Effect of the public release of performance information on patient numbers in the English NHS

• Main Author: Laverty, Anthony
• Other Authors: Millett, Christopher; Aylin, Paul; Smith, Peter
• Published: Imperial College London 2014
• Subjects: 614
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.656589

Introduction: This thesis examines the impact of two special cases of release of quality information on patient utilisation: three NHS trusts heavily publicised as being poor providers of care (the scandal trusts) and NHS trusts highlighted as being the best or worst places for maternity care (the maternity trusts). It also draws on analyses of patient surveys of the information used in making healthcare choices, and whether patient reported quality improved in the maternity trusts. Methods: (1) Analysis of a survey of 2,181 patients recently referred for an outpatient appointment. Logistic regression examined socio-demographic factors associated with information use, as well as the likelihood of attending the 'local' hospital for treatment. (2) Difference-in-difference analysis of patient numbers between the scandal and maternity trusts and comparison groups using data from Hospital Episode Statistics. (3) Analysis of three patient reported quality measures from two surveys of maternity patients. Results: High-profile reports into the quality of care in the scandal trusts had an impact on patient utilisation for only one out of the three trusts, which disappeared six months after report publication. In the maternity trusts there was no change in patient utilisation in trusts highlighted as the best or the worst providers of care. On two out of three measures of patient reported quality trusts publicised as the worst providers did not improve at a faster rate than providers with similar scores at baseline. Discussion: Neither the scandal nor maternity trusts experienced declines in levels of patient utilisation. This lack of effect from high-profile reports, which were unequivocal on the quality of care, casts doubt on whether other forms of information reporting will have an impact on patient utilisation. The limited findings on patient reported quality suggest that in the absence of changes in patient utilisation, reporting on specific clinical areas is not associated with improvements.