Six RCTs with a total of 458 patients were included in the review. Five RCTs were included in the meta-analyses (n=355 with adequate data)
In terms of study quality, 4 studies reported prior power analysis, two reported adequate allocation concealment, five were at least double-blinded and three used intention-to-treat analysis. One study reported no exclusions from analysis; rates of exclusion from analysis in the other 5 studies ranged from 17 to 23.6%.
Acute pain at rest at 24 hours post-surgery (5 studies).
There was no significant difference between interventions in acute pain at rest at 24 hours (WMD -0.27, 95% CI: -0.91, 0.37; favours preemptive TEA). A sixth study, which could not be pooled, also showed no significant difference between the groups. Statistically significant heterogeneity between the studies was detected (I-squared 98.6%). The results of the analysis were not changed by the exclusion of a study in which no opioid was used.
Acute pain at rest at 48 hours post-surgery (5 studies).
There was a significant reduction in acute pain in the preemptive TEA group (WMD -0.59, 95% CI: -1.14, -0.04). Statistically significant heterogeneity between the studies was detected (I-squared 98.2%). The sensitivity analysis, which excluded a study in which no opioid was used, showed no significant difference between the groups (WMD -0.08, 95% CI: -0.30, 0.14).
Acute post-operative pain on coughing (2 studies).
There was a significant reduction in pain on coughing in the preemptive TEA group at 24 hours (WMD -1.17, 95% CI: -1.50,-0.83) and at 48 hours (WMD -1.08, 95% CI: -1.17, -0.99). No statistically significant heterogeneity was detected.
Chronic pain at 6 months post-surgery (3 studies).
The pooled RR showed no statistically significant difference between the groups (RR 1.32, 95% CI: 0.76, 2.30). Statistically significant heterogeneity between the studies was detected (I-squared 73.6%). The results of the analysis were not changed by the exclusion of a study in which no opioid was used.
Adverse events (4 studies).
There were few adverse events and no epidural catheter-related complications.