Dear editor,
We read with great interest, the article by Greenhalgh et al. (2014)[i] on ‘Evidence based medicine: a movement in crisis’. Although the aims of the original EBM proponents to incorporate scientific knowledge more explicitly into clinical decision-making were laudable, Greenhalgh et al.’s incisive critique highlights some of EBM’s unintended consequences.
We agree with Greenhalgh et al.’s assertion that real EBM places the care of individual patients at its heart. To this end, we would like to highlight the concept of “personalised evidence” we proposed in a recent article[ii]. This concept arose out of a project in a mental health trust which sought to evaluate the feasibility of integrating a clinical librarian into mental health teams. We observed that the utility of research for practitioners did not always accord with the “evidence hierarchy” commonly promulgated by EBP advocates. Indeed, lower ‘level’ evidence, such as case reports may have greater ‘salience’ (because they are more personalised) to individual presentations than more generic population-level evidence. We therefore propose a tension between ‘level’ and salience’ of evidence and have illustrated this tension graphically with a matrix diagram[iii].
We believe that the concept of ‘personalised evidence’ can complement personalised medicine. The latter has focused on pharmacogenetic factors which may mediate treatment response. However, these factors will have no effect without patient concordance with treatment. In such cases, lower ‘levels’ of evidence, for example, case reports or evidence in the form of clinicians’ or knowledge of the patient’s unique circumstances can complement the more rationalist knowledge available from personalised medicine[iv].
The use of ‘lower’ levels of evidence is not recommended in frequentialist research traditions. However, the contrasting Bayesian approach integrates findings from newly observed data with prior knowledge[v][vi]. In this way, ‘lower levels of evidence can be combined with emerging higher levels of evidence to increase our certainty about the probably outcome of a course of action’[vii].
With reference to Greenhalgh et al.’s point about the difficulties of applying the ‘evidence base’ to patients with multi-morbidities, we would also highlight the difficulties which can accrue when trying to apply research findings to patients with single diagnoses. For example, although a classic triad of symptoms has been described relating to Wernicke-Korsakoff Syndrome (a neuropsychiatric consequence of dietary deficiency),the classic triad is only present in a minority of patients[viii].A survey found that 21% of Consultant Psychiatrists and 10% of A&E Specialists required all three symptoms to be present before they would administer parenteral thiamine – the treatment and prophylactic agent for Wernicke-Korsakoff[ix]. Thus, it may be problematic to apply best evidence even when an unambiguous treatment is indicated if there are difficulties in diagnosing a condition appropriately.
The Wernicke-Korsakoff example also supports another of Greenhalgh et al.’s propositions – that EBM should not de-value knowledge derived from the basic sciences. Randomised controlled trials have shown that some biologically plausible treatments have either no benefit, or are harmful[x][xi]. However, in the Wernicke-Korsakoff example, best practice guidelines are based on evidence from case reports and clinical experience[xii]. These are low levels of evidence: yet in this case, the role of thiamine as the treatment and prophylactic agent is clear. Parenteral thiamine is, however, under-used in practice, largely due to misplaced fears about provoking anaphylaxis[xiii].
We therefore support Greenhalgh et al.’s critique and would argue for the evolving concept of ‘personalised evidence’ to complement personalised medicine and EBM to optimise more holistic clinical practice.
Yours sincerely,
Rachel Steele and Paul A. Tiffin.
23 June 2014
[i]Greenhalgh T, Howick J, Maskrey N. Evidence based medicine: a movement in crisis? BMJ 2010;348:g3725.
[ii]Steele R, Tiffin, PA. ‘Personalised evidence’ for personalised healthcare: integration of a clinical librarian into mental health services – a feasibility study. Psychiatr Bull 2014;38:29-35.
[iii] Steele R, Tiffin, PA. ‘Personalised evidence’ for personalised healthcare: integration of a clinical librarian into mental health services – a feasibility study. Psychiatr Bull 2014;38:29-35.
[iv] Steele R, Tiffin, PA. ‘Personalised evidence’ for personalised healthcare: integration of a clinical librarian into mental health services – a feasibility study. Psychiatr Bull 2014;38:29-35.
[v]Holloway CA. Decision Making Under Uncertainty: Models and Choices. Prentice Hall, 1979.
[vi]Press SJ. Subjective and Objective Bayesian Statistics: Principles, Models and Applications. John Wiley & Sons, 2002.
[vii] Steele R, Tiffin, PA. ‘Personalised evidence’ for personalised healthcare: integration of a clinical librarian into mental health services – a feasibility study. Psychiatr Bull 2014;38:29-35.
[viii]Harper CG, Giles M, Finlay-Jones R. Clinical signs in the Wernicke-Korsakoff complex: a retrospective analysis of 131 cases diagnosed atnecropsy. J Neurol Neurosurg Psychiatry1986; 49: 341-5.
[ix]Hope LC, Cook CCH, Thomson AD. A survey of the current clinical practice of psychiatrists and accident and emergency specialists in the UK concerning vitamin supplementation for chronic alcohol misusers. Alcohol Alcohol 1999; 34: 862–7.
[x]Torgerson DJ, Torgerson CJ. Designing Randomised Trials in Health, Education and the Social Sciences: An Introduction. Palgrave Macmillan, 2008.
[xi]Howick J. The philosophy of evidence-based medicine. Wiley-Blackwell, 2011.
[xii]Day E, Bentham PW, Callaghan R, Kuruvilla T, George S. Thiamine for prevention and treatment of Wernicke-Korsakoff Syndrome in people who abuse alcohol. Cochrane Database Syst Rev 2013;7:CD004033.
[xiii]Day E, Callaghan R, Kuruvilla T, George S, Webb K, Bentham P. Pharmacy-based intervention in Wernicke’s encephalopathy. Psychiatr Bull 2010;34:234-238.