Epidural analgesia for pancreatoduodenectomy: a critical appraisal.

INTRODUCTION: Epidural analgesia has emerged as a commonly applied method to improve pain management and reduce perioperative complications in major abdominal surgery. However, there is no detailed analysis of its efficacy for pancreatic operations. This study compares clinical and economic outcomes after epidural and intravenous analgesia for pancreatoduodenectomy.
MATERIAL AND METHODS: Data for 233 consecutive patients, who underwent pancreatoduodenectomy, were prospectively acquired and retrospectively reviewed at a single institution, pancreato-biliary specialty practice. From October 2001 to February 2007, all patients were offered thoracic epidural analgesia, and those who declined received intravenous analgesia. Perioperative pain management was dictated as an element of a standardized clinical pathway for pancreatic resections. Clinical and economic outcomes were analyzed and compared for epidural analgesia and intravenous analgesia groups.
RESULTS: One hundred eighty-five patients received epidural analgesia, and 48 received intravenous analgesia, with equivalent baseline patient demographics between the groups. Patients administered epidural analgesia had lower pain scores but significantly higher rates of major complications. Pancreatic fistulae and postoperative ileus occurred more frequently, and patients with epidural analgesia more often required discharge to rehabilitation facilities. A trend towards longer hospitalizations was observed among epidural analgesia patients, but total costs were statistically equivalent between the groups. Further analysis demonstrates that 31% of epidural infusions were aborted before anticipated (fourth postoperative day) because of hemodynamic compromise and/or inadequate analgesia. These select patients required more transfusions, aggressive fluid resuscitation, and subsequently suffered even higher rates of gastrointestinal and respiratory complications, all attributing to higher costs. Multivariate analysis demonstrates that preoperative hematocrit concentration less than 36%, elderly age (>75 years), and chronic pancreatitis predict failure of epidural infusions.
CONCLUSION: Thoracic epidural analgesia after pancreatic resections is associated with hemodynamic instability, which may compromise enteric anastomoses, gastrointestinal recovery, and respiratory function. These outcomes are exacerbated in poorly functioning epidurals and suggest that epidural analgesia may not be the optimal method for perioperative pain control when pancreatoduodenectomy is performed.