Evidence-Based Medicine (EBM) as an idea is usually attributed to Archibald Cochrane and his seminal work “Effectiveness and Efficiency”. The basic idea is to no longer rely on “Eminence-Based medicine”, that is medical knowledge based on authoritative opinions, but on the best available evidence. This has now expanded to an international movement well accepted in medicine, with the Cochrane Collaboration being the most visible institution. Building up a library of systematically reviewed medical evidence and ongoing development of the necessary methods is the most notable achievement. When it comes to clinical evidence, the Cochrane Collaboration/the field of EBM developed quite sophisticated and widely agreed methods in collecting and evaluating the evidence, with meta-analysis of randomized clinical trials collected through systematic reviews as the highest level of evidence.
Evidence-Based Public Health (EBPH) has the goal of applying the principles of EBM to the field of Public Health. Public Health is, simply put, “the science and art of preventing disease, prolonging life and promoting physical health and efficiency through organized community effort”. Generally, the evidence for public health intervention is, as compared to clinical interventions, rather weak.
A crucial difference between EBM and EBPH is that in most cases EBPH does not have the luxury of the experiment. Despite limitations, the experiment is still the gold standard of causal inquiry in science. Hence, not only is generating and/or identifying (policy) evidence more difficult for EPBH, but also the external validity of the evidence is more difficult to establish. Certain findings may only be valid under certain cultural, social, or economic circumstances. In the problem of identifying mechanisms for whole populations/countries, EBPH shares very similar problems set with the field of (macro-)economics/econometrics.
Several analytical tools and approaches for EBPH exist, namely (i) public health surveillance, (ii) systematic reviews and evidence-based guidelines, (iii) economic evaluation, (iv) health impact assessment, and (v) participatory approaches. From these five approaches, health impact assessment (HIA) is a relatively new and comprehensive approach that aims to account for social and environmental determinants when assessing the effects of (all) policies on population health. HIA made great progress in terms of developing causal webs for intentional and unintentional health effects of policies and does not ignore the need to incorporate participatory approaches for both, the target population and decision makers. In particular HIA, championed the use of population health intervention models to quantify and compare the expected effects of competing interventions.
What findings can be used as evidence and how should evidence be ranked?
Clearly, the hierarchy as used by EBM for clinical evidence can only be partly applied. To gather evidence on successful interventions, innovative evidence such as “natural experiments” or econometric methods should be explored and evaluated for their use in EBPH; in this EBPH shares a problem with economics. EBPH lacks a clear ranking of the relative value of novel/alternative types of evidence to each other.
How can evidence from one context being extrapolated and applied to a new context?
From an epidemiological point of view, quite a lot is known about the relationship between a risk-factor and health outcomes on an individual level. However, at the level of a population several factors influence the effect of an intervention on population health. Most notably, the disease and risk-factor profile differ between populations, yet must be accounted for when quantifying the anticipated effect. Moreover social, cultural, and economic effects have an influence on how an intervention is accommodated by the target population(s) and eventually influence individual behavior.
How can EBPH influence public policy/health policy?
To gain credibility in the policy arena, two important conditions for interventions—as suggested by EBPH—are the ability (i) to quantify the consequences of the implementation/non-implementation and (ii) to provide information about the relative (cost-)effectiveness as compared to competing policy proposals. But policy-making is only partly a rational process, it is still unclear why certain suggestion are being adopted by certain groups of decision makers at a certain time whilst others, from an EBPH perspective equally or even more compelling proposals, are not.