Objective To investigate the application value of Clavien⁃Dindo classification (CDC) and comprehensive complication index (CCI) in evaluating early postoperative complications for radical gastric cancer surgery.

Methods The retrospective cohort study was conducted. The clinicopathological data of 1 484 patients who underwent radical gastric cancer surgery in Affiliated Drum Tower Hospital, Medical School of Nanjing University from January 2016 to December 2018 were collected. There were 1 086 males and 398 females, aged (60±17)years. The early complica-tions of patients were classified by the CDC and the CCI was calculated. Measurement data with normal distribution were represented as Mean±SD. Count data are expressed as absolute numbers. Univariate analysis was performed using the chi-square test. Multivariate analysis was conducted using Logistic regression analysis. Correlation analysis between CDC, CCI and duration of postopera-tive hospital stay, total hospital expenses was performed using the Spearman rank correlation. Multiple linear regression analysis was conducted to compare the correlation analysis between CDC, CCI and duration of postoperative hospital stay, total hospital expenses.

Results (1) Distribution of early postoperative complications by CDC and CCI. Among the 1 484 patients, 432 patients developed early postoperative complications, of which 322 cases had a single complication, and 110 cases had multiple complications. Among the 432 patients with complications, the numbers of patients with grades Ⅰ, Ⅱ, Ⅲa, Ⅲb, Ⅳ, and Ⅴ of CDC were 231, 137, 45, 13, 3, and 3 cases, respectively. The CCI in the 432 patients with complications was 17.2±11.7, with peak values of 8.7 and 20.9. (2) Analysis of factors influencing early postoperative complications in patients undergoing radical gastric cancer surgery. Results of multivariate analysis showed that female, age ≥70 years, and preoperative serum albumin <35 g/L were independent risk factors for early postoperative complications in patients undergoing radical gastric cancer surgery (odds ratios=1.391, 1.535, 1.521, 95% confidence intervals as 1.083-1.786, 1.187-1.985, 1.055-2.192, P<0.05). (3) Correlation analysis between CDC, CCI and duration of postoperative hospital stay, total hospital expenses. Among the 432 patients with post-operative complications, both CDC and CCI were positively correlated with duration of postoperative hospital stay (r=0.574, 0.576, P<0.05), and both were positively correlated with total hospital expenses (r=0.413, 0.438, P<0.05). Among the 110 patients with multiple postoperative complications, both CDC and CCI were positively correlated with duration of postoperative hospital stay (r=0.514, 0.537, P<0.05), and both were positively correlated with total hospital expenses (r=0.427, 0.474, P<0.05). (4) Regression analysis between CDC, CCI, and duration of postoperative hospital stay, total hospital expenses. Multiple linear regression analysis of duration of postoperative hospital stay showed that CDC ≥grade 3, age ≥70 years, preoperative C-reactive protein ≥10 g/L, and American Society of Anesthesiology (ASA) classification ≥grade 3 were associated with prolonged duration of postopera-tive hospital stay. The higher CCI indicated longer duration of postoperative hospital stay. The standard b value of CCI was higher than that of CDC ≥grade 3 (0.467 versus 0.212). The regression analysis of total hospital expenses showed that CDC ≥grade 3, age ≥70 years, preoperative serum albumin <35 g/L, ASA classification ≥grade 3, and pathological stage Ⅲ-Ⅳ increased total hospital expenses, while gender of female reduced the expenses. The higher CCI indicated higher hospital expenses. The standard b value of CCI was higher than that of CDC ≥grade 3 (0.449 versus 0.061).

Conclusions Both the CDC and CCI can effectively reflect the severity of postoperative complica-tions following radical gastric cancer surgery. Compared to the CDC, the CCI can better predict post-operative hospital stay and total hospital expenses.