Introduction Predictive markers of anastomotic leak have been identified and include the use of post-operative C-reactive protein (CRP). A rising or post-operative day (POD) 3 CRP ≥ 140 is suggestive of anastomotic leak.

\We performed a retrospective review of post-operative CRP in anastomotic leak against the timing of clinical or radiologically detected anastomotic leak.

Method A prospectively collected database was used to identify all colorectal cancer resections over a 7-year period at a District General Hospital. This included both elective and emergency resections. From this data anastomotic leaks were identified. Post-operative CRP at day 1 and 3 and the time of clinical or imaging confirmed leak was retrospectively recorded for these patients. The data was then analysed.

Hospital stay and in-hospital mortality was also recorded.

Results A total of 1311 patients underwent colorectal cancer resections between January 2008 and December 2014. Of these we identified 45 anastomotic leaks (3.4%). These included 31 anterior resections, 8 right hemicolectomy, 4 left hemicolectomy and 2 other. Of the 31 anterior resections 48% (15/31) were de-functioned.

80% (35/44) of anastomotic leaks had a rising CRP and/or CRP ≥ 140 on POD 3. Of these 64% (29/44) had a POD 3 CRP ≥ 140 on its own. 84% (38/45) of patients underwent a diagnostic CT for clinically suspected anastomotic leak. The mean average days to clinical or radiologically detected anastomotic leak was 7 days.

The mean average length of stay was 33 days (range 10–88). There was a 4% (2/45) associated in-hospital mortality.

Conclusion Overall anastomotic leak rate is low. CRP has been shown to be sensitive and clinically relevant in our cohort. A proactive approach responding to CRP could lead to earlier detection of anastomotic leak. We recommend early CT imaging if CRP ≥140 on POD 3.

Disclosure of interest None Declared.