How to fix evidence based medicine for the next generation of doctors
The experts weigh in at Evidence Live 2015
Evidence based medicine (EBM) is an approach to clinical practice that aims to combine patient values, clinical expertise, and high quality evidence to inform patient care and decisions. Although the movement is more than 20 years old, it has recently been claimed that EBM is in crisis and in need of change. Some of the current challenges include assessing what constitutes good evidence, managing the sheer volume of data available, staying up to date with the latest advances, and difficulty of applying evidence and guidelines to individual patients, especially those with multiple conditions.
On 14 and 15 April leading academics, clinicians, teachers, and medical students gathered in Oxford for Evidence Live 2015. In preparation for the event, Student BMJ held a competition for readers challenging them to answer the question: “What is the most important intervention, change, or idea required to ‘fix’ evidence based medicine over the next 10 years, and how should it be implemented?” Authors who submitted the best responses were invited to attend the conference at a discounted rate and the four winning entries were published in the April issue of Student BMJ. Ideas included changing how EBM is taught at medical school, giving patients greater access to the evidence, ensuring guidelines are applicable to patients with multimorbidities, and encouraging junior doctors to play a more active part in advocating evidence based practice.
The conference also invited leading EBM thought leaders to weigh in with their view at a session called “How to fix EBM for the next generation.” In this article we have summarised the main perspectives, problems, and ideas for the ways forward expressed at the event into three major themes: the creation, evaluation, and application of evidence.
How to fix EBM for the next generation
The creation of evidence—how research is done
A lot of research is wasted and much of the evidence is low quality
Iain Chalmers, editor, James Lind Library
Chalmers identified two problems with the evidence used in EBM. Firstly, research wastage is an ongoing problem with many authors duplicating studies. “Embarking on new primary research without assessing systematically what’s known already is unethical, unscientific, and wasteful,” he said. Secondly, he argued that resources are wasted on poorly conducted studies—for example, low quality animal research: “Systematic reviews of animal studies have revealed gross deficiencies in study designs and conduct and in biased reporting of research so it’s no wonder they have such poor replicability.”
In terms of going forward, he suggested that researchers should be made aware of the Evidence-Based Research Network (www.EBRnetwork.org) and lobby widely for reducing research wastage.
Information from patients is not considered evidence
Paul Wicks, vice president of Innovation, PatientsLikeMe
Drawing on his research in patients with amyotrophic lateral sclerosis, Wicks suggested that one problem with EBM is that patient accounts and anecdotes are not considered valid evidence. Although he did not think that EBM was broken, he argued that researchers are ignoring valuable patient data.
He suggested that gathering these data from patients would complement other forms of evidence collected and proposed using the internet to create this evidence base. He pointed out that researchers should think carefully about what types of patient reporting would be useful to aid evidence based decisions. “A report in neurological disease on whether [patients] can walk or not is probably a lot more credible than asking someone with a heart condition what their blood pressure is,” Wicks said.
Evaluation of evidence—how we assess its clinical value
The written record of published peer reviewed literature is difficult to verify
Tom Jefferson, editor, Cochrane Acute Respiratory Infections Group
Jefferson suggested that studies published in peer review journals, particularly studies on pharmaceutical interventions, may not be reliable as the results are difficult to verify. Two of the main problems are the condensed format of published randomised controlled trial data and the belief that authors have accurately represented their trial data. “The stuff that’s in ten pages cannot be trusted to reflect what is thousands and thousands of pages underlying a trial,” he said. He argued that this problem applies to both trials funded by drug companies and those that are publicly funded.
Jefferson’s solution for this was to change the standard of trial reporting to include all information submitted to regulatory agencies, rather than the “summary” of information typically published in peer review journals.
Drug companies influence prescribing in general practice
Anne Van den Bruel, associate professor, Nuffield Department of Primary Care Health Sciences, University of Oxford
Van den Bruel discussed the free continuing medical education doctors get from drug companies, either from drug representatives or at conferences funded by drug companies and argued that this is problematic. “Doctors don’t realise how influential that information is on their prescribing,” she said. Van den Bruel highlighted the dangers of trusting information that comes solely from drug companies without proper evaluation or assessment of the wider evidence base: “Often, information about safety and harms will not be included in the educational talks.” She said that clinicians should stop seeing drug representatives in general practice and stop going to sales talks.
Application of evidence—how it informs practice
EBM is not as coercive as it should be
Howard Bauchner, editor in chief, JAMA
Bauchner argued that EBM is not broken but that “it remains in its infancy.” However, he suggested that EBM should have a greater impact on decision making. He argued that some clinical decisions are so well supported by evidence that deviation from the indicated decisions is not appropriate—for example, prescribing antibiotics for bacterial meningitis. “I think EBM needs to become more coercive . . . I don’t think it should be coercive over many things, but if we are certain about certain decisions, we shouldn’t allow much deviation from those decisions,” he said.
In terms of how to make this a reality, Bauchner said that we could make use of the electronic medical record. “We should take the five or six major health problems that we face in hospitals, understand the highest grade evidence and we should use the electronic health record to reinforce that that evidence-based medicine diagnostic or therapeutic approach is done with that patient,” he said.
Financial incentivisation of evidence based performance targets distorts clinical practice
Iona Heath, former president, Royal College of General Practitioners
Heath identified three problems with EBM today. Firstly, she said that evidence based guidelines that are financially incentivised through “pay for performance” distort clinical practice as they prevent doctors from thinking about the individual patient and their preferences. “If you’re going to get a substantial amount of money if you meet a certain set of targets around a treatment of blood pressure across your population, you stop thinking so carefully about the individual and their predicament and whether a blood pressure of less than 140 [mm Hg] is good for them and good given their age and morbidity.” Secondly, she argued that guideline based targets may be inappropriate as the evidence underpinning guidelines can quickly become outdated. Thirdly, the evidence on which targets are based often does not apply to many patients, specifically older patients with multimorbidities and on many medications. She said, “Evidence based medicine has almost nothing to offer for the vast majority of people over 75 who have more than three conditions. Where is the research on the taking of multiple medications?”
Despite there being differences of opinion between the experts about the causes of, and solutions to, EBM, the general consensus was that EBM is unfinished and needs to evolve and improve. Some of the panel members identified fundamental flaws which would require a heavy overhaul of EBM—such as reforming the published peer review literature system—whereas others thought EBM isn’t broken but could be tweaked to be more effective.Dylan Collins, PhD student1, Niklas Bobrovitz, PhD student1
1Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
Correspondence to: firstname.lastname@example.org
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.
- Evidence Based Medicine Working Group. Evidence-based medicine. A new approach to teaching the practice of medicine. JAMA 1992;268:2420-5.
- Greenhalgh T, Howick J, Maskrey N; Evidence Based Medicine Renaissance Group. Evidence based medicine: a movement in crisis? BMJ 2014;348:3725.
- Collins D, O’Neill B. How would you fix evidence based medicine? Medical students’ solutions on the way forward. Student BMJ 2015;23:1522.
Cite this as: Student BMJ 2015;23:h2861