Development and validation of a novel nomogram for postoperative pulmonary complications following minimally invasive esophageal cancer surgery.

Abstract

Postoperative pulmonary complications (PPCs) are the most common complications following minimally invasive esophagectomy (MIE) and can be associated with adverse outcomes. This study aims to construct a nomogram based on clinical factors to predict PPCs and investigate related early outcomes. Clinical data of 969 consecutive patients receiving MIE were retrospectively collected. Univariate and multivariate analysis were performed to select independent predictors. Using independent predictors to develop a nomogram and using a bootstrap-resampling approach to conduct internal verification. Early outcomes of PPCs were analyzed. The incidence of PPCs following MIE was 39.6% (384 out of 969). In multivariate analysis, older age (Odds ratio (OR) 1.034, P < 0.001), higher body mass index (OR 0.993, P = 0.003), heavy smoking (OR 1.396, P = 0.027), FEV1/FVC < 105% (OR 1.958, P < 0.001), chemoradiotherapy (OR 0.653, P = 0.039), estimated blood loss ≥ 400mL (OR 2.582, P = 0.018), general anesthesia (vs Combined thoracic paravertebral blockade, OR 1.578, P = 0.014), operative time ≥ 240min (OR 1.388, P = 0.027), squamous cell carcinoma (OR 2.099, P = 0.036) and conversion to thoracotomy (OR 2.820, P = 0.026) were independent predictors for PPCs. These ten independent predictors were used to develop a nomogram, with concordance index (C index) value of 0.662 and good calibration. After internal validation, similarly good calibration and discrimination (C index, 0.654; 95% CI 0.614-0.690) were observed. Patients developing PPCs had higher rates of anastomotic leakage, reoperation, ICU and 30-day readmissions, and prolonged ICU and hospital stays (P < 0.05). Our study identified ten predictors for PPCs, which were associated with poor early outcomes. The proposed nomogram can be a useful tool to identify patients at high risk of PPCs after MIE.