2430. Comorbidity and Severity of Illness Risk Adjustment for Hospital-Onset Clostridioides difficile Infection

Abstract

BackgroundHospital-onset C. difficile infection (HO-CDI) rates are publicly reported. However, patient-level risk factors are not included in the current risk adjustment methodology, and the knowledge as to which risk factors to include is incomplete. This study aimed to determine whether electronically-available comorbidities and laboratory indicators of severity of illness are risk factors for HO-CDI.MethodsWe performed a retrospective cohort study of all adult patients admitted to three hospitals (one academic, two community) in Baltimore, Maryland between January 1, 2016 and January 1, 2018. Information extracted from electronic medical records included demographics, ICD-10 codes, laboratory results within 24 hours of admission (i.e., hematocrit, hemoglobin, platelet count, leukocytes, BUN, CO2, creatinine, glucose, sodium, and potassium), medication administration (i.e., antibiotic and antacid use), and C. difficile test result. Comorbid conditions were assessed by the Elixhauser Comorbidity Index components. HO-CDI was defined by positive laboratory test > 3 days after admission. Potential risk factors for HO-CDI were assessed using bivariate log binomial regression. Multivariable log binomial regression was conducted using significant (P < 0.1) covariates.ResultsAt hospital 1 (academic), 314 of the 48,057 (0.65%) eligible patient admissions had HO- CDI; 41 of the 8,791 (0.47%) and 75 of the 29,211 (0.26%) of patient admissions at community hospitals 2 and 3, respectively, had HO-CDI. In multivariable analysis, Elixhauser Score was a significant risk factor for HO-CDI at all hospitals when controlling for antibiotic and antacid use; for every one-point increase in Elixhauser Score, there was an increased risk of HO-CDI of 1.27 (95% CI: 1.21, 1.32) at hospital 1, 1.38 (95% CI: 1.24, 1.54) at hospital 2, and 1.28 (95% CI: 1.10, 1.31) at hospital 3. Table 1 shows significant risk factors for HO-CDI for each hospital. When individual comorbidities were assessed in the regression analysis, fluid and electrolyte disorders were a significant risk factor for HO-CDI for all hospitals.ConclusionLaboratory values upon admission and electronically available patient comorbidities are important risk factors for HO-CDI and should be considered for future risk adjustment. Disclosures All authors: No reported disclosures.