Ethnic minorities and their perceptions of the quality of primary care

Societies are becoming more culturally diverse and this trend—which is driven by a combination of political and economic forces, globalisation, and improved access to travel—is set to continue. More than two decades of research, mainly from English language speaking countries, shows that ethnic minority people tend to have poorer outcomes than the white majority population for a range of diseases.1 Research from the UK has, for example, shown that the risk of an emergency admission among South Asians with asthma is three times higher than the risk in white people, and is twice as high in black individuals.2 Similar inequalities have also been noted with respect to risk of developing diabetes, ischaemic heart disease, and stroke.3 4 5 6 In the linked study (doi:10.1136/bmj.b3450), Mead and Roland assess why patients from ethnic minorities also tend to give lower evaluations of primary health care than white patients.7

The factors underpinning inequalities in health outcomes are complex, multi-faceted, and not well understood, which partly explains the limited progress in reducing them.8 Part of the explanation does, however, seem to lie in differences in the quality of care received, which has been the dominant explanation of findings from previous patient surveys of primary care provision. On the other hand, Mead and Roland’s analysis suggests that some of the differences in ratings of care might be explained by higher expectations in some ethnic minority groups.7

The General Practice Assessment Questionnaire (GPAQ) is a national (English) patient survey for primary care that was undertaken as part of the quality and outcomes framework. The national use of this tool reflects the increasing importance being assigned by policy makers to patient reported outcome measures. Mead and Roland have performed an innovative and potentially important secondary analysis of this dataset, investigating various parameters of care among the four main ethnic groupings in the UK. The major strengths of this study include the fact that the instruments used in the survey allowed assessment both of perceptions of care and also patient reports of actual care received. In addition, a large number of patients (including ethnic minority patients) completed the questionnaire and were included in the sample, and the authors adjusted for a range of potentially important confounding factors.

But despite these strengths, this secondary analysis has some important limitations (many of which the authors acknowledge). These shortcomings include the lack of a clear sampling frame, the resulting difficulties in trying to ascertain the response rate—let alone compare the characteristics of responding and non-responding patients—the failure to validate the instruments across cultures, the crude ethnic groupings used, and the substantial risk of residual confounding. It is also important to highlight that the dataset used did not involve any objective assessment of the actual quality of care delivered.

Given these limitations, one should be cautious about the suggestion that higher expectations of care in ethnic minority groups is a key factor underpinning the differences in perceptions of care. Indeed, the data suggest that, if anything, issues with patient reports of the quality of care are likely to be an important problem. This is particularly evident in Mead and Roland’s data from Chinese and black respondents’ assessments of receptionist and general practitioner communication, but also in data from Asian patients.8 This explanation is further supported by a large body of evidence showing inequalities in various process measures of care such as provision of preventative care, the responsiveness of services, and referrals for specialist assessment and care provision.9 10

That said, the hypothesis about possible ethnic differences in expectations of care is intriguing and warrants further exploration in qualitative studies and descriptive studies using appropriately validated instruments in carefully defined population groups. Any residual differences in perceptions of care that are not adequately explained by structural factors (for example, socioeconomic considerations and the characteristics of the practice and general practitioner), discrimination, or differences in actual quality of care are more likely to be explained by differences in expectations of the types and models of care provided by health services (such as preferences with respect to evening and weekend opening) than by higher expectations.11 This possibility should, therefore, be a consideration for future research.

Given the increased importance assigned to responsiveness and personalisation of care, the key practical message of this work is that practices need to identify and understand the priorities of their communities and tailor care accordingly. The emphasis thus needs to be on judging care providers by their success in meeting the needs of their local populations rather than on adjusting patient evaluations for the ethnic profile of the practice populations.12