Study objective: Overweight patients seem to have a poorer outcome and a higher risk of complications during their stay in the ICU. We conducted a prospective study in order to examine the relationship between body mass index (BMI) and mortality among these patients.
Design: Prospective clinical study.
Setting: A 24-bed medical ICU in a university-affiliated hospital.
Methods: All patients hospitalized in the ICU over a 1-year period were included except those dying or being discharged from the hospital within 24 h of admission. Overweight patients were defined as those having a BMI > 75th percentile of this selected ICU population. Other data collected were demographic and ICU-related data. The Mann-Whitney test was used to compare numeric data between groups (ie, obese and nonobese populations). Variables that were significantly associated with ICU mortality by univariate analysis were entered into a multiple logistic regression model, allowing the determination of independent predictors.
Results: Eight hundred thirteen patients were included in the study. The limit of the upper quartile of the BMI was 27. This value was used to separate obese (n = 215) and nonobese (n = 598) groups. Significant differences between obese and nonobese patients were observed in age, length of stay in the ICU, simplified acute physiology score (SAPS) II, and ICU mortality. The observed mortality of obese patients was significantly higher than that predicted by SAPS II (32% vs 18%, respectively; p = 0.001). No difference was observed in frequency of nosocomial infection or duration of mechanical ventilation for mortality in ICU patients. Using a multivariate analysis, the predictive factors of mortality were SAPS II (p < 0.0001) and BMI > 27 (p < 0.01).
Conclusion: This is the first prospective study showing high BMI value as an independent prognostic factor of mortality for ICU patients. The prognostic scoring systems currently in use, which were designed to predict the mortality of ICU patients, do not include BMI or do not consider obesity. These may underestimate, therefore, the risk for the specific population of obese patients.