Heterogeneous association of health with patient and general practice characteristics by region, age and chronic condition: pooled cross-sectional study of patient-level data from England.

OBJECTIVES

To investigate the association of individual patient health with their characteristics, including income deprivation, ethnicity and gender, with the size, quality and staffing of their general practices, and how these associations and income-related health inequality vary across chronic conditions, regions and age bands.

DESIGN

Using observational pooled cross-sectional individual patient survey data linked with data on general practice clinical quality and staffing and deprivation at Lower-Layer Super Output Area level. Ordinary least-squares multiple regression models of patient health estimated on the full sample and on eight condition-specific, nine region-specific and six age-specific subsamples. Three concentration indices embodying different value judgements summarise income-related health inequality in the full sample and subsamples.

SETTING

Primary care in England.

PARTICIPANTS

Over 1 million adult patients in 6426 general practices in 2015/2016 and 2016/2017.

PRIMARY OUTCOME MEASURES

Patient-reported health (the 5-level EQ-5D version or EQ-5D-5L).

RESULTS

Patients who are younger, male, more satisfied with their practice, have fewer chronic conditions and live in less-income or education-deprived areas report better health. White ethnicity is associated with worse health up to age 64 and better health from age 65, with better health in five of the eight chronic condition samples, and in the regional samples except for London and Yorkshire and Humber regions. Practice clinical quality is positively associated with health in the full sample but only in 4 of the 23 subsamples. Income-related health inequality is worst for patients with a mental health problem, residents in the northwest and northeast regions and is greatest for those aged 55-64. The three concentration indices are highly positively correlated across chronic condition and age-band samples. One index has a much weaker correlation relationship with the other two indices in the region-specific samples.

CONCLUSION

Income-related health inequality and the associations of health with patient and practice characteristics are heterogeneous by patients' chronic condition, age and region.