Short-term outcomes of pancreaticoduodenectomy in the state of Victoria: hospital resources are more important than volume.

ABSTRACT

BACKGROUND: Pancreaticoduodenectomy (PD) is a high-risk procedure. Australian hospitals perform a relatively low volume of PD. This study sought to gain an understanding of hospital volume and short-term outcomes of the procedure in the Australian state of Victoria. METHODS: The Dr Foster Quality Investigator tool was used to interrogate the Victorian Admitted Episodes Database for the Australian Classification of Health Intervention code for PD (30584) from July 2010 to June 2016. The data set included patients from a peer group of 14 hospitals that included all the public hospitals performing PD during this period. Patient characteristics, inpatient mortality, 30-day readmission rates and median length of stay were reported for each de-identified hospital. RESULTS: There were 547 PD conducted over 6 years in 10 public hospitals. The median patient age was 65 years. Inpatient mortality was 2.7%. There was a significant risk adjusted difference in mortality between principal referral and other public hospitals. Annual hospital volume ranged from 3 to 20 PD, and there was no significant relationship between mortality, readmission rates or length of stay and hospital volume. CONCLUSION: The inpatient mortality associated with PD in Victorian public hospitals is comparable to that seen in overseas studies. While hospital volume is relatively low, there does not seem to be a relationship between volume and short-term outcomes. Variability between hospital peer groups suggests that resource availability is more important than volume. The development of a procedure specific registry would be useful to test the outcomes of this study and determine long-term PD outcomes.