Background: High body mass index (BMI) has long been recognized as a risk factor for postoperative complication among total hip arthroplasty (THA) patients. However, recent studies showed mixed results in the effect of high BMI on surgical outcomes. Our study is to examine the association of preoperative BMI with complication incidence, stratified by age and gender.
Methods: We queried the American College of Surgeons National Surgical Quality Improvement Project database to identify patients who underwent elective primary THA between 2012 and 2016. We examined the associations between BMI as a continuous and a categorical variable and risk of 30-day postoperative complication, using 2 multiple polynomial logistic regression models. We also created predictive plots to graphically assess the relationship between BMI and complication by gender and age.
Results: In total, 117,567 eligible patients were included in the analyses. The predictive probability of all-type postoperative complications showed a U-shaped relationship with continuous BMI values (range 10-65 kg/m2). The lowest complication risks occurred in patients with BMI between 35 and 40. Females had higher complication rate than males across all BMI values. This U-shaped relationship was only observed among patients younger than 60 years old, while the associations appear to be inversely linear among patients aged greater than 60 years.
Conclusion: Our results suggest that the current theory of a linear association between BMI and complication risk may not apply to elective primary THA. Strict BMI cutoffs may not minimize risk, especially among patients over 60 years old. Orthopedic surgeons should factor in patient-specific variables of age and gender when determining acceptable surgical risk given a particular BMI value.
Keywords: NSQIP; body mass index; complication; nonlinear association; obesity; total hip arthroplasty.
Copyright © 2020 Elsevier Inc. All rights reserved.