Cefn Coed Hospital, Swansea, UK
Chase Farm Hospital, Enfield, Middlesex, UK
Correspondence: Dr D. D. R. Williams, Cefn Coed Hospital, Swansea SA2 0GH, UK
![]() |
ABSTRACT |
---|
Aims To demonstrate that evidence-based medicine is not new, sources of evidence are limited and psychosocial aspects of medicine are neglected in this process.
Method Some of the literature is reviewed. Ideas and arguments are synthesised into a critical commentary.
Results These are considered under four headings: evidence-based medicine is not new; what evidence is acceptable; the doctor as therapist; and the emergence of a new utilitarian orthodoxy.
Conclusions It is agreed that a degree of professional consensus is necessary. However, too great an emphasis on evidence-based medicine oversimplifies the complex and interpersonal nature of clinical care.
![]() |
INTRODUCTION |
---|
![]() |
EVIDENCE-BASED MEDICINE IS NOT NEW |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Hard evidence, as defined by the criteria of the EBM movement, has existed for many years and psychiatrists have acted on the findings of several major RCTs since the Second World War. It is worth recalling three of these.
Insulin coma therapy became the treatment of choice for schizophrenia after its introduction in 1933 at Vienna (Sakel, 1938). The evidence of its value was never fully convincing, but after an RCT (Ackner & Oldham, 1962) showed that it conferred no benefit over barbiturates for inducing coma it was soon abandoned.
In the early 1960s the Medical Research Council undertook a landmark study of the treatment of depressive illness (Medical Research Council, 1965). This large RCT compared the efficacy of electroconvulsive therapy, imipramine, phenelzine and placebo. It clearly established the efficacy of imipramine, which has been used since as the gold standard for comparing the efficacy of subsequent drugs with antidepressant potential. This study, carried out 40 years ago, would have complied with current EBM standards and it is worth noting that a certain Professor A. L. Cochrane was a member of the committee running the trial.
Also in the early 1960s, The National Institute of Mental Health (NIMH) carried out a large RCT of the treatment of schizophrenia (NIMH-PSC Collaborative Study Group, 1964). Fluphenazine and thioridazine were compared with chlorpromazine and placebo. It showed the efficacy of drug treatment in acute schizophrenia and showed that there was no evidence for the view that chlorpromazine was more effective in patients requiring sedation or that the piperazine derivative fluphenazine was more effective in withdrawn patients. It also cast doubt on the view that extrapyramidal dysfunction was necessary for improvement to occur.
There is little doubt that psychiatrists acted on the results of these trials. Their broad conclusions provided the foundations for the modern treatment of depressive illness and acute schizophrenia.
![]() |
WHAT EVIDENCE IS ACCEPTABLE? |
---|
Limitations of the evidence
Evidence deemed acceptable by the EBM movement as that on which treatment
must be based is essentially information derived from RCTs or meta-analyses.
This narrow approach diminishes consideration of other types of evidence
available from naturalistic enquiry, case material and experiential sources
and, for some doctors, is too blinkered a view.
Many clinicians regret the demise of the psychiatric case report. We feel that this significantly impoverishes the literature and further increases the tilt towards publications of an evidence-based nature and away from those considering individual patients. In 1999 this journal was encouraged to publish case reports for two reasons.
"The first is that the case is capable of generalising some fresh conjecture that is then amenable to empirical testing. The second is that the case embodies an effective refutation of a particular hypothesis." (Farmer, 1999)
For clinicians these are not the only values: in addition, case studies can be a particularly helpful way of stimulating new clinical and psychodynamic insights and of illustrating nuances of treatment.
In clinical practice we are concerned with not missing cases and with not misdiagnosing non-cases as cases. In research the most important consideration is to have a specific and defined sample. Most psychiatric data are continuous, whereas EBM treatment data are categorical.
There may be no evidence. "It is a short step from without substantial evidence to without substantial value" (Bradley & Field, 1995). However, the absence of evidence of effectiveness is not the same as absence of effectiveness. Not all therapies are studied to the same extent.
Limitations of RCTs
Because the conclusions of RCTs are frequently deemed to be the only
acceptable form of evidence, some observations about their design are
pertinent. Most trials have an upper age limit of 64 years for inclusion. In
many trials, women of child-bearing age or pregnant are excluded. Patients
with mixed diagnoses and comorbidity are also excluded. Drop-outs are a major
problem in RCTs. People may not cooperate for a variety of reasons:
side-effects are often cited, nevertheless personality and social factors must
play important roles. Randomised controlled trials rarely give information
about these important issues. It often appears that those who are motivated to
cooperate in RCTs are not a typical cross-section of our patients.
