On-line version ISSN 2078-5135
SAMJ, S. Afr. med. j. vol.98 n.6 Cape Town Jun. 2008
ISSUES IN MEDICINE
Alan D Rothberg
Professor Rothberg is a paediatrician, sub-specialised in neonatology. He was Head of Paediatrics at Witwatersrand University and Deputy Dean of the Faculty until 1994, when he entered the private health funding arena. He returned to the University and Faculty in 2005 as Head of the School of Therapeutic Sciences
Addressing racial profiling in South African medical research,1 Dan Ncayiyana highlighted the debate over the past decade: from the position that data should not routinely be collected according to population group,2 through editorial policies that rejected unwarranted reference to race or use of race as an explanatory variable, to the current position that identifying health disparities of social groups is important. One should be clear as to causation of disparities, and not attribute differences to race or ethnicity when they are the result of social, educational or environmental deprivation. Having justified categorisation, Ncayiyana warns that while the goal of race-based data collection might be to monitor post-apartheid transformation, it risks becoming an affirmation of the validity of race classification rather than a means to erase it.
Race classification in South African health care
The 'Health Charter'3 indicates that we are far from eliminating race from our datasets. Race is entrenched as a variable to be captured and tracked in the quest for transformation and equity, whether dealing with access to health interventions or services, needs of vulnerable groups, health care funding, education and training of professionals, quality of health care, racial composition of the health care industry, or service agreements between parties. Some goals are broadly stated, e.g. the human capital programme aims over the next 15 years to 'address the demographics and diversity of the people being trained', and seeks 'support for initiatives that will increase the number of black people matriculating in science'. Others are specific, e.g. by 2014 the workplace will be 70% black across the value chain, equity ownership in the health care sector will be 51%, and procurement from black-owned firms should increase to 80%.
The Health Charter addresses two major issues: affordable, accessible, appropriate, quality health care for all South Africans, and broad-based black economic empowerment. The latter is supported by legislation.4 While the former broadly involves policy issues, embodied therein is a specific call for action from academic institutions, stating that the speed with which the health sector can be transformed will depend on the pace of their transformation.
Challenges for faculties of health science
Pressures to transform academic medicine have existed since the days of racial quotas and segregated universities, through the years of international academic rejection and isolation, to our democratisation and the challenges of accommodating applicants from differentiated and inferior secondary educational systems. Ironically, student selection for health sciences was a simpler process in early post-apartheid years when demographic targets could be set based on race and gender, and after compensatory adjustments the best applicants could be taken from the various race-based matriculation systems. Current public school inadequacies5 and the introduction of a standardised matriculation examination are likely to frustrate transformation, since simply accepting the best-performing applicants will largely select the historically and recently empowered, and probably not produce the desired demographic profile.
Similarities in the North American and South African health-related race debates are of interest and importance. An Institute of Medicine (IOM) committee tasked with addressing racial and ethnic disparities in health care identified socioeconomic and health insurance status, health systems, and patient and provider attitudes, behaviour and expectations as major contributors.6 Their recommendations echo many of those in the Health Charter,3 e.g. improvements in medical care financing and allocation of care, community-based care, and the need for cross-cultural education to foster provider-patient communication and understanding. In response to this report the Harvard Institute of Health Policy emphasised the role of academic medicine in collecting data and reporting on access and utilisation by race/ethnicity, encouraging the use of evidence-based guidelines, supporting the use of language and interpretation services in the clinical setting, increasing awareness of racial/ethnic disparities in health care, increasing the proportion of under-represented minorities in the health care workforce, integrating cross-cultural education into the training of all health care professionals, and conducting further research into identifying sources of disparities and remedial interventions for them.7
Promotion of diversity and 'cross-cultural sensitisation' feature on our universities' agendas, and language-interpretation services and cross-cultural education are common in health sciences curricula. Paradoxically, while these initiatives cater for university life and clinical training, they tend to leave graduates unprepared for the attitudes, behaviour, expectations and demands of providers and patients in the private sector, which many will enter after qualifying.
