Thoracic Epidural Analgesia and Mortality in Acute Pancreatitis: A Multicenter Propensity Analysis

OBJECTIVE: Recent preclinical and clinical data suggest that thoracic epidural analgesia, a technique primarily aimed at decreasing pain, might exert anti-inflammatory effects, enhance splanchnic and pancreatic blood flow during acute pancreatitis; however, the influence of epidural analgesia on mortality remains under investigated in this setting. This study was therefore designed to assess the impact of epidural analgesia on mortality in ICU patients with acute pancreatitis. DESIGN: Multicenter retrospective, observational, cohort study. SETTING: Seventeen French and Belgian ICUs. PATIENTS: All patients admitted to with acute pancreatitis between June 2009 and March 2014. INTERVENTIONS: The primary exposure was thoracic epidural analgesia versus standard care without epidural analgesia. The primary outcome was 30-day mortality. Propensity analyses were used to control for bias in treatment assignment and prognostic imbalances. MEASUREMENTS AND MAIN RESULTS: One thousand three ICU patients with acute pancreatitis were enrolled, of whom 212 died within 30 days. Epidural analgesia was used in 46 patients and was associated with reduced mortality in unadjusted analyses (4% vs. 22%; p = 0.003). After adjustment for baseline variables associated with mortality, epidural analgesia was still an independent predictor of 30-day mortality (adjusted odds ratio, 0.10; [95% CI, 0.02-0.49]; p = 0.004). Using propensity score analysis, the risk of all-cause 30-day mortality in patients with acute pancreatitis receiving epidural analgesia was significantly lower than that in matched patients who did not receive epidural analgesia (2% vs. 17%; p = 0.01). CONCLUSIONS: Among critically ill patients with acute pancreatitis, mortality at 30 days was lower in patients who received epidural analgesia than in comparable patients who did not. These findings support ongoing research on the use of epidural analgesia as a therapeutic intervention in acute pancreatitis.