Many patients have multiple diagnoses and problems. Diagnosis alone is a poor predictor of treatment outcome. Personality characteristics and social circumstances influence therapeutic response. Efficacy in RCTs is no guarantee of effectiveness in the field, and effectiveness in the field is no guarantee of effectiveness in the individual patient. Efficacy and effectiveness may be conflated, to the confusion of clinicians, researchers and policy-makers (Chiesa & Fonagy, 1999; Wells, 1999).
Randomised controlled trials only provide us with information about groups, not individuals. A drug treatment may be helpful for a patient in the context of an individually tailored community care package or with psychological support, but not otherwise. There may be little or no hard evidence on which the treatment regime for a particular patient can be based. "Current emphasis on EBM and clinical effectiveness has meant that the prescribing behaviour of clinicians, including psychiatrists, is high on the research agendas" (Buston et al, 1998). However, other factors must be taken into consideration.
Limitation of meta-analyses and systematic reviews
Because the results from meta-analysis seem very precise and convincing and
are beginning to have an impact on practice, clinicians must be reminded of
the "file drawer problem"
(Rosenthal, 1979). For any
given research area, one cannot tell how many studies have been conducted but
never reported. Considerable uncertainty exists. If the studies that show
non-significant results were known, could these negate completely the
conclusions of the meta-analysis? Easterbrook et al
(1991), in a careful
retrospective study of 487 research projects in Oxford, confirmed the presence
of a publication bias in a cohort of clinical research studies and suggested
that conclusions based only on a review of published data should be
interpreted cautiously. They emphasised the need to identify the results of
unpublished as well as published studies.
Caution needs to be exercised in reading systematic reviews, particularly if the amount of information found is small; few trials and small numbers lessen the reliability of a review. It is rare to find a systematic review devoted to adverse effects (Anonymous, 1999).
Role of qualitative research
Databases for EBM are founded on metaanalyses of quantitative research and
therefore miss information that can be gained from a naturalistic enquiry:
from qualitative research where the prime goal is not to enumerate
(Buston et al, 1998).
Qualitative research methods provide an in-depth examination of a small number
of patients rather than a limited examination of large numbers. They are also
able to investigate researchers' interpretations and meanings. Necessarily, in
hypothesis-driven quantitative research the variables are reduced and
information is honed down at the start the entire complexity of the
setting in which the subjects/patients find themselves is missing.
Qualitative research places great emphasis on examining why a particular treatment did or did not work and generating new ideas from these facts. A recent series (Mays & Pope, 2000; Meyer, 2000; Pope et al, 2000) described the application and usefulness of qualitative research methods in different settings and outlined how this alternative to quantitative research is coming to be accepted increasingly in health care research. This approach has the capacity to explore what meanings the symptom, the consultation and the treatment may have for the patient.
Assessing the evidence
Critical appraisal is clearly important. Some journal clubs have
successfully introduced it (Warner &
King, 1997; Geddes,
1998). However, one can be in danger of not seeing the wood for
the trees. Simple reiteration of the paper is clearly not satisfactory but
journal clubs where the critically appraised parts of the paper are
exhaustively discussed risk losing a lively discussion of the subject of the
paper perhaps even including "interesting patients I have
seen". This type of discussion may be dismissed as mere anecdote but
perhaps the role of narrative as explanation is as important for the doctor as
for the patient.
Narrative of illness provides a framework for approaching a patient's problems holistically and may uncover diagnostic and therapeutic options (Greenhalgh, 1999). In addition to emphasising the interpretive (the discernment of meaning) component of taking a history, it offers a method for addressing existential qualities such as inner hurt, despair, hope, grief and moral pain, which frequently accompany and may even constitute people's illness. A recent series of articles (Elwyn & Gwyn, 1999; Greenhalgh, 1999; Greenhalgh & Hurwitz, 1999; Jones, 1999; Launer, 1999) made a very strong plea that the lost tradition of narrative should be revived in the teaching and practice of medicine.
A considerable amount of time now is being spent on critical appraisal exercises. Psychiatric trainees might spend some of it better by devoting more time to studying individual patients in depth and with a critical attitude to acquiring basic psychotherapeutic skills. Busy practising doctors often do not have the time or the expertise regularly to carry out accurate and useful critical appraisal. Most rely on expert abstracting journals such as Journal Watch to do this for them. The French pharmacological journal Prescrire has taken this a step further. A quick glance at the cartoon icon beside each review tells readers whether the drug is a major breakthrough (little man joyfully jumping up and down), a me too drug (little man dropping a capsule into a box full of others) or simply ineffective (little man booting a capsule out of the window). This interesting innovation puts regular critical appraisal into context for busy clinicians.
