Association of Obesity With Septic Complications After Major Abdominal Surgery: A Secondary Analysis of the RELIEF Randomized Clinical Trial

Abstract

Importance: Body mass index (BMI) has been the most common obesity measure to assess perioperative risk; however, cardiometabolic risk is associated with the burden of visceral fat. Definitive evidence on the association of visceral fat measures, such as waist circumference and waist-to-hip ratio (WHR), with postoperative complications is lacking.

Objective: To compare the value of waist circumference with the value WHR and BMI in predicting adverse outcomes, including major septic complications and persistent disability, following major abdominal surgery.

Design, setting, and participants: This planned secondary analysis of the Restrictive vs Liberal Fluid Therapy for Major Abdominal Surgery (RELIEF) randomized clinical trial took place at 47 centers in 7 countries between October 2013 and September 2016, with 90-day follow-up. A total of 2954 adult RELIEF participants were coenrolled in this secondary analysis. Data analysis took place from December 2018 to September 2019.

Exposures: Waist circumference, WHR, and BMI measurements.

Main outcomes and measures: The primary outcomes were 30-day major septic complications and 90-day persistent disability or death.

Results: Of 2954 eligible participants, 2755 were included (mean [SD] age, 65.9 [12.9] years; 1426 [51.8%] men) in the final analysis. A total of 564 participants (20.6%) experienced at least 1 major septic complication within 30 days after surgery (sepsis, 265 [9.7%]; surgical site infection, 409 [14.9%]; anastomotic leak, 78 [2.8%]; pneumonia, 104 [3.8%]). Waist circumference had a statistically significantly larger odds ratio (OR) and discrimination indices as well as a smaller prediction error than WHR or BMI for 30-day major septic complications or death (waist circumference: OR, 1.44; 95% CI, 1.28-1.62; P < .001; area under the receiver operating characteristic curve, 0.641; net reclassification index, 0.266; integrated discrimination improvement [score × 104], 152.98; Brier score, 0.162; WHR: OR, 1.15; 95% CI, 1.03-1.28; P = .01; area under the receiver operating characteristic curve, 0.621; net classification index, 0.199; integrated discrimination improvement [score × 104], 28.47; Brier score, 0.164; BMI: OR, 1.33; 95% CI, 1.17-1.50; P < .001; area under the receiver operating characteristic curve, 0.629; net reclassification index, 0.205; integrated discrimination improvement [score × 104], 85.61; Brier score, 0.163) but not for any other outcomes.

Conclusions and relevance: In this secondary analysis of the RELIEF randomized clinical trial, waist circumference was observed to be superior to other adiposity indices in predicting 30-day major septic complications alone or in conjunction with death following elective major abdominal surgery. Findings suggest that waist circumference is a useful adiposity measure that should be incorporated in preoperative risk assessment for such complications.

Trial registration: ClinicalTrials.gov identifier: NCT01424150.

Conflict of interest statement

Conflict of Interest Disclosures: Drs Forbes and Martin reported receiving grants from Monash University during the conduct of the study. No other disclosures were reported.

Figures

Figure.. Study Flow Diagram

Of 2755 eligible patients, 2742 (99.6%) had data on septic complications and pulmonary edema at 30 days, 2683 (97.5%) had data on death and disability at 90 days, 2663 (96.7%) had data on acute kidney injury, 2751 (99.9%) had data on unplanned intensive care unit admission at 30 days, and 2668 (96.9%) had data on quality of recovery at 30 days. Of these, 2636 (96.1%) with data on septic complications, 1695 (61.8%) with data on pulmonary edema, 2661 (99.2%) with data on death or disability to 90 days, 2598 (97.6%) with data on acute kidney injury, 2661 (96.7%) with data on unplanned intensive care unit admission at 30 days, and 2668 (100%) with data on quality of recovery at 30 days were included in the analysis.