Simple preoperative risk scale accurately predicts perioperative mortality following esophagectomy for malignancy

Surgery remains one of the major treatment options available to patients with esophageal cancer, with high mortality in certain cohorts. The aim of this study was to develop a simple preoperative risk scale based on patient factors, hospital factors, and tumor pathology to predict the risk of perioperative mortality following esophagectomy for malignancy. The Nationwide Inpatient Sample database was used to create the risk scale. Patients who underwent open or laparoscopic transhiatal and transthoracic esophageal resection were identified using International Classification of Diseases, 9th edition codes. Patients <18 years and those with peritoneal disease were excluded. Multivariate logistic regressions were used to define a predictive model of perioperative mortality and to create a simple risk scale. From 1998 to 2011, a total of 23 751 patients underwent esophagectomy. The observed overall perioperative mortality rate for this cohort was 7.7%. Minimally invasive techniques, and operations performed in higher volume centers were protective, whereas increasing age, comorbidities and diagnosis of squamous cell carcinoma were independent predictors of mortality. Based on this population, a risk scale from 0-16 was created. The calibration revealed a good agreement between the observed and risk scale-predicted probabilities. A set of sensitivity/specificity analyses was then performed to define normal (score 0-7) and high risk (score 8-16) patients for clinical practice. Mortality in patients with a score of 0-7 ranged from 1.3-7.6%, compared with 10.5-34.5% in patients with a score of 8-16. This simple preoperative risk scale may accurately predict the risk of perioperative mortality following esophagectomy for malignancy and can be used as a clinical tool for preoperative counseling.