To investigate the risk factors of postoperative short-term complications in patients with rectal cancer and to establish a nomogram model to predict the postoperative short-term complications.

A retrospective study was conducted to collect and analyze the clinicopathological data and related surgical information of patients who underwent radical resection of rectal cancer in the Second Affiliated Hospital of Nanjing Medical University from January 2015 to September 2018. A total of 204 patients were collected, including 135 male patients and 69 female patients. Pathological TNM staging: 11 cases of stage 0, 42 cases of stage Ⅰ, 62 cases of stage Ⅱ and 89 cases of stage Ⅲ. The postoperative complications of grade Ⅱ and above according to the modified Clavien-Dindo classification of surgical complications (Version 2019) were defined as clinically significant complications. The relationship between the general baseline characteristics, pathological characteristics, operation information characteristics and the occurrence of clinically significant complications was analyzed and the clinical variables were analyzed by univariate analysis and multivariate analysis. The nomogram prediction model was established by R software (R4.0.3) and the nomogram model was verified and evaluated by ROC curve and C-index.

Of the 204 patients, 39 (19.11%) had clinically significant complications. The common postoperative complications were anastomotic leakage (14 cases, 9.03%; 155 cases of Dixon and coloanal anastomosis), intestinal obstruction (14 cases, 6.86%), pulmonary infection (13 cases, 6.37%), stoma related complications (5 cases, 5.95%; 84 cases of Miles operation, Hartmann operation or preventive stoma),postoperative bleeding (9 cases, 4.41%), abdominal infection/pelvic infection (6 cases, 2.94%), dysuria/urinary retention/urinary tract infection (5 cases, 2.45%), incision infection (3 cases, 1.47%), venous thrombosis (1 case, 0.49%). Univariate analysis showed that basic diseases (χ²=6.677, P=0.010), previous abdominal surgery history (χ²=5.260, P=0.022), preoperative albumin＜40 g/L (χ²=9.495, P=0.002), preoperative CEA increased (χ²=4.976, P=0.026), distance between the lower edge of the tumor and the anal edge ≤7 cm (χ²=6.683, P=0.010), intraoperative bleeding＞100 mL (χ²=10.694, P=0.001) were risk factors of clinically significant complications after radical resection of rectal cancer. Multivariate analysis showed that basic diseases (OR=2.770, 95% CI: 1.175~6.531, P=0.020), previous abdominal surgery history (OR=2.538, 95% CI: 1.023~6.295, P=0.044), distance between the lower edge of the tumor and the anal edge ≤7 cm (OR=2.376, 95% CI: 1.077~5.239, P=0.032), intraoperative blood loss＞100 mL (OR=3.154, 95% CI: 1.339~7.427, P=0.009), preoperative albumin＜40 g/L (OR=3.403, 95% CI: 1.514~7.653, P=0.003) were independent risk factors for clinically significant complications after radical resection of rectal cancer and the nomogram clinical prediction model was established. Basic diseases were for 78 points, previous abdominal surgery history was for 72 points, distance between the lower edge of the tumor and the anal edge ≤7 cm was for 74 points, preoperative albumin＜40 g/L was for 100 points and intraoperative blood loss＞100 mL was for 80 points. Adding all the points was the total score and the probability corresponding to the total score is that the model predicts the incidence of clinically significant complications after radical resection of rectal cancer. The discrimination test showed that the AUC value was 0.775 (95% CI: 0.694~0.855), and the c-index index was 0.747 after repeated sampling internal validation. The calibration curve showed a good consistency between the prediction results and the actual results.

Basic diseases, previous abdominal surgery history, distance between the lower edge of the tumor and the anal edge ≤7 cm, intraoperative blood loss＞100 mL, preoperative albumin＜40 g/L are the independent risk factors for clinical significant complications after radical resection of rectal cancer. Adequate preoperative evaluation, nutritional support, accurate surgical manipulation and reduction of blood loss are effective measures to reduce the incidence of clinically significant complications in patients with rectal cancer. The nomogram prediction model constructed in this study has high predictive value for the probability of clinically significant complications after radical resection of rectal cancer.