A decade of mortality reductions in major oncologic surgery: the impact of centralization and quality improvement

For more than a decade, health policy groups have recommended concentrating care for certain high-risk oncologic procedures into high-volume centers. The degree to which practice patterns and outcomes have changed over that time period is unclear.

To evaluate temporal trends in the mortality and concentration of high-risk oncologic procedures.

Retrospective cohort study using data from the Nationwide Inpatient Sample on 93,108 adult patients undergoing pancreatectomy, esophagectomy, gastrectomy, or major lung resection for organ-specific cancers from 1997 to 2006. The main outcome measure was in-hospital mortality.

Risk- and volume-adjusted mortality decreased over time for all 4 procedures (P < 0.05). When hospitals were categorized into terciles of case volume, mortality gaps persisted between high- and low-volume centers in all procedures throughout the study period and did not significantly narrow over time. Patient volumes shifted toward high-volume centers over time for all procedures (P <0.001), although at the end of the study period, low-volume centers still cared for one-quarter to one-third of patients undergoing each of the studied procedures. Most of the overall improvements in mortality were attributable to decreasing mortality within volume categories as opposed to the effects of care concentration.

Modest concentration into higher-volume centers has taken place over the past decade, but improvements within volume categories have driven most of the generalized decreases in mortality. Significant outcome gaps between high- and low-volume centers still persist; further centralization may retain the potential to lower overall mortality.