UNDERSTANDING RISK FACTORS FOR CLOSTRIDIUM DIFFICILE INFECTION ACROSS A FIVE HOSPITAL HEALTH SYSTEM IN THE BALTIMORE/WASHINGTON DC AREA

Statement of Problem – The objective of this study is to understand conditional risk factors for Clostridium difficile infection (CDI) over time and between hospitals Methods – Retrospective cohort of patients admitted to five adult hospitals in the Baltimore/Washington DC area in calendar year 2016 and 2017. International Statistical Classification of Diseases and Related Health Problems (ICD) 10 codes, the Premier’s CareScienceTM Mortality Risk Score, room and clinical data were obtained through Premier CareScience reports and EPIC electronic medical record reports. Days of therapy (DOT) for antibiotics, probiotic supplementation and proton pump inhibitor use (PPI) were cumulative through each day of hospitalization. Colonization pressure was calculated as days of exposure multiplied by the number of patients with CDI in a unit that month. A multivariable logistic regression was performed for each day of admission separately and for each hospital. Results – There were 1922 (1.0%) cases of CDI among 192,522 patients. The Premier’s CareScienceTM Mortality Risk Score expected mortality score was the greatest contributor to risk of CDI particularly in the first 5 days of hospitalization. Probiotic supplementation exposure was a significant risk factor in the first (OR 2.67) and third day (OR 1.46) of the hospital course. ICU exposure was protective with 37-33% lower odds of CDI on day 1 and 2 but by day 8 and 11 became a significant risk factor for CDI (OR 1.15, OR 1.12). Colonization pressure risk decreased over time, but remained significant for the first 8 days of admission (OR 1.19-OR 1.03). Carbapenem, penicillin, cephalosporin, quinolone and other antibiotic use was a risk factor for CDI however the conditional risk decreased over time. Clindamycin was protective in our cohort. The Premier’s CareScienceTM Mortality Risk Score was the greatest predictor for CDI in 4/5 hospitals (OR 6-30). Colonization pressure was a significant risk factor for CDI in three hospitals. Carbapenem use was a risk factor in two hospitals and cephalosporin use was only a risk factor in one hospital. Conclusions – Risk factors for CDI vary over time and by hospital and interventions to prevent CDI need to be tailored by hospital and time to be effective