Objective: Anastomotic leak after ileocolectomy or right colectomy occurs infrequently and there are little published data on the risk factors for this complication. Our aim was to identify the clinical and operative factors that are associated with ileocolic anastomotic (ICA) leak.
Methods: From 1995 to 2005, all patients who underwent open or laparoscopic ileocolic resection or right colectomy with ICA were reviewed. The incidence of ileocolic anastomotic leak was determined by identification of the subset of patients who were either diagnosed with the ICD-10 codes for “complication of intestinal anastomosis”, “abdominal abscess”, “pelvic abscess”, or underwent repeat laparotomy within 30 days of ICA. ICA leak was defined as an anastomotic complication that required operative intervention. The subset of patients with ICA leak (Group A) was then compared in a 1:2 case-match format to patients who underwent ICA with no leak (Group B). The groups were compared for age, gender, diagnosis, type of surgery (open or laparoscopic), ASA, BMI, comorbidities, steroid use, preoperative hemoglobin, operative time, anastomotic technique, and intra-operative blood loss (EBL). Statistical analysis was performed with the Chi-squared and Wilcoxon Rank tests.
Results: In the study period, the overall leak rate was 29/1973 (1.5%). In these 29 patients, the median age was 47 years and 15(52%) were males. The diagnosis was neoplasia (48%) and Crohn's disease (52%). Operative treatment was disconnection of anastomosis and end ileostomy in 16 (55%), repair of leak with diverting loop ileostomy in 11 (38%), and repair without diversion in 2 (7%). The perioperative mortality was 14%. In group A, pre-op anemia, operative time, and EBL were significantly increased compared to Group B (p≤.05) (Table). Group A and B were similar with regard to comorbidities, ASA, steroid use, BMI, and anastomotic technique.
Conclusions: ICA leaks were more common in patients with pre-operative anemia, long operations, and increased operative blood loss. This suggests that ICA leak is more likely to occur in debilitated patients who undergo complex operations. In this subset of patients, a heightened awareness of the increased risk for leak and a decreased threshold for construction of a diverting ileostomy may prevent a deadly complication.

	GROUP A (Leak)†	GROUP B (No Leak)†	p value
Preoperative Hb (g/dl)	11.7 (8.8-14.6)	13 (10.7-16.3)	0.05*
Operation time (min)	150 (100-260)	120 (90-180)	0.04*
EBL (ml)	200 (50-620)	100 (50-270)	0.01*

EBL = estimated blood loss *Wilcoxon rank test † median(IQR)