Regulatory bodies and insurers evaluate hospitals using obstetrical outcomes as quality measures. However, it is unclear which outcomes, or group of outcomes, are meaningful. Our goal was to assess the consistency of hospital performance across four obstetrical outcome measures.
Data were obtained by trained abstractors, with ongoing data edits and audits, from maternal and neonatal charts of all deliveries on 365 randomly selected days at 25 hospitals over a three-year period. Four outcome measures, selected a priori and rigorously defined, were chosen: postpartum hemorrhage (PPH), maternal peripartum infection, perineal trauma (3rd or 4th degree laceration) at spontaneous vaginal delivery (LAC), and a composite adverse neonatal outcome. Outcome rates, adjusted for institutional differences in patient characteristics, were calculated for each hospital, and hospitals were ranked from lowest to highest based on each outcome's frequency. Spearman correlations were calculated to determine risk adjusted hospital performance for each measure.
115,502 deliveries were evaluated. The hospital rates for PPH and LAC were negatively correlated (rho=-.49, p=.01) whereas hospital rates for PPH and infection were positively correlated (rho=.41, p=.04) (Figures
). None of the other outcome measures rates were significantly correlated: neonatal vs. infection (p=.9), vs. LAC (p=.53), vs. PPH (p=.59), or LAC vs. infection (p=.13).
Quality measures for different obstetric outcomes are poorly correlated, and may be inversely correlated in some cases. Therefore, performance based on a single outcome quality measure cannot be generalized. Multiple quality measures should be assessed to gain a better understanding of overall hospital performance.
© 2012 Mosby, Inc. Published by Elsevier Inc. All rights reserved.