Association between thoracic epidural analgesia and gastrointestinal failure in acute pancreatitis: a prospective exploratory cohort study in Vietnam

Abstract Severe acute pancreatitis is characterized by profound inflammation and splanchnic hypoperfusion, which is often exacerbated by opioid-induced bowel dysfunction. We aimed to evaluate the association between thoracic epidural analgesia (TEA) and early gastrointestinal failure (GIF) in this population. In this single-center observational cohort study, patients with severe and predicted severe acute pancreatitis who received either TEA or systemic analgesia (non-TEA cohort) were prospectively evaluated. The primary outcome was the rate of Day 3 GIF (GIF score ≥ 3). The secondary outcome was the cumulative 72-hour systemic fentanyl consumption (mg). Multivariable logistic regression models were used to assess the independent associations of both TEA and opioid consumption with early GIF. A total of 66 patients were included (TEA cohort: n = 32; non-TEA cohort: n = 34). Baseline characteristics, including age, sex and disease severity scores, were comparable between the cohorts. At ICU admission, GIF was highly prevalent, affecting 87.5% of the TEA cohort and 82.4% of the non-TEA cohort. By Day 3, the unadjusted rate of GIF was significantly lower in the TEA cohort compared to the non-TEA cohort (9.4% vs. 32.4%, p = 0.034). However, this association did not reach statistical significance after adjusting for baseline severity (adjusted Odds Ratio: 0.258; 95% CI, 0.062–1.063; p = 0.061). Furthermore, patients receiving TEA consumed significantly less systemic fentanyl over the first 72 h. In the multivariable regression model, cumulative fentanyl consumption was independently associated with an increased risk of Day 3 GIF (adjusted Odds Ratio: 2.17 per 1 mg increase, 95% CI, 1.33–3.54; p = 0.002). In this exploratory cohort, TEA demonstrated no major procedure-related complications. Although its independent association with early resolution of gastrointestinal failure did not reach statistical significance after adjustment, TEA may relate to clinical benefits partly through significant reductions in systemic opioid consumption.