Patient-Specific Risk Adjustment Improves Comparison of Infection Rates Following Posterior Fusion for Adolescent Idiopathic Scoliosis

There is increasing interest in surgeon "performance measures." Understanding patient factors that increase the risk of adverse events is important for the comparison of such metrics among surgeons and centers. The purpose of this study was to ascertain whether patient characteristics, beyond the control of the surgeon, were associated with increased risk of postoperative infection following posterior spinal fusion for the correction of adolescent idiopathic scoliosis (AIS) and to establish a "risk-adjusted" method of reporting postoperative infection rates. We reviewed the data of patients from 14 participating scoliosis treatment centers who experienced an infection within 90 days following posterior spinal fusion for the treatment of AIS. Patients with a deep infection (irrigation and debridement performed) were compared with those without an infection with regard to age, sex, body mass index (BMI) percentile for age, Lenke classification of curve type, primary curve magnitude, and estimated 3-dimensional sagittal kyphosis (T5-T12). A regression model was created to identify variables that were associated with infection, and the performance of the risk model was evaluated. The actual infection rate by site was divided by the predicted infection rate for that site and multiplied by the overall rate to create a risk-adjusted rate. Of 2,122 patients analyzed, 21 (1.0%) had an infection within 90 days following surgery. Obesity was the only significant risk factor (odds ratio [OR], 7.6; p ≤ 0.001), with the resultant model demonstrating good discrimination and calibration. For the 8 sites that enrolled ≥100 patients, the predicted infection rates based on the proportion of obese patients ranged from 0.8% to 1.2%. The range of the risk-adjusted infection rates varied more substantially, from 0.2% to 2.0%. For the posterior approach to surgical correction of AIS, the only identified patient risk factor for postoperative infection was a BMI percentile for age of ≥95%. To assess infection rates, we suggest adjusting for the proportion of obese patients in the cohort. A risk-adjusted infection rate for posterior spinal fusion with instrumentation for AIS allows for more accurate assessment of performance and comparison among centers. Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.