292 PREOPERATIVE RISK STRATIFICATION IN ESOPHAGEAL SURGERY: COMPARING RISK MODELS WITH THE CLINICAL JUDGEMENT OF THE SURGEON

Abstract

Abstract Numerous prediction models have been proposed to estimate the risk of complications after esophagectomy. However, these models are not commonly used in practice and surgeons generally trust on their own clinical judgment. The aim of this study is to compare the clinical judgment of the surgeons with the existing risk stratification models with regard to predicting complications after esophageal surgery. Methods Patients with resectable esophageal cancer who underwent an esophagectomy between March 2019 and January 2020 were included in this prospective study. Literature was searched to identify prediction models predicting the incidence of postoperative complications after esophagectomy. Clinical judgment of three surgeons was assessed using a standardized form where surgeons could indicate their estimated risk for postoperative complications. The best performing prediction model was compared with the judgment of the surgeons, using the Net Reclassification Improvement (NRI). A higher NRI correlates with better estimation by the surgeon and a negative NRI indicates a better prediction by the prediction model. Results Fifty-three patients were included, 36 patients (68%) developed a complication. Two risk classification models were identified in literature: Model 1 (Lagarde et al, Annals of thoracic surgery, 2008) and model 2 (Reeh et al, Medicine, 2016). Model 1 had a better discriminative ability than model 2 (Area Under the Receiver Operator Curve 0.738 versus 0.609). The NRI for the surgeons combined was −2%, meaning that model 1 outperforms the combined judgment of the surgeons. However, there was a large difference in clinical judgment between surgeons. Figure 1 shows the NRI for all three surgeons separately. Conclusion Surgeons‘assessment does not outperform prediction models in predicting the incidence of postoperative complications after esophageal surgery. However, there is a poor agreement between surgeons regarding their risk assessment based on their clinical judgment. Some surgeons might individually outperform existing risk stratification models. Surgeon’s assessment can therefore still be important when counseling patients about the risks of esophageal surgery in addition to risk stratification models.