Introduction Anastomotic leakage is generally assumed to occur from approximately 5 days after surgery unless there has been a technical error in anastomotic construction. In order to investigate whether this assumption is correct, clinical changes associated with anastomotic leakage were studied in a homogeneous group of patients.

Method Individuals were included from a prospectively collected database of patients undergoing laparoscopic total colectomy and ileorectal anastomosis (LTC and IRA) for familial adenomatous polyposis (FAP) between 2006–2013 at a single institution. Patients suffering anastomotic leakage (case group) were compared with controls that did not leak. These controls were selected as the consecutive case before and after each leak case (2 controls for each case). Panel data was collected as repeated measures of continuous post-operative parameters and time-series regression performed using a double subscript structure. A generalised least squares multivariate approach was applied in a random effects setting to calculate correlations for observations with the dependent outcome of anastomotic leak. Regression calculations were performed according to individual observations at each time point to identify outcome correlations.

Results Data from 95 patients who underwent LTC and IRA for FAP between 2006 and 2013 was available for analysis. All operations were performed laparoscopically within an enhanced recovery programme. 10 patients (10.5%) developed International Classification Grade C leakage (requiring laparotomy), with a mean time to diagnosis of 7.5 +/- 2.6 days. The mean age of the 10 leak patients and 20 controls was 31.9 (+/-17.0) years. There was no significant difference between groups in age, sex, ASA grade, physiology, and operative score. Multivariate analysis of physiological parameters demonstrated a significant difference between leak patients and controls in respiratory rate, urine output, diastolic blood pressure, and heart rate within 48 h of operation. Biochemical parameters, including white cell count, neutrophil count, albumin, and C-reactive protein, did not differ significantly.

Conclusion Following anastomotic leakage, significant changes in physiological parameters are identifiable within 48 h of surgery. Leakage may occur earlier than has previously been thought and identification may provide opportunities for earlier intervention in order to improve outcome.

Disclosure of interest None Declared.