Comparison of the Clavien-Dindo and Comprehensive Complication Index systems for grading of surgical complications after colorectal resections

Abstract

Introduction

Postoperative complications are associated with prolonged hospital stay and a rise in costs of treatment. The Comprehensive Complication Index (CCI) was developed as a scoring system that does not only take the most severe complication into account but all complications after surgery. Our aim was to compare the Clavien-Dindo scoring system with the CCI in predicting length of hospital stay (LOHS) and in-hospital costs after colorectal resections.

Methods

Complications occurring after surgical procedures, performed between October 2012 and September 2013, were prospectively recorded. During this period 164 patients developed complication(s). Only patients that underwent a colorectal resection were included. Multivariable linear regression analysis was performed to find independent predictors of in-hospital costs and LOHS.

Results

64 patients (age (range): 69 (10–91) years, M/F: 36/28) were retained. 46 (71.9%) patients had a Clavien-Dindo score ≥ IIIb. Median (IQR) CCI was 40 (30.2–53.9). Mean (±SD) in-hospitals costs for all patients were €12,920 ± €10,229. The adjusted difference (95% CI, p-value) in in-hospital costs for minor and major (Clavien-Dindo ≥ IIIb) complications was 10,021 (€4283 to €15,759, p = 0.001). A 10 point increase in CCI increased in-hospital costs by €2040. Multivariable analysis retained CCI > 40 as the only independent risk factor for increased in-hospital costs (Standard Beta Coeffic (p-value): 8063 (p = 0.022).

Conclusion

CCI is a better predictor of in-hospital costs than Clavien-Dindo score to classify complications after colorectal resections, as it captures all complications. Further research is warranted to extrapolate our findings to other sub-specialities of surgery.

Keywords:

• Clavien-Dindo classification
• Comprehensive Complication Index
• colorectal resections
• postoperative complications
• socio-economic impact of complications

Disclosure statement

No potential conflict of interest was reported by the author(s).