Is more always better?
The IOM report6 notes that racial and ethnic minorities tend to receive lower quality health care even when factors such as health insurance status and income are controlled. The sources of these disparities are complex, rooted in past and present inequities, and involve several levels (health systems, health professionals and patients). Disparities were confirmed in several areas including cardiovascular care, cancer screening and treatment, diabetes care, renal transplantation, maternal and child health, and rehabilitative and nursing home services, with minorities experiencing more care in some situations, e.g. bilateral orchidectomy and amputations. It concludes that disparities exist and are unacceptable because they are usually associated with worse outcomes. An example of an apparent clinical benefit of racial disparity is that black patients dialysed for end-stage renal failure have better survival statistics.8 However, black transplant rates are lower,9 perhaps suggesting that the higher mortality of white patients on dialysis is due to a concentration of medically or socially 'untransplantable' patients in this group.
One might question the relevance of the IOM report to the South African situation because the USA relates to minority groups whereas South Africa is concerned about disparate care for the majority of citizens. However, the critical common factor is disempowerment. Ultimately we are all concerned about the quality of care, defined as involving access to care, process/es of care, health care system/s, patient experience, and the outcome of these.10 Racially based health care disparities under apartheid would often have resulted from disempowered blacks having inadequate access in a flawed system that offered sub-optimal care from culturally insensitive providers. Such circumstances would almost certainly compromise outcomes. What we need to know now is how much has changed since the end of apartheid, and are we moving towards quality care and best practice for all? Some early results show that mental health services have not improved, the situation being worse for black patients,11 and that cervical cancer screening is more successful for whites although blacks are at higher risk.12 Reasons for ongoing disparities include differences in socioeconomic status, education, access to services and cultural background. In the all-important area of HIV/AIDS, South African results are likely to be similar to those in the IOM report,6 viz. there are racial disparities in accessing and adhering to antiretroviral treatment, if only because of the substantial personal costs involved13 and white patients being better able to absorb such costs.
While there are many opportunities for research into racial disparities in health care in the public sector, e.g. in non-academic peripheral hospitals previously reserved for whites but now integrated, the private sector also offers a rich source of data through its medical scheme administrators who all capture gigabytes of data according to diagnostic and procedural codes. Published work includes the observation that in some low-cost, predominantly black medical schemes the benefit design provides nurse-based antenatal care in a private sector clinic but offers delivery in an unattached and unrelated private hospital, usually by a specialist who has not seen the patient during the pregnancy. This fragmentation cannot be regarded as quality care or best practice.14 In another study Mathabathe reviewed the frequency of dental procedures and services within medical schemes at opposite ends of the socioeconomic spectrum.15 Extractions and removable dentures were more common in the lower-cost, predominantly black medical schemes, whereas restorative services such as crowns and inlays were more frequent in the higher-cost, predominantly white schemes. The range of benefits was similar for these medical schemes, although members in the higher-cost schemes might have had more to spend on the services. Work is required to establish whether the differences are due to black patients' inability to pay, cultural predisposition to dental extraction rather than conservation, stage of disease at first presentation, or provider attitudes.
Studies of interracial differences in the type and frequency of interventions for ischaemic heart disease in the USA have shown that revascularisation procedures are performed less frequently in blacks than in whites.6 In this country an analysis of total expenditure for several medical and surgical interventions revealed that in the case of ischaemic heart disease, within the same medical schemes and with access to the same benefits, total expenditure per event was ±50% higher for white males between the ages of 45 and 65 (±R32 500 for whites v. ±R20 000 for blacks).16 The question was whether the higher cost represented overuse in one group or underuse in the other. Research from the USA indicates that too few clinically indicated revascularisation procedures are performed in blacks, while whites receive more clinically indicated procedures and many clinically unnecessary ones.17 Overall the data show that racial care disparities usually imply inferior care, even when blacks receive more of a particular intervention.
