Related Experiment Video
Updated: Jan 8, 2026

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Temporal trends in ICU outcomes by BMI: A retrospective propensity-score matched study
Abhi Chand Lohana1, Subhash Chander2, Fnu Sorath3
1Department of Internal Medicine, West Virginia School of Medicine/Camden Clark Program, Parkersburg, WV, USA.
Importance:
Obesity has been associated with adverse outcomes in critical illness, including increased mortality, organ dysfunction, and prolonged ventilation. The "obesity paradox," suggesting improved survival in obese patients, remains controversial in ICU populations. The independent role of body mass index (BMI) on outcomes with modern organ support remains uncertain.
Objective:
To evaluate whether a BMI of 30-44.9 kg/m2 increases the risk of mortality, organ failure, and use of advanced life support compared with BMI of 18.5-29.9 kg/m2 in ICU patients.
Design:
Retrospective multicenter cohort study using a real-world data platform. Patients were stratified by BMI and matched using propensity scores for demographic and clinical factors.
Exposure:
BMI category at ICU admission (30-44.9 vs 18.5-29,9 kg/m2.
Main Outcomes And Measures:
Outcomes were assessed during early and late weeks of ICU stay. Primary outcomes were all-cause mortality; secondary outcomes included cardiac arrest, hemodialysis initiation, ARDS, and use of ECMO.
Results:
Among 29,617 matched patients, obesity was associated with higher early (HR 1.213 [95 % CI 1.16-1.267]) and late (HR 1.213 [95 % CI 1.16-1.267]) mortality. Hemodialysis use was more frequent early (HR 1.49 [95 % CI 1.387-1.6]) and late (HR 1.537 [95 % CI 1.352-1.747]). VV-ECMO was also higher early (HR 1.4 [95 % CI 1.022-1.916]) and late (HR 2.206 [95 % CI 1.14-4.161]). No significant differences were observed in cardiac arrest, ARDS, or ECMO.
Conclusions And Relevance:
Obesity independently increased ICU mortality and need for renal hemodialysis, underscoring the need for targeted preventive and management strategies in this high-risk population.
More Related Videos
Related Concept Videos
Bias in Epidemiological Studies
Regression Toward the Mean
Relative Risk

