The bad news for many medical students and doctors is that these days one cannot
practise medicine well without some understanding of statistics. Most papers in
medical journals use statistical techniques to summarise or interpret observations,
and even at clinical meetings it is difficult to escape at least occasional reference
to standard deviations and P values. A doctor with no knowledge of statistics is
unable to evaluate much of the scientific information that is crucial to the optimal
care of patients.
Increasingly, a similar challenge is now faced also by other health professionals
such as nurses and physiotherapists. This dependence on statistical methods is reflected
in undergraduate and postgraduate curricula.
The good news is that clinicians need not be mathematicians to use statistics.
In the same way that a report of a plasma creatinine concentration can be interpreted
and acted on without a detailed understanding of the laboratory methods underlying
the assay, so it is not necessary for a clinician to understand all the mathematical
intricacies of a statistical calculation in order to apply its results. Just as
we normally trust the biochemist to use an appropriate analytical method, so we
may put faith in medical statisticians to get their sums right. It helps if clinicians
can carry out simpler tasks such as calculating means and standard deviations, but
more important is the ability to communicate with statisticians—to formulate a problem
in such a way that a statistician can advise on an appropriate analytical method;
to recognise and assess any biological assumptions that are inherent in the statistical
analysis; and particularly to understand the results of statistical calculations
when they are presented.
This book is aimed at doctors, medical students and other health professionals
who view statistics as a necessary evil. It is not intended as a manual for those
wishing to carry out their own statistical analyses beyond the simple summarising
of data that might be required for a clinical meeting. Rather, it sets out to explain
the principles of statistics that must be understood in order to read journals and
practise clinical disciplines competently, and it does this without using any mathematics
beyond the level needed for school leavers. For readers who prefer a more detailed
and mathematical approach, many other texts are available, some of which are listed
on page 106.