A rigid application of research based on evidence-based medicine (EBM) – an approach that has dominated medical thinking for 30 years – is ill-suited to address many complex health issues where social aspects such as poverty play a key role, say experts who argue for greater emphasis on other sources of evidence that EBM overlooks.
Published today in the latest issue of Public Health Research & Practice, a peer-reviewed journal of the Sax Institute, the paper examines the limitations of EBM in a research context, which the authors define as the systematic analysis of published medical and health research as the basis for clinical decision making – for example, the decision about which medicine to use to treat an illness.
The paper’s authors, Dr Susan Jacups and Dr Clare Bradley, both from the School of Public Health at the University of Queensland, say that the “gold standard” status that EBM bestows on randomised controlled trials (RCTs), while other forms of evidence are accorded less weight, is creating difficulties for research into socially-based problems, because the RCT is an ill-suited study design to address these complex health issues.
The paper argues for a move away from “the current obsession with randomised controlled trial-based EBM”, which continues to drive unsuccessful, often laboratory or medically-based research, in populations where the disease process is recognised to be socially driven. Instead, more comprehensive solutions are needed that draw from a wider range of relevant information, such as expert opinion, reasoning and increased service delivery pilot trials.
Although EBM regards RCTs as preferential sources of evidence over other study designs, which in turn encourages research projects to be designed as RCTs, the authors say that in complex situations, it can be difficult to design RCTs that adequately account for all the factors that exert an influence on disease initiation or progress – at least some of which are likely to be unknown or unmeasurable.
They use the example of treating a type of ear infection called otitis media, which is common in children in remote Aboriginal populations, and which can be linked to a complex web of overlapping factors. In addition to physiological issues, social factors such as nutrition, sleep, housing, financial stress and interpersonal violence can all contribute to the prevalence of these infections in Aboriginal children.
The authors report that although RCTs have been carried out testing a number of treatments for otitis media in these populations, the findings all tend either to contradict each other or show no statistically significant difference between experimental and control groups.
For example, a trial that tested antibiotic ear drops in Aboriginal children in one remote location found a benefit for those receiving the treatment, compared with those who were given a placebo. But when a trial with the same design was repeated in a different remote location, the treatment reported no significant effect.
This not only makes it difficult to base clinical decisions on the evidence, but the authors note that some of the interventions tested in randomised controlled trials may actually be harmful to the trial participants – for example, the insertion of tubes into the ear (grommets) can introduce bacterial pathogens into the middle ear.
Rather than more RCTs, the authors suggest the solution to improving ear health in Aboriginal children may come through expert clinicians tailoring treatments to individual children and their families. Clinicians and patients may also need to accept that sometimes the best treatments don’t work. A focus on improving underlying living standards, healthcare access and health literacy is also critical, they say.
They add that for complex cases such as this, research funding bodies such as the National Health and Medical Research Council (NHMRC) should place greater emphasis on what are currently considered lower levels of evidence, such as expert opinion or case-series studies – a potentially significant step given the widespread acceptance of the EBM-derived ‘hierarchy of evidence’ in which these other sources of information are seen as less reliable.
“Clinical randomised controlled trials addressing complex health issues where social determinants play a major contributing role are unlikely to return meaningful results,” the authors conclude.
“We believe that the tide may turn, and society will return to trusting expert clinical knowledge as empirical evidence.”
Please acknowledge Public Health Research & Practice as the source for any stories on our papers.
PUBLICATION DATE
15 March 2023
MEDIA ENQUIRIES
0451 122 146
hugo.wilcken@saxinstitute.org.au
