Comparing hospital quality performance estimates based on different patient‐reported outcome measures

Abstract

Generic as well as disease‐specific PROMs have been collected by hospital providers in the English National Health Service (NHS) since April 2009 for four elective procedures: hip and knee replacement, varicose vein surgery and hernia repair. These measures provide information about the self assessment of patients’ health status. The aim of this study is to compare the provision of health services in the NHS according to the different patient reported outcome (PRO) measures and to analyse whether our judgement about hospital performance depends on the choice of PRO instrument. In order to do that we carry out a literature review searching for papers that make direct comparisons between the generic, EQ‐5D or EQ‐VAS, and specific measures OHS, OKS or AVVQ. In addition, we estimate fixed effects models for 20,509 patients treated in 153 hospitals who have completed the questionnaires before and after hip replacement. This methodology will allow for the analysis of hospital effects, i.e. the variation in the measures at provider level. The results show a high positive correlation (over 0.70) between the measures indicating that in general high/low scores in one of the measures are associated to high/low scores in the other measure used for comparisons. However, there are some hospitals judged outliers according to one measure but not the other one.