|| Checking for direct PDF access through Ovid
Risk calculators are an emerging tool that provide granular, individualized risk estimation. Presently, there is a paucity of risk calculators specific to plastic surgery. Abdominoplasty is a popular plastic surgery procedure associated with moderate risks of complications, such as surgical-site infection and dehiscence, and would benefit from the ability to provide patients with accurate, personalized risk assessment.Abdominoplasties from the National Surgical Quality Improvement Program 2005 to 2014 database were identified by Current Procedural Terminology code. Relevant perioperative variables included age, body mass index, sex, smoking history, diabetes, American Society of Anesthesiologists class, pulmonary comorbidities, hypertension, bleeding disorders, and operative duration. Multiple logistic regressions were used to generate 30-day risk models for medical complications, surgical-site infection, wound dehiscence, and reoperation. Internal validation of model performance was conducted using C-statistics, Hosmer-Lemeshow tests, and Brier scores.Among the 2499 cases identified, complication rates were as follows: medical complications, 3.8 percent; superficial surgical-site infection, 2.4 percent; deep or organ-space surgical-site infection, 1.6 percent; wound dehiscence, 1.0 percent; and reoperation, 2.0 percent. Risk prediction models were constructed and all demonstrated good predictive performance, with mean predicted risks closely matching observed complication rates. The distributions of predicted risk were wide and contained outliers with very high risk. A user-friendly, open-access online interface for these models is published at AbdominoplastyRisk.org.The authors developed an internally valid risk calculator for which individual patient characteristics can be input to predict 30-day complications after abdominoplasty. Given that estimated risk can vary widely, individualized risk assessment is a way to enhance shared decision-making between surgeon and patient.