Abstract
Patient‐reported outcome measures (PROMs) are now routinely collected in the English National Health Service and used to compare and reward hospital performance within a high‐powered pay‐for‐performance scheme. However, PROMs are prone to missing data. For example, hospitals often fail to administer the pre‐operative questionnaire at hospital admission, or patients may refuse to participate or fail to return their post‐operative questionnaire. A key concern with missing PROMs is that the individuals with complete information tend to be an unrepresentative sample of patients within each provider and inferences based on the complete cases will be misleading. This study proposes a strategy for addressing missing data in the English PROM survey using multiple imputation techniques and investigates its impact on assessing provider performance. We find that inferences about relative provider performance are sensitive to the assumptions made about the reasons for the missing data. © 2015 The Authors. Health Economics Published by John Wiley & Sons Ltd.
Highlights
Non-response is a major concern in health surveys because individual non-respondents tend to be systematically different from those providing complete data (Cohen and Duffy, 2002; Perneger et al, 2005; Schenker et al, 2006)
The English Patient-reported outcome measures (PROMs) survey provides valuable evidence about the effectiveness of the health care delivered by the National Health Service (NHS) in terms of health improvements to patients
As with other self-reported, voluntary health surveys, PROMs are prone to large proportions of missing data because of reasons that are associated with both the patient and the provider
Summary
Non-response is a major concern in health surveys because individual non-respondents tend to be systematically different from those providing complete data (Cohen and Duffy, 2002; Perneger et al, 2005; Schenker et al, 2006). The reasons for the non-response are rarely completely independent from both observed and unobserved values, meaning that data are not missing completely at random (MCAR). If the differences between respondents and non-respondents can be explained entirely by differences in the observed data, such as characteristics of the patients, data are said to be missing at random (MAR). It is possible to condition analyses on observed factors, thereby correcting the bias caused by missing data. Differences between respondents and non-respondents may depend on unobserved values, in which case data are missing not at random (MNAR). If missingness is associated with unmeasured factors, conditioning on the observed data will not eliminate entirely potential bias
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
