Online Program Home
My Program

Abstract Details

Activity Number: 583 - Statistical Methods in Health Services and Performance Profiling
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 2:00 PM to 3:50 PM
Sponsor: Health Policy Statistics Section
Abstract #330564 Presentation
Title: Bayesian Reliability Assessment of Facility-Level Patient Outcome Measures
Author(s): Jianghua He* and Nancy Dunton
Companies: University of Kansas Medical Center and University of Kansas Medical Center
Keywords: Intraclass correlation coefficient; Emperical Bayesian Statistics; Quality of patient care

Patient health outcome measures at facility-level are often used as quality indicators of patient care. Within-facility variations of such measures often differ among facilities, for which the intra-class correlation coefficient based on equal within-subject variation may not be applicable. Signal-to-noise approach can be used to assess the facility-specific reliability of a measure with different within-subject variation among facilities. Previous studies showed that the reliability score provided by the signal-to-noise approach had a poor correlation with facility ranking based on the outcome measure. In this study, we propose a modified signal-to-noise approach within the Bayesian framework for assessing the reliability of patient outcome measures at facility-level. Simulation studies demonstrated that our approach provides a reliability measure that has a much stronger correlation with facility ranking compared to previous approaches. Our approach is applied to assess the reliability of nursing quality indicators from the National Database of Nursing Quality Indicators.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program