It is often asserted that more evidence is needed to take action on the social determinants of health. In this guest post Ted Schrecker identifies such claims as a key obstacle to achieving health equity. He argues that to overcome this obstacle, we must recognize that decisions about how much evidence is enough are irrevocably bound together with important ethical and political choices. Ted is a Professor of Global Health Policy at Durham University.
We should now be familiar with many hard facts about health equity. In the United Kingdom, for example, despite rhetorical commitments by the previous government to reduce health disparities between rich and poor, by 2007 such disparities were on many measures greater than at any point since the 1930s. This was before the economic crisis and subsequent austerity measures, which have disproportionately affected the UK’s poorest regions, including the one where I live and work.
Yet in discussions of policy responses, a frequent refrain is that the evidence is not strong enough to justify addressing the “inequitable distribution of power, money, and resources” that was one of the foci of the WHO’s Commission on Social Determinants of Health. Tobacco control initiatives and encouraging people to eat a healthy diet are fine, but not so challenges to “the inequality machine [that] is reshaping the whole planet,” in the words of the editor of Le Monde Diplomatique. Since the Canadian experience shows that a healthy diet is often unaffordable for benefit recipients or the working poor, and more than 47 million people in the United States are relying on the government vouchers known as food stamps, that would seem to be a major omission.
Debates about the strength of evidence are hardly new: think about tobacco, or climate change, or any number of environmental and workplace exposures whose lethality is now widely acknowledged. The role of ethical and political choices about standards of proof (how much evidence is enough) in these debates is often neglected. I began a recent article on epidemiology and social determinants of health with an analogy to the case of former professional athlete O.J. Simpson. Acquitted of the murder of his estranged wife and her friend in a criminal trial, he was nevertheless found liable for damages in a civil proceeding initiated by the survivors of his alleged victims. The difference simply reflects the much higher standard of proof that must be met, in common law countries, in criminal proceedings.
My points were that (a) the concept of a standard of proof is crucial for public health policy; (b) the choice of a standard of proof with respect to social determinants of health, as for environmental exposures, is a matter of public health ethics with respect to which scientists qua scientists have no special competence; and (c) unreflective insistence on a definition of scientific rigour organized around avoiding false positives, or Type I errors, can be highly destructive of health, and in particular health equity, under conditions of uncertainty.
The complexity of the causal pathways that connect macro-scale economic and social processes with health disparities means that some degree of uncertainty is inescapable. A recent report on structural influences on obesity from the Scottish Collaboration for Public Health Research and Policy makes this point effectively, noting that “many strategies aimed at obesity prevention may not be expected to have a direct impact on BMI [Body Mass Index], but rather on pathways that will alter the context in which eating, physical activity and weight control occur. Any restriction on the concept of a successful outcome … is therefore likely to overlook many possible intervention measures that could contribute to obesity prevention.” Conversely, if the evidentiary bar is set high enough, it can always be claimed that nothing works, or that more research is needed … but waiting for more evidence is itself a decision about risks and benefits. This point has been understood for decades, yet it continues to be either ignored or willfully misunderstood by (for example) some protagonists in the current debate over European policy toward endocrine-disrupting chemicals in the environment.
Choice of a standard of proof is one of a larger class of issues and choices at the interface of science, values and politics. Understanding that interface, and in particular its political dimension, is critical to reducing health inequity. Thus, when I read an article that exhorts social epidemiologists to concentrate on narrowly defined questions amenable to experimental or quasi-experimental study designs that will generate “the kind of evidence wanted by policymakers,” my immediate reaction is one of revulsion. The quality of evidence that is demanded by “policymakers” – and the term is itself curiously decontextualized – depends entirely on what those in power have at stake. Often, no evidence or imaginary evidence is sufficient; think about the weapons of mass destruction that Iraq was confidently declared to possess, or the nonexistent jobs into which George Osborne wants to herd poor under-25s. Producing research findings that are not “wanted” by Osborne and his kind should be viewed as an ethical imperative. When public health practitioners and the organisations in which they work are sincerely committed to reducing health inequities in a hostile environment, progressive health researchers should provide all the support we can. But we must choose allies and audiences with care, and often the most appropriate algorithm for our interactions with those in power is the three R’s: Resist, Ridicule, and Replace. More about that in my next posting.