Poor Analgesic Efficacy of Epidural Analgesia in Critical Care Patients After Pancreaticoduodenectomy

Objectives The objective of this study was to investigate the analgesic efficacy of functional and prematurely aborted epidurals after pancreaticoduodenectomy in critical care, as this is unknown. Methods Data from elective pancreaticoduodenectomy recipients admitted to the critical care unit over 44 months were prospectively collected. Epidural (0.1% bupivacaine and 2 mu g/mL fentanyl) analgesic efficacy was assessed with a ranked categorical verbal pain score (primary end point). If no epidural was placed, intravenous (IV) fentanyl patient-controlled analgesia (PCA) was used. Results Eighty-six pancreaticoduodenectomy patients had a mean age of 66.5 years; 61.6% were men; and 73 received an epidural, whereas 13 received an IV PCA. Epidural abortion rate was 42.5%, associated with a higher 24-hour (P = 0.02) but not 48-hour pain score. Overall, fewer patients reported any pain (P = 0.010; number needed to harm, 3.2; 95% confidence interval, 1.7-3.2) or severe pain (P = 0.006; number needed to harm, 2.9; 95% confidence interval, 2.1-4.7) with functional epidurals. Pain (sensitivity, 93.8%) and severe pain (specificity, 87.8%) were predictive of epidural abortion. Most postepidural analgesia was IV PCA (P = 0.097) after both functional and aborted epidurals. Conclusions Premature epidural abortion rate was high and associated with analgesic morbidity. Pain score was predictive of epidural abortion. Thus, preference toward epidural analgesia cannot be supported.