ROUND TABLE DISCUSSION
Health inequalities impact assessment
Donald Acheson
Former Chief Medical Officer of England; Chairman of the International Centre for Health and Society, Department of Epidemiology, University College London, 119 Torrington Place, London WC1E 6BT, England
I have been asked to consider Davidson Gwatkins article in the light of the Report of the United Kingdoms recent Independent inquiry into inequalities in health which I chaired (1). While it is clearly inappropriate to generalize from the experience of one country in terms of particular policies, the weight of scientific evidence suggests, as indeed does Davidson Gwatkins article, that a socioeconomic explanation of health inequalities is likely to be relevant for all countries rich and poor throughout the world.
The socioeconomic model traces the roots of ill-health far beyond health services to such determinants as income, education and employment as well as to the material environment and lifestyle. This has the practical implication that the necessary policy developments to reduce health inequalities will extend far beyond the remit of departments of health, some of them relating to the government as a whole while others will fall within the terms of reference of a range of other government departments.
As has been the experience elsewhere, the Inquiry found that although health in England (as judged by reductions in mortality rates) had over the past 50 years on average improved greatly, in recent decades inequalities in health had either remained static or widened. These inequalities can be identified at all stages of the life course from pregnancy to old age.
The Report selected the following three of its 37 recommendations as crucial.
- All policies likely to have an influence on health should be evaluated in terms of their impact on health inequalities.
- A high priority should be given to the health of families with children.
- Further steps should be taken to reduce income inequalities and improve the living standards of poor households.
While the reasoning which supports the second and third of these priorities is self-evident, the first, which recommends an important development in health impact assessment, needs further explanation.
Experience shows that a well-intended policy which improves average health in a population may have no effect on inequalities. Indeed it often widens them by having a greater impact on the better-off. This has happened in some initiatives concerned with immunization and cervical screening, as well as in some campaigns to discourage smoking or promote breastfeeding.
These examples highlight the need for health policies to focus extra attention on the health of the less well-off. This could be done both by policies directed specifically at the less well-off, and by an approach which would require inequalities to be considered wherever universal services are provided (such as publicly funded education or health care) and where other policies are likely to have an impact on health.
The Report identified a wide range of areas for future policy development relevant to the reduction of health inequalities judged on the scale of their potential impact and the weight of the evidence. These policy areas include poverty; tax and benefits; education; employment; housing and environment; mobility; transport and air pollution; and nutrition. In addition, a number of other policies were put forward in relation to stage in life course for mothers, children and families; young people and adults of working age; and older people and in relation to ethnic and gender inequalities.
An important aspect of the Inquirys work was to confirm once again the findings of Black (2) and others that it is an over-simplification to consider socially related ill-health and attenuated life span as restricted to those living in poverty. Poverty is defined here as households in receipt of less than 50% of the average income. It is a rule with few exceptions that whenever it has been possible to relate mortality or morbidity to a graduated social indicator such as income, extent of education, or skill of work task, a gradient has emerged. In England, gradients across the whole social spectrum were found for both men and women. These exist for men in respect of mortality from all causes, coronary heart disease, lung cancer, stroke, accidents and suicide, and for women for all causes of death and for coronary heart disease. Socioeconomic gradients were also shown for longstanding illness in both sexes, and for obesity and high blood pressure in women.
These findings carry an important lesson for policy-makers, namely that measures aimed exclusively at helping those in poverty, or at the bottom of the social hierarchy, will deal with only a small proportion of the burden of socially related ill-health and premature death.
A new direction for public policy which explicitly addresses inequalities is therefore needed. It was the view of the Inquiry that reductions in inequalities in health were most likely to be achieved if all relevant policies (and this will include many social policies, far beyond health care) are formulated with the reduction of inequalities in mind.
Hence the Inquirys first recommendation:
"As part of health impact assessment all policies likely to have direct or indirect effects on health should be evaluated in terms of their impact on health inequalities, and should be formulated in such a way that by favouring the less well-off they will, whenever possible, reduce such inequalities."
The transformation of health impact assessment to health inequalities impact assessment will be a prerequisite for such policies. Their effective formulation and evaluation will be a major challenge for public health workers and should be a priority for research and development. This will be a major step on the path "from analysis to action" suggested by Davidson Gwatkin. 
1. Acheson D. Independent inquiry into inequalities in health report. London, The Stationery Office, 1998.
2. Black D et al. Inequalities in health; report of a research working group. London, Department of Health and Social Services, 1980.