C-reactive protein and anastomotic leakage after colorectal cancer surgery: useful marker or not?

Exsel, U. van (Ursula) (2015) C-reactive protein and anastomotic leakage after colorectal cancer surgery: useful marker or not? thesis, Medicine.

Text ExselvanU.pdf Restricted to Registered users only Download (1MB)

Abstract

Introduction: Anastomotic leakage is a dreaded complication after colorectal surgery. Several studies have shown absolute ‘C-reactive protein’ (CRP) as marker for anastomotic leakage. However, the accuracy of absolute CRP as marker for anastomotic leakage is questionable. CRP cutoff values differ considerably between studies. Also between individual patients absolute CRP values can be different. Therefore we hypothesize that the postoperative course of CRP is a more reliable parameter to predict the development of anastomotic leakage. The aim of this study was to evaluate the postoperative course of CRP and its potential correlation with the development of anastomotic leakage after elective colorectal cancer surgery. Methods: In a retrospective single-center study 873 patients were analyzed who underwent colorectal surgery between 2008 and 2013. The diagnostic accuracy of absolute CRP for anastomotic leakage was examined by ‘receiver operator characteristic’ (ROC) curves. Optimal CRP cutoff values were determined. Logistic regression analysis was used to evaluate the correlation between the postoperative CRP course and development of anastomotic leakage. With the combination of postoperative CRP course and absolute CRP values, a prediction model for anastomotic leakage was created. Results: After exclusion, 403 patients were analyzed who underwent elective colorectal cancer surgery with primary anastomosis. Anastomotic leakage was diagnosed in 24 patients (6.0%). ROC curves showed good diagnostic accuracy of absolute CRP for leakage on postoperative day (POD) 2 till 7. Optimal cutoff values, positive and negative predictive values are given in table B. Logistic regression analysis showed a significant association between the development of anastomotic leakage and increase in postoperative CRP, regardless of how much increase and the initial absolute CRP value (p < 0.05) (see table B). The prediction model for anastomotic leakage, which combined increase in postoperative CRP and absolute CRP values, was accurate. Conclusion: Our study shows that increase in the postoperative CRP course is significantly correlated with the development of anastomotic leakage. Prediction of anastomotic leakage using only absolute CRP seems inaccurate for individual patients. A prediction model with both ‘increase in CRP’ as well as ‘absolute CRP’ shows an innovative practical tool for the prediction of anastomotic leakage.

Item Type:	Thesis (Thesis)
Supervisor name:	Noo, Dr. M.E. de and Deventer Hospital and Department Surgery
Faculty:	Medical Sciences
Date Deposited:	25 Jun 2020 10:56
Last Modified:	25 Jun 2020 10:56
URI:	https://umcg.studenttheses.ub.rug.nl/id/eprint/1679

Actions (login required)

View Item