Compression versus hand-sewn and stapled anastomosis in colorectal surgery: a systematic review and meta-analysis of randomized controlled trials.

Anastomotic leaks are a feared complication of colorectal resections and novel techniques that have the potential to decrease them are still sought. This study aimed to compare the anastomotic leak rates in patients undergoing compression anastomoses versus hand-sewn or stapled anastomoses. Randomized controlled trials (RCTs) comparing outcomes of compression versus conventional (hand-sewn and stapled) colorectal anastomosis were collected from MEDLINE, Embase and the Cochrane Library. The quality of the RCTs and the potential risk of bias were assessed. Pooled odds ratios (OR) were calculated for categorical outcomes and weighted mean differences for continuous data. Ten RCTs were included, comprising 1969 patients (752 sutured, 225 stapled, and 992 compression anastomoses). Most used the biofragmentable anastomotic ring. There was no significant difference between the two groups in terms of anastomotic leak rates (OR 0.80, 95 % confidence interval (CI) 0.47, 1.37; p = 0.42), stricture (OR 0.54: 95 % CI 0.18, 1.64; p = 0.28) or mortality (OR 0.70; 95 % CI 0.39, 1.26; p = 0.24). Compression anastomosis was associated with an earlier return of bowel function: 1.02 (95 % CI 1.37, 0.66) days earlier (p < 0.001) and a shorter postoperative stay; 1.13 (95 % CI 1.52, 0.74) days shorter (p < 0.001), but significant heterogeneity among studies was observed. There was an increased risk of postoperative bowel obstruction in the compression group (OR 1.87; 95 % CI 1.07, 3.26; p = 0.03). There was no significant difference in wound-related and general complications, or length of surgery. Compression devices do not appear to provide an advantage over conventional techniques in fashioning colorectal anastomoses and are associated with an increased risk of bowel obstruction.