LAUSR

BACKGROUND The use of risk-stratified pancreatectomy care pathways (RSPCPs) is associated with reduced length-of-stay (LOS). This study sought to evaluate the impact of successive pathway revisions with the hypothesis that high-risk patients require iterative pathway revisions to optimize outcomes. STUDY DESIGN A prospectively maintained database (10/2016-12/2021) was evaluated for pancreatoduodenectomy (PD) patients managed with RSPCPs preoperatively assigned based on pancreatic fistula (POPF) risk. Launched in 10/2016 (Version 1-V1), RSPCPs were optimized in 2/2019 (V2) and 11/2020 (V3). Targeted pathway components included earlier nasogastric tube removal, diet advancement, reduced intravenous fluids and opioids, institution-specific drain fluid amylase cutoffs for early day 3 removal, and patient education. Primary outcome was LOS. Secondary outcomes included major complications (Accordion Grade III+), POPF (ISPGS Grade B/C), and delayed gastric emptying (DGE). RESULTS Of 481 patients, 234 were managed by V1 (83 high-risk), 141 V2 (43 high-risk), and 106 V3 (43 high-risk). Median LOS reduction was greatest in high-risk patients with a 7d reduction (pre-RSPCP-12d, V1-9d, V2-7d, V3-5d), compared to low-risk patients (pre-pathway-10d, V1-6d, V2-5d, V3-4d). Complications decreased significantly among high-risk patients (V1-45%, V2-33%, V3-19%, p< 0.001), approaching rates in low-risk patients (V1-21%, V2-20%, V3-14%). POPF (V1-33%, V2-23%, V3-16%, p< 0.001) and DGE (V1-23%, V2-22%, V3-14%, p< 0.001) improved among high-risk patients. CONCLUSIONS Risk-stratified pancreatectomy care pathways are associated with reduced LOS, major complications, Grade B/C fistula, and DGE. The easiest gains in surgical outcomes are generated from the immediate improvement in the patients most likely to be fast-tracked, but high-risk patients benefit from successive application of the learning health system model.