Development of an Individualized Surgical Risk Calculator for Abdominoplasty Procedures

Background: Individualized risk calculation for surgical procedures is supplanting coarser, population-based risk estimates because it provides more accurate predictions better understood by clinician and patient alike. By developing publically accessible and user-friendly risk calculators, clinicians can better inform patients about the risks and expectations of aesthetic surgical procedures such as abdominoplasties. To accomplish this goal, we used data from the Tracking Operations and Outcomes for Plastic Surgeons (TOPS) database to develop a surgical risk calculator for abdominoplasty procedures, designed for use as a clinical aid to predict the chance of complication based on patient information.

Methods: Abdominoplasties performed between 2008-2011, including those with concomitant liposuction, were identified from the TOPS database. Duplicate case IDs were excluded to remove cases with multiple procedures. Age, BMI, smoking history, diabetes, American Society of Anesthesiologists' (ASA) Class, facility type, and admission status were examined as relevant clinical parameters. Seroma, dehiscence, surgical site infection (SSI), reoperation, and overall complications were the outcomes of interest for which we generated logistic regression models to predict the chance of their occurrence. These models were evaluated for accuracy using p-value, c-statistic, Hosmer-Lemeshow (H-L), and Brier score.

Results: 22,289 abdominoplasty cases were found, with 5,786 cases ultimately meeting inclusion criteria. Seroma, dehiscence, SSI, and reoperation occurred in 4.1%, 4.4%, 2.0%, and 1.1% of cases respectively. Overall, 9.9% of cases suffered at least one of these complications. Logistic regression models were produced for each of these complications using the clinical parameters of interest, yielding beta values for each parameter (Table 1). Each risk model demonstrated satisfactory performance based on p-value, c-statistic, H-L, and Brier score. The distribution of predicted overall risk of surgical complication was wide and positively skewed (Figure 1), highlighting the inadequacy of a population-based risk assessment and therefore the importance of individualized risk assessment. These models were collectively packaged and published online at AbdominoplastyRisk.org for user-friendly and public use.

Conclusion: With the advent of large surgical outcomes databases, a significant amount of data is available to assist patients with making personalized surgical decisions. We have leveraged this abundance of data to develop a risk calculator that evaluates absolute risk of complications following an abdominoplasty based on personal details about a patient's individual situation in order to yield accurate predictions for their surgical outcomes.