Statistics from Altmetric.com
In practice, we seek to use the best evidence to underpin our decisions. That evidence is likely to be a mix of robust research evidence, clinical experience, patient preferences and experiences, and local information 1 So research is an important part, but not the only part, of your decision about care.
Although there is much talk about evidence-based policy, research evidence has to compete with a range of additional factors that influence decisions about what will become policy. These factors may include experience, expertise, judgement, values, resources, habits and traditions.2 In addition, The Economist3 commented ‘the ambiguities of science sit uncomfortably with the demands of politics. Politicians, and the voters who elect them, are more comfortably with certainty.’ p13. Krebs4 discusses the relationship between scientific evidence, uncertainty and risk. He argues that turning scientists' risk assessments into policy is a political judgement, and political decisions involve not only scientific evidence but also the economic and public acceptability of risk.
In a discussion of government commitment to evidence-based policy, Delamothe5 discusses the claim that horse riding is more dangerous than taking ecstasy. Nutt6 reported that horse riding is associated with acute harm once every 350 episodes and that the figure for ecstasy is once every 10 000 episodes. The home secretary of the day told Professor Nutt, (then chairman of the Advisory Council on the Misuse of Drugs), that she was ‘profoundly disappointed’ by his comparison. Delamothe reports that her being profoundly disappointed by this truth made a deeper impression on him than the original statement on risk. He argues that ‘… government would prefer the evidence to align more closely with its policies.’ p1117.
So science may at times be uncertain, or even when strong, may be giving a politically unpopular message. Recently, Katikireddi et al7 questioned how evidence-based public health policy was in England, and concluded many of the interventions proposed lacked evidence of effectiveness or were likely to be ineffective. This was despite the Secretary for State for Health8 stating that ‘We must only support effective interventions that deliver proven benefits.’
A recent example of the same evidence about breast implants having different policy outcomes was reported by Reuters.9 The French government offered removal of certain implants because there was a danger of rupture and inflammation. In Britain, the Chief Medical Officer10 reported ‘we have no evidence of …an increased risk of rupture.’ Different national governments acted on evidence in different ways.
Wells et al2 presented a government department's perspective on handling uncertainty in health sciences. They argue that science can reduce uncertainty, and can be a ‘start point for a dialogue between decision makers and researchers.’ p4853. They highlight that researchers and policy makers have different notions of evidence. Greenhalgh and Wieringa11 describe evidence-based policy as ‘a process of argumentation to decide what is right and reasonable’ although they caution against ‘potentially sinister forms of symbiosis between government, industry and science’ p507.
So evidence-based policy is not as straightforward as only basing policy on the best research evidence. Research evidence may be incomplete or uncertain. Other factors such as values, resources and public opinion may be influential. If policy is based on partial evidence, this needs to be clear and an evaluation of its effect is important. More dialogue between researchers and policy makers seems likely to improve understanding of different perspectives, but a balance is needed between research addressing pressing policy issues and research that may currently be outside those policy priorities.
In clinical practice, just how the healthcare practitioner should combine or meld the different forms of evidence as they make a decision is a neglected area. Similarly in evidence-based policy, the way in which different factors are taken into account in making policy needs to be made explicit, together with a recognition that people will come to evidence-based policy with different perspectives and priorities.
Competing interests None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.