Regional versus metropolitan pancreaticoduodenectomy mortality in Australia

Background: This retrospective, population-based cohort study aims to determine if differences in the regional distribution of procedures or variation in regional mortality contributes to the variable pancreaticoduodenectomy (PD) mortality between Australian states and territories. Methods: De-identified procedural data from public hospitals between 1 July 2005 and 30 June 2015 from the Australian Institute of Health and Welfare were analysed. The regional distribution of procedures and variation in perioperative mortality rate (POMR) were investigated in New South Wales (NSW), Victoria and Queensland (QLD) using logistic regression analysis. Results: NSW performed the highest proportion of city-based procedures (93.8%) while QLD performed the highest proportion of regional procedures (15.3%). QLD demonstrated the lowest city mortality (1.9%) and lowest POMR overall (2.0%). City, regional and statewide mortality was highest in NSW (5.0%, 8.4% and 5.3%). No significant difference in POMR was demonstrated between regional and city hospitals in each of the states (P = 0.46) or across all states (P = 0.50). Conclusion: This study demonstrates comparable regional PD distribution across Australia. The difference in PD POMR between city and regional areas was not found to be statistically significant. NSW exhibited the highest city, regional and overall PD POMR, potentially warranting further investigation.