Centralisatie pancreaschirurgie reduceert postoperatieve sterfte in Nederland*

OBJECTIVE: To analyse the extent of centralisation of pancreaticoduodenectomy (Whipple procedure) and changes in in-hospital mortality rates in the Netherlands. DESIGN: Retrospective analysis. METHODS: Data on patients who had undergone pancreaticoduodenectomy (PD) during the 2004-2009 period was acquired from the Kiwa Prismant registry. Based on the number of procedures performed annually, hospitals were divided into 4 volume-categories: very-low (<5), low (5-10), medium (11-19) and high (>/=20). Changes in volume and in-hospital mortality were analysed per volume category. A subgroup analysis based on age was also performed. RESULTS: 2155 patients who had undergone PD were included. The number of hospitals performing PD decreased from 48 in 2004 to 30 in 2009 (p = 0.01). The proportion of patients who had undergone PD in a medium- or high-volume hospital increased from 52.9% to 91.2% (p < 0.001). Post-operative mortality rates decreased from 9.8% to 5.1% (p = 0.04). Average mortality was 14.7%, 9.8%, 6.3% and 3.3% in very low-, low-, medium-, and high-volume hospitals, respectively (p < 0.001). The difference in mortality between medium- and high-volume hospitals was statistically significant (p = 0.004). The mortality rate in patients >/= 70 years was 10.4% compared with 4.4% in younger patients (p < 0.001). CONCLUSION : Nationwide centralisation of PD is occurring in Netherlands, and this is associated with a decrease in in-hospital mortality. Further centralisation is likely to further decrease in-hospital mortality, especially in the elderly.