Amit is entirely correct in identifying the problem - that physicians
(and other clinicians) do not understand probability and cannot explain
its consequences. Every now and again I re-read the introduction to
"Clinical Epidemiology: a basic science for clinical medcine", where
Dave Sackett observes that "it dawned on him that applying these
epidemiologc principles (plus a few more from biostatistics) to the
beliefs, judgements and intuitions that comprise the art of medicine
might substantially improve the accuracy and efficiency of diagnosis and
prognosis, the effectiveness of management..." etc.
The point is that understanding these principles and being able to
explain them are fundamental skills that are required for clinicians who
claim to practice EBM. Unfortunately, the lack of such skills among most
clinicians has led to EBM being seen as the same as the implementation
of research findings, and to the choice of what research to implement
and how it is to be done being left in the hands of the pharmaceutical
and guidelines industries.
When clinicians are faced with (as in the new GP contract in the UK) a
set of quality indicators based on achieving targets for prescribing
specified drugs, and biological markers such as systolic BP and HbA1C,
problems are bound to arise. The first is that rather than reaching
shared decisions based on understanding of the risks and benefits of
interventions, doctors are pressured into promoting the interventions,
and have a financial interest in doing so. This seems to me to undermine
both the rights of the individual and the role of the doctor. The second
is that in treating patients as groups rather than individuals, personal
wishes and views are excluded. Different patients have different views
on risk and what measures they are prepared to take to avoid them. The
task of a clinician should be to assess the risk and benefits, explain
htem to the patient - and check that the patient understands - and to
help the patient reach a decision that suits them rather than the suits
on guideline committees (please forgive the irresistible pun).
That is why EBM skills are important, but I have to confess that I don't
know how their widespread dissemination can be achieved. They are not
easily acquired, and in a climate that sees "implementation of research"
and quality indicators as the key to improved care, it is clear that
teaching EBM to clinicians is not widely (if at all) supported in health
ministries and healthcare organisations. In fact, rather the reverse. so
much the worse for patients.
On that cheerful note, cheers and happy new year.
Toby
In message <[log in to unmask]>,
Ghosh, Amit K., M.D. <[log in to unmask]> writes
>. When explaining risk to patients, clinicians are typically focused on
>providing the facts about the risk (objective information), while patients
>typically are more interested in knowing how a bad outcome might affect them
>(subjective information). Patients, of course, also want to know the identity of
>the risk (death, disability, pain), its timing (early versus late), and the
>nature of the bad event (temporary, permanent), [JAMA 1999;281:1037-41.] When
>faced with a situation of conflicting evidence by several national guidelines,
>the practitioner is faced with a solitary dilemna. What do you do in the face of
>uncertainty and how do you explain probability. We have recently shown [Do
>medical students and physicians understand probability? Quart J Med
>2004;97:53-55. ] that physician might be have difficulty with probability and
>using natural frequancy format might be a better and simpler form of explaining
>risks.
--
Toby Lipman
General practitioner, Newcastle upon Tyne
Chair, Northern Faculty Board, Royal College of General Practitioners
R&D lead, Newcastle upon Tyne Primary Care Trust
Tel 0191-2811060 (home), 0191-2437000 (surgery)
Northern and Yorkshire Evidence-Based Practice Workshops
http://www.eb-practice.fsnet.co.uk/
|