The Impact of an Inpatient Pancreatitis Service and Educational Intervention Program on the Outcome of Acute Pancreatitis

BACKGROUND We introduced an inpatient pancreatitis consultative service aimed to 1) provide guideline-based recommendations to acute pancreatitis inpatients and 2) educate inpatient teams on best practices for acute pancreatitis management. We assessed the impact of pancreatitis service on acute pancreatitis outcomes. METHODS Inpatients with acute pancreatitis (2008-2018) were included in this cohort study. Primary outcomes included length of stay and refeeding time. The educational intervention was a guideline-based decision support tool, reinforced at hospital-wide educational forums. In Part A (n = 965), we compared outcomes pre-service (2008-2010) to post-service (2012-2018), excluding 2011, when the pancreatitis service was introduced. In Part B (n = 720, 2012-2018), we divided patients into 2 groups based on if co-managed with the pancreatitis service, and compared outcomes, including subgroup analysis based on severity, focusing on mild acute pancreatitis. RESULTS In Part A, for mild acute pancreatitis, length of stay (111 vs 88.4 h, P = .001), refeeding time (61.8 vs 47.4 h, P = .002), and infections (10.0% vs 1.87%, P < .001) were significantly improved after the pancreatitis service was introduced, with multivariable analysis showing reduced length of stay (odds ratio 0.83; 95% confidence interval, 0.82-0.84; P < .001) and refeeding time (odds ratio 0.75; 95% confidence interval, 0.74-0.77; P < .001). In Part B, for mild acute pancreatitis, refeeding time (44.2 vs 50.3 h, P = .123) and infections (5.58% vs 4.70%, P = .80) were similar in patients cared for without and with the service. Length of stay was higher in the pancreatitis service group (93.3 vs 81.2 h, P = .05), as they saw more gallstone acute pancreatitis patients who had greater length of stay and magnetic resonance cholangiopancreatography. In the post-service period, a majority of patients with moderate/severe acute pancreatitis and nearly all intensive care unit admits received care from the pancreatitis service. CONCLUSIONS Implementation of an inpatient pancreatitis service was associated with improved outcomes in mild acute pancreatitis. Guideline-based educational interventions have a beneficial impact on management of mild acute pancreatitis by admitting teams even without pancreatitis consultation.