Yan Feng1, Hugh Gravelle2. 1. Institute of Population Health Sciences, Queen Mary University of London, England, UK. Electronic address: yan.feng@qmul.ac.uk. 2. Centre for Health Economics, University of York, England, UK.
Abstract
OBJECTIVES: To examine the association of self-reported health of patients in general practices, as measured by the EQ-5D-5L, with practice clinical quality and patient-reported satisfaction with accessibility and consultations. METHODS: We used data from the General Practitioner (GP) Patient Survey to construct a practice-level EQ-5D-5L index as the health outcome. Key explanatories were patient-reported measures of satisfaction with access and consultations (also derived from the GP Patient Survey) and clinical quality measured by the achievement of clinical quality indicators reported in the Quality and Outcomes Framework. We estimated practice-level linear panel data models with random and fixed practice effects and practice and patient covariates using 2012/13 to 2016/17 data on more than 7500 English general practices. RESULTS: Bivariate correlations of the EQ-5D-5L index with quality measures were 0.048 for clinical quality, 0.071 for satisfaction with access, and 0.107 for satisfaction with GP consultations (all with P<.001). In both fixed effects regressions, which allow for unobserved time invariant practice characteristics, and random effects regressions which do not, the EQ-5D-5L index was positively associated with 1-year lags of patient satisfaction with access and GP consultations. Patient-reported health was positively associated with clinical quality in the fixed effects regressions. The implied effects were small in all cases. CONCLUSION: Practice-level EQ-5D-5L is positively associated with clinical quality and with 1-year lags of patient-reported satisfaction with access and GP consultations.
OBJECTIVES: To examine the association of self-reported health of patients in general practices, as measured by the EQ-5D-5L, with practice clinical quality and patient-reported satisfaction with accessibility and consultations. METHODS: We used data from the General Practitioner (GP) Patient Survey to construct a practice-level EQ-5D-5L index as the health outcome. Key explanatories were patient-reported measures of satisfaction with access and consultations (also derived from the GP Patient Survey) and clinical quality measured by the achievement of clinical quality indicators reported in the Quality and Outcomes Framework. We estimated practice-level linear panel data models with random and fixed practice effects and practice and patient covariates using 2012/13 to 2016/17 data on more than 7500 English general practices. RESULTS: Bivariate correlations of the EQ-5D-5L index with quality measures were 0.048 for clinical quality, 0.071 for satisfaction with access, and 0.107 for satisfaction with GP consultations (all with P<.001). In both fixed effects regressions, which allow for unobserved time invariant practice characteristics, and random effects regressions which do not, the EQ-5D-5L index was positively associated with 1-year lags of patient satisfaction with access and GP consultations. Patient-reported health was positively associated with clinical quality in the fixed effects regressions. The implied effects were small in all cases. CONCLUSION: Practice-level EQ-5D-5L is positively associated with clinical quality and with 1-year lags of patient-reported satisfaction with access and GP consultations.