Epidural analgesia and risk of anastomotic leakage

Based on case reports of early anastomotic leakage in patients receiving epidural analgesia with local anesthetic and data to document a stimulatory effect of epidural block on gastrointestinal motility, it has been suggested that continuous infusion of epidural local anesthetic may lead to an increased incidence of anastomotic leakage. Therefore, we examined the association between continuous epidural local anesthetic and anastomotic leakage by reviewing the literature.

Review of controlled, randomized clinical trials aiming to investigate postoperative complications in which continuous postoperative epidural local anesthetic was administered in patients scheduled for colorectal surgery with an anastomosis. Data were obtained from a Medline search (1966-May 2000), previous review articles, references cited in original papers, and personal communication with investigators. Twelve trials including a total of 562 patients met the inclusion criteria.

Sixteen of 266 patients (6.0%, 95% confidence interval [CI]: 3.5% to 9.6%) receiving postoperative epidural local anesthetic or epidural local anesthetic-opioid mixtures developed anastomotic leakage, compared with 10 of 296 patients (3.4%, 95% CI: 1.6% to 6.1%) receiving epidural or systemic opioid-based analgesia (P >.05 between groups, Fisher's test). The risk of overlooking a significant difference (type II error) was approximately 67% (power: 33%). Studies including more than 1,037 patients in each group are needed to demonstrate an increased risk of anastomotic leakage from 3.4% to 6.0% with 80% power and 2alpha = 0.05. There was no significant difference (P >.05 between groups, Fisher's test) between subgroups of study design: Epidural local anesthetic-versus systemic or epidural opioid, or epidural local anesthetic-opioid mixtures versus systemic or epidural opioid.

So far, there is no statistically significant evidence from randomized trials to indicate epidural analgesia with local anesthetic to be associated with an increased risk of anastomotic breakdown. However, relatively few patients have been included in randomized trials, indicating a need for more studies to secure valid conclusions.