Aaditya Narendra1, Peter D Baade2, Joanne F Aitken3, Jonathan Fawcett4, Bernard Mark Smithers5. 1. University of Queensland, Princess Alexandra Hospital, Burke Street Centre, Level 1, B2, 2 Burke Street, Woolloongabba, QLD, 4102, Australia. Electronic address: narendra@tcd.ie. 2. Cancer Council Queensland, 553 Gregory Terrace, Fortitude Valley, QLD, 4006, Australia. 3. University of Queensland, University of Southern Queensland, Cancer Council Queensland, 553 Gregory Terrace, Fortitude Valley, QLD, 4006, Australia. 4. University of Queensland, Hepato-pancreatico-biliary Unit, Princess Alexandra Hospital, 199 Ipswich Rd, Woolloongabba, QLD, 4102, Australia. 5. Cancer Alliance Queensland, University of Queensland, Upper-GI, Soft Tissue and Melanoma Unit, Princess Alexandra Hospital, 199 Ipswich Rd, Woolloongabba, QLD, 4102, Australia.
Abstract
BACKGROUND: An association between higher hospital-volume and better "quality of surgery" and long-term survival has not been reported following pancreatic cancer surgery in low resection-volume regions such as in Australia. Using a population-level study, we compare "quality of surgery" and two-year survival following pancreaticoduodenectomy between Australian hospitals grouped by resection-volume. METHODS: Data on all patients undergoing pancreaticoduodenectomy for adenocarcinoma in the Australian state of Queensland, between 2001 and 2015, were obtained from the Queensland Oncology Repository. Hospitals were grouped into high (≥6 resections annually) and low (<6) volume centres. Following adjustment for case-mix, "quality-of-treatment" indicators were compared between hospital groups using multivariate logistic regression and Poisson regression analysis; and two-year cancer-specific and overall survival were compared using multivariate Cox proportional hazard models. RESULTS: Compared with high-volume centres, low-volume centres had worse two-year cancer-specific survival (Adjusted HR = 1.31; 95% CI:1.03-1.68), higher 30-day mortality (Adjusted IRR = 3.81; 95% CI: 1.36-10.62) and fewer patients received "high-quality surgery" (Adjusted OR = 0.55; 95% CI: 0.33-0.90). Differences in 30-day mortality, or "quality-of-treatment" indicators did not entirely explain the observed survival difference between hospital-volume groups. CONCLUSION: In an Australian environment, a "high" hospital-volume was significantly associated with better quality surgery and two-year survival following pancreaticoduodenectomy.
BACKGROUND: An association between higher hospital-volume and better "quality of surgery" and long-term survival has not been reported following pancreatic cancer surgery in low resection-volume regions such as in Australia. Using a population-level study, we compare "quality of surgery" and two-year survival following pancreaticoduodenectomy between Australian hospitals grouped by resection-volume. METHODS: Data on all patients undergoing pancreaticoduodenectomy for adenocarcinoma in the Australian state of Queensland, between 2001 and 2015, were obtained from the Queensland Oncology Repository. Hospitals were grouped into high (≥6 resections annually) and low (<6) volume centres. Following adjustment for case-mix, "quality-of-treatment" indicators were compared between hospital groups using multivariate logistic regression and Poisson regression analysis; and two-year cancer-specific and overall survival were compared using multivariate Cox proportional hazard models. RESULTS: Compared with high-volume centres, low-volume centres had worse two-year cancer-specific survival (Adjusted HR = 1.31; 95% CI:1.03-1.68), higher 30-day mortality (Adjusted IRR = 3.81; 95% CI: 1.36-10.62) and fewer patients received "high-quality surgery" (Adjusted OR = 0.55; 95% CI: 0.33-0.90). Differences in 30-day mortality, or "quality-of-treatment" indicators did not entirely explain the observed survival difference between hospital-volume groups. CONCLUSION: In an Australian environment, a "high" hospital-volume was significantly associated with better quality surgery and two-year survival following pancreaticoduodenectomy.