Ncayiyana concludes that racial profiling is generally offensive, but is legitimate and important in unequal societies, provided one is aware of what is being measured, and when race or ethnicity are proxies for other measures.1 This article proposes that we have an obligation, in some cases statutory, to categorise by race. Racial disparities research in health care must involve academic medicine, identify inferior care, and determine appropriate care and best practice. As with coronary revascularisation procedures, one must establish whether higher cost and utilisation in one group implies better care or a waste of scarce resources,17 or as in dialysis and transplantation,8,9 one must 'tease out' whether the denial of definitive treatment to one group falsely indicates a survival advantage over the other. Are disparities in one group due to financial advantage, better education, computer literacy and internet-derived health information, or is there cultural reluctance in another group to submit to a high-risk intervention? Or are there subconscious biases among providers that favour the use of specific drugs, procedures or devices in one group at the expense of others? These important issues must be addressed now and repeatedly over time because it is change that must be measured. Health systems, patients and providers are all experiencing change as a result of revised health and welfare policies, promotion and advancement of empowerment within previously disadvantaged communities, and a changing medical practitioner profile in terms of gender, race and ethnicity. These factors affect access to care, the processes of care and patients' experiences, and it is hoped will improve clinical outcomes.
1. Ncayiyana DJ. Racial profiling in medical research: what are we measuring? S Afr Med J 2007; 92: 1225-1226. [ Links ]
2. Ellison GT, de Wet T, Ijsselmuiden CB, Richter LM. Desegregating health statistics and health research in South Africa. S Afr Med J 1996; 86: 1257-1262. [ Links ]
7. Betancourt JR, Maina AW. The Institute of Medicine Report 'Unequal Treatment': implications for academic heath centres. Mt Sinai J Med 2004; 71: 314-221. [ Links ]
8. Agodoa L, Eggers P. Racial and ethnic disparities in end-stage kidney failure - survival paradoxes in African-Americans. Semin Dial 2007; 20: 577-585. [ Links ]
9. Stolzmann KL, Bautista LE, Gangnon RG, McElroy JA, Becker BN. Trends in kidney transplantation rates and disparities. J Natl Med Assoc 2007; 99: 923-932. [ Links ]
10. National Quality Measures Clearinghouse. Agency for Health care Research and Quality, www.qualitymeasures.ahrq.gov (accessed 18 February 2008). [ Links ]
11. Kohn R, Szabo CP, Gordon A, Allwood CW. Race and psychiatric services in post-apartheid South Africa: a preliminary study of psychiatrists' perceptions. Int J Soc Psychiatry 2004; 50: 18-24. [ Links ]
12. Bailey RS, Selvey CE, Bourne D, Bradshaw D. Trends in cervical cancer mortality in South Africa. Int J Epidemiol 1996; 25: 488-493. [ Links ]
13. Rosen S, Ketlhapile M, Sanne I, de Silva MB. Cost to patients in obtaining treatments for HIV/AIDS in South Africa. S Afr Med J 2007; 97: 524-529. [ Links ]
14. Rothberg AD, McLeod H. Private-sector caesarean sections in perspective. S Afr Med J 2005; 95: 257-260. [ Links ]
15. Mathabathe N. Trends in oral health among patients on medical aid. MDent Research Report, University of the Witwatersrand, 2002. [ Links ]
16. Black TJ. An investigation into the influence of age, ethnicity and gender on the cost of existing prescribed minimum benefits in South African medical schemes in 2001. BBusSci Research Report, University of Cape Town, 2003. [ Links ]
17. Schneider EC, Leape LL, Weissman JS, Piana RN, Gatsonis C, Epstein AM. Racial differences in cardiac revascularization rates: does 'overuse' explain higher rates among white patients? Ann Intern Med 2001; 135: 328-337. [ Links ]
A D Rothberg