Prediction model for complications after low anterior resection based on data from 33,411 Japanese patients included in the National Clinical Database

Background Low anterior resection is associated with a relatively high incidence of postoperative morbidities, including anastomotic leakage and other operative site infections, which sometimes result in postoperative mortality. Therefore, recognition of the incidence and risk factors of postoperative complications following low anterior resection is essential. Methods Data from the National Clinical Database on patients who had undergone low anterior resection in 2011 and 2012 were retrospectively analyzed. Multiple logistic regression analyses were performed to generate predictive models of postoperative complications. Receiver-operator characteristic curves were generated, and the concordance index was used to assess the model's discriminatory ability. Results The number of patients who had undergone low anterior resection was 33,411. Seven complications, namely, overall operative site infections except for leakage, anastomotic leakage, urinary tract infection, pneumonia, renal failure, systemic sepsis, and cardiac events, were selected to construct statistical risk models. The concordance indices for the first 2 complications, which were dependent on the operative procedure, were relatively low (0.593–0.625), and the other 5, unrelated to operative procedures, showed high concordance indices (0.643–0.799). Conclusion This study created the world's second risk calculator to predict the complications of low anterior resection as a model based on mass nationwide data. In particular, this model is the first to predict anastomotic leakage.