Medical Versus Surgical ICU Obese Patient Outcome: A Propensity-Matched Analysis to Resolve Clinical Trial Controversies

Objectives: To determine the short- and long-term mortality of obese ICU patients following medical as opposed to surgical admission and the relation between obesity and mortality. Design: Retrospective analysis of prospectively collected data, using a propensity score-matched analysis of patients with medical or surgical admission. Setting: One French mixed medical-surgical ICU. Patients: Critically ill obese patients (body mass index >= 30 kg/m(2)) and nonobese patients admitted during a 14-year period. Interventions: None. Measurements and Main Results: Seven-hundred ninety-one obese patients and 4,644 nonobese patients were included, 338 (43%) and 2,367 (51%) medical and 453 (57%) and 2,277 (49%) surgical obese and nonobese patients, respectively. Mortality was significantly higher in medical than in surgical obese patients in ICU (25% vs 12%; p < 0.001) and up to 365 days (36% vs 18%; p < 0.001) post ICU admission. One-to-one propensity score matching generated 260 pairs with well-balanced baseline characteristics. After matching on propensity score, mortality was still significantly higher in medical patients both in the ICU (21% vs 13%; p = 0.03) and up to 365 days (30% vs 20%; p = 0.01) post ICU admission. Obesity was not significantly associated with mortality both in univariate analysis (140 obese patients [15%] in the dead group vs 651 [14%] in the alive group; p = 0.72) and multivariate analysis (odds ratio, 1.09 [95% CI, 0.86-1.38]; p = 0.49) after adjustment for Simplified Acute Physiology Score II, age, category of admission, history of cardiac disease, and history of respiratory disease. Conclusions: After careful matching, the data suggest that ICU mortality in obese population was higher in the medical group than in the surgical group and remains significantly higher 365 days post ICU admission.