Impact of prior hospital mortality versus surgical volume on mortality following surgery for congenital heart disease

Objective: Our objective was to assess the relationships of a hospital's past adjusted in-hospital mortality and surgical volume with future in-hospital mortality after surgery for congenital heart disease. Methods: Using the Pediatric Health Information Systems database, we (1) calculated hospital surgical volume and standardized mortality ratio (= observed number of deaths/expected number of deaths adjusted for surgery type) for January 2004 through June 2006 for children (0-18 years) after surgery for congenital heart disease at 38 hospitals and (2) assessed the relationship between these values and subsequent mortality during July 2006 through December 2008. We constructed Poisson regression models to estimate risk of mortality, adjusting for age, race, sex, genetic syndrome, insurance type, and surgery type (using the Risk Adjustment in Congenital Heart Surgery method). Results: There were 49,792 hospital encounters during 2004 through 2008 for pediatric patients having surgery for congenital heart disease, with an overall in-hospital mortality of 3.45%. For the 24,112 eligible encounters during July 2006 through December 2008, a hospital's prior standardized mortality ratio was significantly associated with postoperative in-hospital mortality (P < .0001), and a hospital's prior surgical volume had only borderline significance (P = .0792). On stratified analysis, past standardized mortality ratio was associated with mortality for both lower-and higher-risk surgical risk categories (P = .0105 and .0015, respectively). Hospital surgical volume was not significantly associated with mortality for lower-risk categories (P = .4122), but it was borderline significant for higher-risk categories (P = .0678). Conclusions: In this data set, prior hospital surgical volume tended to be associated with improved mortality after higher-risk operations in pediatric patients with congenital heart disease, whereas prior hospital postoperative mortality was significantly associated with mortality across all risk strata of congenital heart surgery. (J Thorac Cardiovasc Surg 2011; 142: 882-6)