Debunking the "Obesity Paradox": Obesity is Associated With Adverse Outcomes in Emergency General Surgery

Introduction: Obesity is increasingly prevalent both nationwide and in the emergency general surgery (EGS) population. While previous studies have shown that obesity may be protective against mortality following EGS procedures, the association between body mass index (BMI) and postoperative outcomes, as well as intraoperative decision-making, remains understudied. Methods: The National Surgical Quality Improvement Program 2015-2019 database was used to identify all adult patients undergoing an open abdominal or abdominal wall procedure for EGS conditions. Our outcomes included 30-d postoperative mortality, composite 30d morbidity, delayed fascial closure, reoperation, operative time, and hospital length of stay (LOS). Multivariable logistic regression models were used to explore the association between BMI and each outcome of interest while adjusting for patient demographics, comorbidities, laboratory tests, preoperative and intraoperative variables. Results: We identified 78,578 patients, of which 3121 (4%) were categorized as underweight, 23,661 (30.1%) as normal weight, 22,072 (28.1%) as overweight, 14,287 (18.2%) with class I obesity, 7370 (9.4%) with class II obesity, and 8067 (10.3%) with class III obesity. Class III obesity was identified as a risk factor for 30-d postoperative morbidity (adjusted odds ratio 1.14, 95% CI, 1.03-1.26, P < 0.01). An increase in obesity class was also associated with a stepwise increase in the risk of undergoing delayed fascial closure, experiencing a prolonged operative time, and having an extended LOS. Conclusions: Obesity class was associated with an increase in delayed fascial closure, longer operative time, higher reoperation rates, and extended hospital LOS. Further studies are needed to explore how a patient's BMI impacts intraoperative factors, influences surgical decision -making, and contributes to hospital costs. <feminine ordinal indicator> 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.