An analysis of publicly reported pediatric heart surgery data and patient mortality implications

J. Matthew Austin, Jordan M. Derk, Peter J. Pronovost


Background: An association exists between higherhospital volumes and better patient outcomes for high-risk pediatric andcongenital heart surgery. This relationship suggests that patient deaths couldpotentially be avoided if the highest complexity operations performed in low-and medium-volume hospitals were shifted to high-volume hospitals. Usingpublicly available data, this study investigates the number of deaths ofpediatric and congenital heart surgery patients in the USA that theoreticallycould be avoided if highest complexity patients had surgery at high-volumehospitals.
Methods: Data were extracted from The Societyof Thoracic Surgeons website for 61 hospitals that voluntarily publiclyreported their outcomes. Each hospital was categorized based on their meanannual volume of pediatric and congenital heart surgeries. For each volume andpatient risk category combination, we calculated an observed-to-expectedmortality ratio. Using high-volume hospitals as the reference, we calculatedthe theoretical difference in number of deaths in medium- and low-volumehospitals if they had the same mortality performance as high-volume hospitals.
Results: Over the 4-year reporting period, 104deaths (overall 26% reduction in observed deaths) theoretically might have beenprevented if higher-risk operations done at low- and medium-volume hospitalswere performed with outcomes comparable to those of high-volume hospitals.
Conclusions: This analysisidentified a large relative risk reduction in mortality of high-risk pediatricand congenital heart surgery patients that theoretically could be achieved ifhigher-risk operations performed at low- and medium-volume hospitals couldshift to high-volume hospitals. A number of potential solutions exist to reducethese potentially preventable deaths, including regionalization.