Should They Stay or Should They Go? The Utility of C-Reactive Protein in Predicting Readmission and Anastomotic Leak...

Hospital readmission and anastomotic leak following colorectal resection have a negative impact on patients, surgeons, and the health care system. Novel markers of patients unlikely to experience these complications are of value in avoiding readmission.

This study aimed to determine the predictive value of C-reactive protein for readmission and anastomotic leak within 30 days following colorectal resection.

This is a retrospective review of a prospectively compiled single-institution database.

From January 1, 2013, to July 20, 2017, consecutive patients undergoing elective colorectal resection with anastomosis without the presence of proximal intestinal stoma, who had C-reactive protein measured on postoperative day 3, were included.

The primary outcome measured was the predictive value of C-reactive protein measured on postoperative day 3 for readmission or anastomotic leak within 30 days after colorectal resection.

Of the 752 patients examined, 73 (10%) were readmitted within 30 days of surgery and 17 (2%) had an anastomotic leak. Mean C-reactive protein in patients who neither had an anastomotic leak nor were readmitted (127 ± 77 mg/L) was lower than for patients who were readmitted (157 ± 96 mg/L, p = 0.002) and lower than for patients who had an anastomotic leak (228 ± 123 mg/L, p = 0.0000002). The area under the receiver operating characteristic curve for the diagnostic accuracy of C-reactive protein for readmission was 0.59, with a cutoff value of 145 mg/L, generating a 93% negative predictive value. The area under the curve for the diagnostic accuracy of C-reactive protein for anastomotic leak was 0.76, with a cutoff value of 147 mg/L generating a 99% negative predictive value.

This study was limited by its retrospective design and because all patients were treated at a single center.

Patients with a C-reactive protein below 145 mg/L on postoperative day 3 after colorectal resection have a low likelihood of readmission within 30 days, and a very low likelihood of anastomotic leak. See Video Abstract at http://links.lww.com/DCR/A761.