BACKGROUND: Patients admitted with Clostridium difficile infection are managed in a variety of settings. If their care is inadequate, these patients can rapidly deteriorate. OBJECTIVE: The purpose of this study was to evaluate whether mortality for patients admitted with C difficile differed between medical and general/colorectal surgery services. DESIGN: This was a retrospective cohort study with multivariable logistic regression used to evaluate the effect of admitting service on in-hospital mortality rates, with propensity score matching used to validate this relationship. SETTINGS: The study was conducted at a single, tertiary care center. PATIENTS: Inpatients with a positive C difficile stool test within 24 hours of admission to medical or surgical services were identified (2005-2015) using institutional electronic data sources. MAIN OUTCOME MEASURE: We measured inpatient mortality rate. RESULTS: Of 1175 patients, 985 (83%) were admitted to medical services, whereas 190 (17%) were admitted by surgeons. Medical patients were older (63.9 vs 58.9 years; p = 0.001) and had a mean of 0.6 additional comorbidities (p < 0.001); cohorts were similar regarding vasopressors, peak white blood cell counts, and rate of intensive care unit admissions. Mortality was lower among surgery patients (2.6% vs 6.8%; p = 0.028), and logistic regression demonstrated lower odds of mortality for this group OR = 0.18 (95% CI, 0.05-0.58)). After propensity score matching for age, comorbidities, and severity of disease, this difference was confirmed (2.6% vs. 9.5%). A higher incidence of total colectomy for surgery patients (14.2% vs 0.4%) was a causal factor in their longer lengths of stay and higher total hospital costs. The time between orders for stool testing and metronidazole therapy was shorter in the surgery group (1.8 vs 3.8 hours; p = 0.002), although this trend was not observed with vancomycin therapy. LIMITATIONS: This was a retrospective study from a single institution, thereby limiting generalizability, with a lack of information regarding premorbid creatinine levels, ileus, or megacolon. CONCLUSIONS: In-hospital mortality rates for patients admitted with C difficile were lower on surgery services, perhaps in part related to higher colectomy rates.
All Science Journal Classification (ASJC) codes