Comparison of the Comprehensive Complication Index and Clavien-Dindo Classification for grading of postoperative complications after Ivor-Lewis esophagectomy in a high-volume center

Abstract

Background: The Clavien-Dindo Classification (CDC) is widely used to measure postoperative complications, yet it only counts the most severe complication in each patient and therefor tends to underestimate the impact of multiple complications in one case. For a more precise analysis of surgical outcome the Comprehensive Complication Index (CCI) was developed, it is based on the CDC but takes into account all complications in each patient. We compared CDC and CCI after Ivor-Lewis esophagectomy in our high-volume center. Methods: Confounder analysis was performed by a retrospective analysis of data using chi-squared test and U-test. Correlation of CCI with length of hospital stay (LOS) was determined using Pearson correlation and correlation of CDC with LOS was determined using Spearman's Rho. The difference between CCI and CDC systems in LOS was controlled by comparison of 1 complication versus >1 complications using analysis of variance (ANOVA). Results: Two hundred and seventy-nine patients underwent esophagectomy from 2014 to 2018 in our center. Overall, 277 complications were recorded in 162 patients. Age >65 and ASA-classification III/IV were found to be risk factors for major complications (CDC >IIIa). CCI stronger correlates with LOS than CDC as the correlation coefficient from dependent samples was significantly higher [CCI (0.74) vs. CDC (0.661), P=0.01]. Patients with more than one complication have significantly prolonged LOS over patients with one complication [mean LOS 21 vs. 31 days (95% CI: 19-22 vs. 27-36 days)]. Conclusions: LOS is described better in complicated courses using CCI than CDC. Therefore, CCI should be used in the literature when describing postoperative complications as well as when evaluating quality of surgical treatment.