpatient billings and/or hospital ICD-10 codes. Logistic regression models were used to examine the risk of CDI as a function of time-varying antibiotic exposures, while controlling for 14 different risk factors. Based on the models, we estimated the comparative risks of specific antibiotic regimens.ResultsWe identified 1,944 cases of CDI, for an incidence of 1.60 per 100,000 person-days. The 90-day risk among residents without antibiotics was 0.84 per 1,000 residents (‰), compared with 1.85‰ in those with a 7-day course of antibiotics. Preventing a 7-day course reduced CDI risk by 45% (see table, adjusted relative risk [RR] = 0.55, 95% confidence interval [CI] = 0.50, 0.60). The antibiotics conferring the highest risks were clindamycin at 4.1‰, moxifloxacin at 3.2‰, and amoxiclav at 3.0‰. Comparing 7-day courses of antibiotics with similar indications: nitrofurantoin engendered 37% less risk than ciprofloxacin, amoxicillin resulted in 31% less risk than amoxicillin-clavulanate, and cephalexin had 51% less risk than clindamycin. Reduced antibiotic durations were associated with less C. difficile risk. Compared with a 10-day course, a 7-day course was associated with 12% less risk, while a 5-day course was associated with 21% less risk.ConclusionWe have quantified, using a real-world population-based cohort of nursing home residents, the reduction in CDI risk incurred by preventing unnecessary antibiotic initiations, preferring low-risk agents over high-risk agents, and reducing duration. These figures will help clinicians compare risks and benefits of different prescribing choices with regards to CDI prevention. Disclosures All authors: No reported disclosures.