![]() |
THE DOCTOR AS THERAPIST |
---|
Naturally the importance of the doctor as the therapist or the therapy has changed over time and varies from culture to culture. As specific treatments, both surgical and medical, have been developed the role of doctor as therapist may have diminished. Nevertheless, the doctor-patient relationship remains a potent factor and is particularly important in chronic ill health, terminal illness, rehabilitation and any disorder that may be influenced psychologically. Even in illnesses where there is a specific treatment, the full potential of the placebo response should be realised because it reflects the positive psychological need to overcome illness.
In the last decade of the 20th century a curious paradox developed. Doctors are now being urged at one and the same time to take users' wishes for treatment into account and to follow the edicts and restrictions of EBM. As the medical profession is embracing EBM, more and more patients are seeking out alternative and mostly unevaluated therapies. A recent survey revealed that the number of people using non-conventional treatments had doubled in the previous 6 years (Gregoriadis, 1999). One in five Britons had turned to complementary or alternative therapies, revealing that osteopathy, acupuncture, reflexology, aromatherapy and yoga have been gaining in popularity. The survey also found that 80% of people believe that such treatments will become more popular. The medical profession needs to understand the reasons for this trend. A number of suggestions have been put forward in explanation, including: the belief that conventional therapies do not work: less willingness to put up with side-effects; a desire to spend longer with a therapist than is possible with their general practitioner, where the average consultation lasts 8 min; and the demand for a more holistic approach to problems. Conventional medicine, by overemphasising its technical and scientific aspects, risks losing the art and humanity of its practice. It will be regrettable if two medical streams become established separately: an orthodox school linked with EBM and a range of alternative therapies.
It is essential for the National Health Service to provide a technically successful medical service that is packaged in a personal, empathic and holistic way. This approach was outlined clearly at the beginning of the 20th century by William Osler (Bliss, 1999), who understood that medicine has to be rooted in science but that doctors also must be healers.
![]() |
AN EMERGING NEW UTILITARIAN ORTHODOXY? |
---|
Archie Cochrane was not a clinician a fact that must be kept in mind and he may not have had the special experience of seeing a patient respond dramatically to treatment. Clinicians see this at first hand and have to balance these insights with evidence from accepted EBM sources. This is currently a problem in old age psychiatry. The EBM data on donepezil show a marginal improvement in patients with mild to moderate dementia. On the other hand, there are good examples of individual patients showing an impressive improvement (Dening & Lawton, 1998; Manchip & Morrison, 1999). Evidence of this nature must be recognised and not dismissed as anecdotal.
This emerging orthodoxy is leading to problems in three areas: the macro-economic level of purchasing health; the individual transaction between the doctor and patient; and the inhibiting influence of political correctness.
Information from EBM is being used to encourage minimalistic purchasing in the name of science. This pseudoscience may lead to rationing and the non-purchase of care that clinical judgement says is useful. A number of purchasers have not allowed the prescription of donepezil, rivastigmine and galantamine, citing the evidence for denying patients this therapeutic possibility, rather than allowing a 3-month trial to discover those individuals who will benefit. The National Institute for Clinical Excellence (2001) recently has seen the evidence in a different light and recommended guidelines for the use of anti-dementia drugs, regardless of the patients' address.
The medical profession is a broad church and often there are different ways of treating patients. In the EBM culture it is difficult to imagine how inspired new ideas can be developed and how new treatments can be initiated. The history of medical progress provides many examples of therapeutic developments as a result of the successful treatment of one patient (e.g. the first surgical operation for mitral stenosis).
Unfortunately the EBM approach is also a covert assertion that only factors that can be measured are recognised as important. This perspective reinforces an enduring weakness in British medicine: a dismissive attitude towards the importance of psychological and social factors in the causation and treatment of disease in the individual. We are convinced that scientific medical practice must be underpinned by the need to understand and respond empathically to the illness in accord with the patient's experiential perspective. This view is confirmed by the launch of a new journal Medical Humanities (Evans & Greaves, 1999). This will provide a forum for inquiry and serious discussion of the need to engage the humanities in medical education and professional development. It will explore how the scientific understanding of disease can be integrated with the human understanding of the experience.
![]() |
DISCUSSION |
---|
Our views are in accord with the criticisms that have been expressed already (Black, 1998; Kelly, 1999; Laugharne, 1999). Treating patients is a complex process. Evidence-based medicine must be only one element in the complexity of clinical decision-making. Patients need to be managed in a holistic way, in the context of their culture, unique psychological make-up and their relationship with their physician. Kafka's aphorism is as valid today as in 1917: "Prescribing is so easy, understanding people so hard" (Kafka, 1917).
![]() |
Clinical Implications and Limitations |
---|
LIMITATIONS
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
REFERENCES |
---|
Anonymous (1999) Bandolier 68, 6, 4.
Black, D. (1998) The limitations of evidence. Journal of the Royal College of Physicians of London, 32, 23-25.[Medline]
Bliss, M. (1999) William Osler: a Life in Medicine. Oxford: Oxford University Press.
