BACKGROUND: Nonsteroidal anti-inflammatory drugs have been associated with anastomotic leakage. It was studied if diclofenac affects anastomoses differently depending on the location in the gut. METHODS: Ninety-five rats were randomized to 6 groups with an anastomosis in either ileum (IL), proximal colon (PC), or distal colon (DC). Groups IL + (n = 10), PC + (n = 30), and DC + (n = 10) received diclofenac (3 mg/kg/day) from day 0 until sacrifice on day 3. Group PC-- (n = 15) did not receive diclofenac. Groups PC1 + and PC2+ (n = 15 each) were given diclofenac from day 1 to 4 and from day 2 to 5. RESULTS: Leak rates were 10/10 in group IL +, 22/30 in PC +, 1/10 in DC +, and 1/15 in PC --. Delayed administration of diclofenac by 1 or 2 days (6/15, P = .05) resulted in reduced leakage rates. Mechanical strength results corresponded with leak rates. CONCLUSIONS: Diclofenac causes leakage of anastomoses in rat IL and PC, but not in the DC. This suggests a role for the ileal and proximal colonic content in diclofenac-induced leakage. [ABSTRACT FROM AUTHOR]