Tackling inequalities in health

Great need for evidence based interventions

What can be done about socioeconomic inequalities in health? A report by the King's Fund contains an impressive agenda for action.1 The objective of the report was “to outline a number of practical and affordable ways in which the situation could be substantially improved, if the political will existed to recognise that tackling inequalities in health is a fundamental requirement of social justice for all citizens.”

The report identifies four areas for intervention: the physical environment, social and economic factors, barriers to adopting a healthier personal lifestyle, and access to appropriate and effective health and social services. For each area one factor has been selected to illustrate possible policy initiatives: housing, income maintenance, smoking, and access to health care. The initiatives range from the development of innovative health education programmes to investments in social housing (to be financed by, among other things, the abolition of tax relief on mortgages) and from ensuring an equitable allocation of NHS resources to changes in the tax system (for example, an increase in the highest rate of income tax).

This brief summary of the report cannot do justice to the richness of its ideas. This richness makes the report a welcome complement to a lucid but much thinner discussion paper by the World Health Organisation that was published a few years ago.2 It also reinforces a recent paper on inequalities in health issued by the BMA, which concluded that “a total rather than service-orientated approach is needed across all sectors of government.” The BMA's paper identified a wider range of policy areas that should be involved in this strategy: economic policy (“particularly taxation policy”), provision of welfare benefits, education and child care, unemployment, environment, housing, transport, and leisure.3

The broad and varied approach advocated in these reports certainly fits the scale and nature of the problem: inequalities in health are a widespread phenomenon, resulting from a complex interplay of many different factors; substantial reductions in them are unlikely unless some of their root causes, such as inequality of income, are addressed. On the other hand, the wide range of policy options also shows uncertainty about which measures are necessary and likely to be effective. Much more helpful would be a more parsimonious package that targeted several specified key areas of concern within the larger domain of socioeconomic inequalities in health, focused on the known causes of these inequalities, and used interventions of established efficacy in reducing these inequalities. Unfortunately, current knowledge allows only the first two of these requirements to be (partially) met.

This can be shown for one key area of concern: the widening of the difference in mortality between rich and poor people. Over the past three or four decades there has been consistent evidence of increasing socioeconomic inequalities in mortality, both in Britain and in several other industrialised countries.4 5 6 7 8 9 There are two competing explanations. One emphasises the possible role of increasing inequalities in income and one focuses on the possible contribution of changes in the distribution of behavioural risk factors. For Britain, evidence exists that the increases in the size of socioeconomic inequalities in mortality are related to increases in the size of inequalities in income,10 11 and this relation also emerges from international comparisons at one point in time.12 The competing explanation for the widening difference in mortality focuses on changes in the social distribution of behavioural risk factors. One of the main contributors to the widening difference in mortality is ischaemic heart disease, which over the past 40 years has changed from being an upper class to a lower class disease. These changes are mirrored by comparable changes in risk factors for ischaemic heart disease, such as smoking and obesity.13

Evidence thus suggests that both reducing inequalities in income and reducing the social gradient in smoking and other behavioural risk factors could result in a slowing down or perhaps even reversal of the widening difference in mortality. Unfortunately, however, this is not enough. Both types of intervention have a price, in political, economic, or simply monetary terms. To justify the costs and to enable policymakers to choose rationally from the available policy options, quantitative information is needed on the effectiveness of these interventions. Public health researchers and practitioners frequently criticise those practising clinical medicine for the lack of evidence of the effectiveness of medical interventions.14 They argue, legitimately, that knowledge of the aetiology and pathophysiology of a disease is not enough to justify treatments that seemingly address these factors. What is also needed is evidence of the effectiveness of medical interventions, preferably collected in carefully controlled experiments. The same “evidence based” rigour should be striven for in public health.

Scientifically valid evidence of the effectiveness of the type of interventions advocated by the reports mentioned above is extremely scarce.15 Randomised controlled trials will usually be impossible, but other powerful research designs are available,16 17 and advantage should also be taken of the results of “natural” experiments, in which favourable changes occur for reasons other than to reduce inequalities in health (for example, changes in employment opportunities, housing, or prices of cigarettes). Whenever possible, effects should be expressed in terms of changes in the size of socioeconomic inequalities in health; various such measures are available.18

Political will should be mobilised, not only to implement some of the recommendations of the King's Fund's and similar reports but also to sponsor the research necessary to evaluate the effects. After researchers have documented the existence of inequalities in health and have studied the explanation of inequalities in health there is one further contribution that they can make to health policy: studying the effectiveness of interventions aimed at reducing inequalities in health. This is a substantial undertaking, for which international cooperation is likely to be required: no single country can provide all the “natural” and induced variation that is necessary to evaluate the relevant policy options.

In the Netherlands a government sponsored five year programme aimed at evaluating several possible interventions to reduce inequalities in health is already under way19; the World Health Organisation's discussion paper contains other examples of national, regional, and local initiatives.2 Attempts at reducing inequalities in health cannot be but illustrations of Rudolf Virchow's famous dictum “Medicine is a social science and politics nothing but medicine on a grand scale.”20 Let's try to make this a rational form of medicine.