Intended for healthcare professionals

The Nhs's 50th Anniversary Getting research findings into practice

Making better use of research findings

BMJ 1998; 317 doi: (Published 04 July 1998) Cite this as: BMJ 1998;317:72

A rigid following of dogma will not be best for all patients

We read with interest your series on ‘Getting research findings into
practice' (Ref 1). As day-by-day GPs who also run a training scheme, we
found it most thought provoking.

The most difficult area of general practice is arriving at a
‘diagnosis'. The articles seem to take off after the crucial decisions of
general practice have been taken, after the ‘art' has been practiced. We
are concerned that with the emphasis on decision analysis after a
diagnosis has been formed, less emphasis will be put on arriving at the
correct diagnosis.

In 1948 J. Ryle (Ref 2) wrote ‘The three main tasks of the
clinician….. are diagnosis, prognosis and treatment. Of these diagnosis is
by far the most important, for upon it the success of the other two

Often the diagnosis of disease is much less objective than we would
like, and not based on one simple test; consider angina, asthma, epilepsy
or depression as examples. Patients often come to see us with a loose
amalgam of symptoms, of which we try to make sense with a mixture of open
and closed questions, postures, pauses and tests.

We have found it helpful to use a framework to explain to doctors on
our course why, as GPs, our knowledge of patients is so important. We
feel that:

Every patient has their own inherent set of predictive values and
that these may change in time depending on the patients experience.

A patient who consults infrequently is more likely to have disease
with a given symptom than a patient who consults frequently in whom
numerous investigations over the years have failed to identify significant
disease; we learn to assign increased validity to the concerns of an
experienced mother; doctors may reflect on their student days when they
feared that every minor symptom they developed heralded the onset or a
rare and life threatening syndrome, yet as qualified doctors they may
ignore serious problems in themselves (Ref 3).

Predictive values of individuals can help us determine the best
course of action and enhance resource utilisation. They must not be used
to deny patients tests or referrals, but they help understand what is
going on, and pre-empt results. Indeed the concept is likely to
facilitate earlier detection of disease in patients who appear to have a
high predictive value. Patient with thin files merit great caution.

Students and housemen are often taught the blunderbuss approach to
diagnosis; ask enough questions and organise enough tests and the ‘answer'
will appear. We hope to help doctors become aware of the bigger picture
and not to rush in to assign labels of disease at an inappropriately early
stage. We recognise that EBM demands the right questions and correct
interpretation of answers, but a rigid following of dogma will not be best
for all patients.


1. Getting research into practice. A series of 8 articles. BMJ
1998;317 July-Aug.

2. Rye J. The natural history of disease. 1948 2nd
Edition Oxford University Press.

3. Allibone A, Oakes D, Shannon H. The health and health care of doctors.
J R Coll Gen Pract 1981;31:728-734.

Competing interests: No competing interests

20 November 1998
Andrew Ross