Assessment of thoracic endografting operative mortality risk score: Development and validation in 2,000 patients

Background In this study we derive and validate a composite risk index termed the Assessment of Thoracic Endografting Operative Mortality, or ATOM, risk score. Methods All thoracic endovascular aortic repairs (TEVAR) in the American College of Surgeons National Surgical Quality Improvement Project (NSQIP) between 2005 and 2012 were identified. The primary outcome was operative mortality. After evaluating the association of over 60 preoperative variables and operative mortality in univariate analysis, a multivariable model was developed. Significant risk factors were assigned points equivalent to their odds ratio rounded to the nearest whole integer in the final multivariable model. Results Overall, 1,981 patients comprised the study population, including 1,486 (75.0%) in the derivation and 495 (25.0%) in the validation cohort. There were 173 (8.7%) operative mortalities. A 30-point risk score incorporating 10 risk factors was generated and found to be highly predictive of operative mortality in the derivation (odds ratio [OR] 1.36, p < 0.001) and validation cohorts (OR 1.24, p < 0.001). The models used to create and validate the ATOM score were robust (C indices 0.84 and 0.83, respectively). There was strong correlation between predicted mortality rates based on the derivation cohort and actual mortality rates in the validation cohort (r = 0.75, p < 0.001). Operative mortality based on low (ATOM < 5), moderate (ATOM 5 to 9), and high risk (ATOM ≥ 10) was 1.3%, 6.6%, and 24.0%, respectively (p < 0.001). Higher ATOM scores also correlated with higher complication rates and longer hospital stays. Conclusions The ATOM score is a significant predictor of operative mortality in TEVAR and can be used for preoperative risk stratification.