| · PSNC
· Antibiotics
· Packaging
· Medicines use reviews
· Smoking cessation
· Statistics
· Statins
· Drug misuse
· Agenda for Change
· Palliative care
· The Society (2)
Letters to the Editor
|
Statistics
Risk reduction figures should be explicitly clear and differentiated
From Mr J. Bland, MRPharmS
The Broad spectrum
article by Chris Brewer (PJ, 11 March, p290) summed
up perfectly the practice in the NHS of prescribing drugs, such as the
statins, to benefit society as a whole, as opposed to guaranteeing benefit
for any one individual patient. In order for a patient to make an informed
choice, as to whether or not to take preventive treatment, the pros and
cons need to be clearly stated and understood by the patient. To this
end, risk reduction figures where available can be quoted and discussed
between prescriber and the patient.
This sounds simple enough but unfortunately it is not. The reason for
this is that risk reduction figures can either be stated as absolute
risk reduction (ARR) or relative risk reduction (RRR), there being a
world of difference between these two statistical terms. In general,
RRR figures are much higher than ARR figures, so if a patient is presented
with the RRR as opposed to the ARR this will undoubtedly have an influence
on the patient’s decision-making process. In my opinion it is the
ARR which should be quoted to the patient, because it is the most easily
understood and the most clinically relevant.
In Mr Brewer’s article it is an ARR (of 2.4 per cent) that is stated
as well as the equally understandable number needed to treat (NNT), which
is the inverse of the ARR. Now contrast how Mr Brewer presented his argument,
with issue number 21 of MeReC Extra, which reviewed the National Institute
of Health and Clinical Excellence appraisal of statins. MeReC states
that statins “significantly reduce cardiovascular morbidity and
mortality by between 20 per cent and 30 per cent”. No mention was
made that this was indeed a relative risk reduction so, statistically,
this statement was meaningless. To a lay person though, or even to a
health care professional not attuned to the niceties of statistical percentage
statements, quoting a reduction in death or disability of 20–30
per cent would be taken at face value as just that.
Is it asking too much for all reputable publications, including the PJ,
to make explicitly clear when quoting risk reduction figures whether
it is an RRR or an ARR being stated? Perhaps the PJ could make it a policy
always to state, if it is known, the ARR and NNT figures for future trial
results that it publishes, so that everybody, patients and health care
professionals alike, can easily understand the significance of the results.
Jonathan Bland
Specialist Clinical Pharmacist
Newark Hospital, Nottinghamshire
|
The
Pharmaceutical Journal