Risk-adjusted prolonged length of stay as an alternative outcome measure for pediatric congenital cardiac surgery

Abstract

Background: Morbid events after pediatric congenital cardiac surgery are increasingly used for better outcome measurement and quality comparisons. This study was undertaken to evaluate the relationship between a hospital's risk-adjusted prevalence of prolonged postoperative length of stay (PLOS) and its risk-adjusted mortality rate to investigate whether PLOS can serve as an appropriate quality measure for pediatric congenital cardiac surgery.

Methods: Risk-adjusted prevalence of prolonged PLOS for 12 programs in New York State was estimated using data from 4,776 operations in the New York State pediatric Cardiac Surgery Reporting System (2006-2009). We used logistic regression analysis to adjust for case mix and patient risk factors. The hospital-level correlation between risk-adjusted prolonged PLOS and risk-adjusted mortality rates was examined using Spearman correlation coefficient analysis.

Results: Risk-adjusted prevalence of prolonged PLOS ranged from 7.48% to 36.52% for pediatric cardiac programs in New York State during the study period. The Spearman correlation test showed a strong positive relationship between a hospital's risk-adjusted prolonged PLOS and mortality rate (r = 0.83; p = 0.0008).

Conclusions: Prolonged PLOS can be used in lieu of risk-adjusted mortality rates when it is not practical to use mortality rates owing to low case volume or decreasing mortality rates of some procedures.