Bradley, F. & Field, J. (1995) Evidence-based medicine (letter). Lancet, 346, 838-839.
Buston, K., Parry-Jones, W., Livingston, M., et al (1998) Qualitative research. British Journal of Psychiatry, 172, 197-199.[Medline]
Chiesa, M. & Fonagy, P. (1999) From the efficacy to the effectiveness model in psychotherapy research. The APP multicentre project. Psychoanalytic Psychotherapy, 13, 259-272.
Dening, T. & Lawton, C. (1998) New drug
treatment for Alzheimer's Disease. British Medical
Journal, 317,
945.
Easterbrook, P. J., Berlin, J. A., Gopalan, R., et al (1991) Publication bias in clinical research. Lancet, 337, 867-872.[Medline]
Elwyn, G. J. & Gwyn, R. (1999) Stories we
hear and stories we tell: analysing talk in clinical practice.
British Medical Journal,
318,
186-188.
Evans, M. & Greaves, D. (1999) Exploring
the medical humanities. British Medical Journal,
319, 1216.
Farmer, A. (1999) The demise of the published case report is resuscitation necessary? British Journal of Psychiatry, 174, 93-94.[Medline]
Ford, G. & Godber, G. (1998) Richard Henry
Lionel Cohen (obituary). British Medical Journal,
316, 1096.
Geddes, J. (1998) Evidence-based psychiatry: a practical approach. Psychiatric Bulletin, 22, 337-338.
Greenhalgh, T. (1999) Narrative based medicine
in an evidence based world. British Medical Journal,
318,
323-325.
Greenhalgh & Hurwitz, B. (1999) Why study
narrative? British Medical Journal,
318, 48-50.
Gregoriadis, L. (1999) Doctors call for better regulations as more turn to alternative cures. The Guardian, 24 August.
Jones, A. H. (1999) Narrative in medical
ethics. British Medical Journal,
318,
253-256.
Kafka, F. (1917) A Country Doctor. The Penal Colony, Stories and Short Pieces (translated by W. & E. Muir, 1961), p. 140. New York: Shocker Books.
Kelly, D. (1999) In conversation with David Tait. Psychiatric Bulletin, 23, 678-681.
Laugharne, R. (1999) Evidence-based medicine, user involvement and the post-modern paradigm. Psychiatric Bulletin, 23, 641-643.
Launer, J. (1999) Narrative based medicine: a
narrative approach to mental health in general practice. British
Medical Journal, 318,
117-119.
Manchip, S. & Morrison, C. (1999) A case report. British Medical Journal, 139, 1510.
Mays, N. & Pope, C. (2000) Qualitative
research in health care. Assessing quality and qualitative research.
British Medical Journal,
320, 50-52.
Medical Research Council (1965) Report by Clinical Psychiatry Committee. Clinical trial of the treatment of depressive illness. British Medical Journal, 1, 881-886.
Meyer, J. (2000) Qualitative research in health
care. Using qualitative methods in health related active research.
British Medical Journal,
320,
178-181.
National Institute for Clinical Excellence (2001) Guidance on the use of donepezil, rivastigmine and galantamine for the treatment of alzheimer's disease. Technology Appraisal Guidance No. 19.
NIMH-PSC Collaborative Study Group (1964) Phenothiazine treatment in acute schizophrenia. Archives of General Psychiatry, 10, 246-261.
Pope, C., Ziebland, S. & Mays, N. (2000)
Qualitative research in health care. Analysing qualitative data.
British Medical Journal,
320,
114-116.
Rosenthal, R. (1979) The file drawer problem and tolerance for null results. Psychological Bulletin, 86, 638-641.[CrossRef]
Sackett, D. L., Rosenberg, W. M. C., Gray, J. A. M., et
al (1996) Evidence-based medicine: what it is and what
it isn't. British Medical Journal,
312, 71-72.
Sakel, M. (1938) The Pharmacological Shock Treatment of Schizophrenia. Nervous and Mental Diseases Monograph Series No. 62. New York: Nervous and Mental Disease Publ. Co.
Warner, J. P. & King, M. (1997) Evidence-based medicine and the journal club: a cross-sectional survey of participants' views. Psychiatric Bulletin, 21, 532-534.
Watson, M., Haviland, J. S., Greer, S., et al (1999) Influence of psychological response on survival in breast cancer. A population-based cohort study. Lancet, 354, 1331-1336.[CrossRef][Medline]
Wells, K. B. (1999) Treatment research at the
cross roads: the scientific interface of clinical trials and effectiveness
research. American Journal of Psychiatry,
156, 5-10.
WKCM (1988) A. L. Cochrane (obituary) British Medical Journal, 297, 614.
Received for publication March 29, 2000. Revision received February 20, 2001. Accepted for publication March 7, 2001.
Related articles in BJP: