Post-Pancreatoduodenectomy Outcomes and Epidural Analgesia: A 5-Year Single Institution Experience

Abstract

Introduction Optimal pain control post-pancreatoduodenectomy is a challenge. Epidural analgesia (EDA) is increasingly utilized despite inherent risks and unclear effects on outcomes. Methods All pancreatoduodenectomies (PD) performed from 1/2013-12/2017 were included. Clinical parameters were obtained from retrospective review of a prospective clinical database, the ACS NSQIP prospective institutional database and medical record review. Chi-Square/Fisher’s Exact and Independent-Samples t-Tests were used for univariable analyses; multivariable regression (MVR) was performed. Results 671 consecutive PD from a single institution were included (429 EDA, 242 non-EDA). On univariable analysis, EDA patients experienced significantly less wound disruption (0.2% vs. 2.1%), unplanned intubation (3.0% vs. 7.9%), pulmonary embolism (0.5% vs. 2.5%), mechanical-ventilation >48hrs (2.1% vs. 7.9%), septic shock (2.6% vs. 5.8%), and lower pain scores. On MVR accounting for baseline group differences (gender, hypertension, pre-operative transfusion, labs, approach, pancreatic duct size), EDA was associated with less superficial wound infections (OR 0.34; CI 0.14-0.83; P=0.017), unplanned intubations (OR 0.36; CI 0.14-0.88; P=0.024), mechanical ventilation >48 hrs (OR 0.22; CI 0.08-0.62; P=0.004), and septic shock (OR 0.39; CI 0.15-1.00; P=0.050). EDA improved pain scores post-PD days 1-3 (P<0.001). No differences were seen in cardiac or renal complications; pancreatic fistula (B+C) or delayed gastric emptying; 30/90-day mortality; length of stay, readmission, discharge destination, or unplanned reoperation. Conclusion Based on the largest single institution series published to date, our data support the use of EDA for optimization of pain control. More importantly, our data document that EDA significantly improved infectious and pulmonary complications.