The impact of patient choice of provider on equity: analysis of a patient survey.

Objectives: To understand the impact on equity of giving patients a choice of provider.

Methods: A postal survey of 5997 patients in four areas of England about choice at their recent referral and, using a discrete choice experiment, how they would choose in hypothetical situations. Binary logistic regression and a series of multinomial and nested logit models were used to analyse the data to discover whether patients with particular characteristics were more likely to: think choice is important; be offered a choice; and, choose a non-local provider.

Results: The response rate was 36%. Choice was more important to older patients aged 51-80 years, patients from non-white backgrounds, women, those with no qualifications and those with a bad past experience of their local hospital. There were no significant differences in who was offered a choice in terms of education, age group or ethnicity. In both real and hypothetical situations patients with no formal qualifications and those living in urban centres were more likely to choose their local hospital, and patients with a bad or mixed past experience at the local hospital were more likely to choose an alternative. In hypothetical choices those who do not normally travel by car and without Internet access were more likely to choose their local hospital irrespective of that hospital's characteristics.

Conclusions: More educated, affluent patients were no more likely to be offered a choice than other population groups, but there does appear to be a social gradient in who chose to travel beyond the local area for treatment. If these results were replicated across England, there is at least the potential risk that when local hospitals are failing, patient choice could result in inequitable access to high quality care, rather than enhancing equity as the policy's architects had hoped.