Impact of Epidural Failures on the Results of a Prospective, Randomized Trial

We conducted a prospective randomized trial to evaluate the merits of two established postoperative pain management strategies: thoracic epidural (EPI) versus patient-controlled analgesia (PCA) with intravenous narcotics after minimally invasive repair of pectus excavatum. Pain scores favored the EPI group for the first two postoperative days only. Critics of the trial suggest that if the epidural failure rate was not so high, results may have favored the EPI group. Therefore, we performed a subset analysis of the EPI group to evaluate the impact of these failures.

Patients for whom epidural catheter could not be placed or whose catheters were removed early owing to dysfunction were compared with those with well-functioning catheters. Those with well-functioning catheters were also compared with the PCA group. A two-tailed independent Student t-test and a two-tailed Fisher exact test were used where appropriate.

Of 55 patients in the EPI group, 12 patients (21.8%) had failed placement or required early removal. Comparing those with failed placements with the rest of the group, there was no difference in daily visual analogue scale pain scores or measures of hospital course. Likewise, comparing those with well-functioning catheters only to those in the PCA group, the results of the trial are replicated in terms of pain scores, hospital course, and length of stay.

In patients with failed epidural therapy, there is no significant difference in postoperative hospital course. Comparing those with well-functioning catheters to those in the PCA group, trial results are replicated-that is, no significant difference in length of stay, time to regular diet, or time to transition to oral medications. Therefore, failure rate in the EPI group did not influence the results of the trial.