Lies, Damn lies and Statistics

Medical research needs consistency in terminology. If Dr X calls it “mild hypertension” and Dr Y calls it “severe hypertension”, how can we agree on whether the blood pressure tablets are working?

So we need pidgeon holes, labels and agreed definitions.  In that regard the hardest part of medicine to slot neatly into categories is psychiatry.  The difficulties that this posed were recognised and in 1952 the American Psychiatric Association developed the Diagnostic and Statistical Manual – called DSM for short.  It was very useful for consistency in labelling diagnoses and of course for statistical purposes. It was perhaps less useful in knowing how best to treat the patient in front of you.

Over the past 60 years, DSM has undergone a series of major revisions and several minor ones. Diseases were added, others were removed. For example, until 1972 homosexuality was regarded as a psychiatric condition; since 1986 it has been removed completely from the DSM.

The latest complete revision of DSM (the 5th major rewrite – hence called DSM-V) was released last week. It has caused enormous controversy in the medical world.  Aspergers disease no longer “exists” (it’s now part of Autism Spectrum Disorder).  Disruptive Mood Disregulation Disorder (children who have more than three outbursts of rage a week for over a year) has been added.  Most controversially, grief is no longer an exclusion from the criteria for Major Depression – so someone grieving for loss of a relative for more than a few months may now be labelled as Depressed rather than grieving.

Critics of DSM-V say that it has been influenced too much by drug companies and by psychiatrists who are associated with them. Defenders say that the critics are misrepresenting the criteria, and even if the labels have been moved, there is no obligation on a doctor to use medication unless the individual patient requires it.

The lunch-time table-talk at Wellness on Wellington is that most doctors don’t like the changes. Whilst on occaission we may be required to use the labels “for diagnostic and statistical purposes”, our treatment of individual patients will remain as always – focussed on how we think we can best help you, regardless of labels.

If it’s Wednesday there must be new guidelines….

Once upon a time doctors were encouraged to use their own judgement. Now we are encouraged to follow guidelines (though they are really just somebody else’s judgement about what is best for the population as a whole, not specifically targeted at the patient sitting in front of us, and certainly not as immutable as the 10 Commandments).

So it’s pretty confusing for patients – and highly confusing for doctors – to realise that there are currently – count them – seven sets of “official” guidelines in Australia for what is best practice in testing for prostate cancer. And they all differ. It would be nice if the learned Colleges of Pathology, Surgery, General Practice, the NHMRC, Medicare etc could agree.

Let’s add one more to the list. A new study from Sweden (www.bmj.com/content/346/bmj.f2023) suggests that men should only be tested three times in their lives – around the ages of 45, early 50s and age 60. If each of those tests is normal, their lifetime risk of prostate cancer is very low. If any of the tests is above average, then testing more often is called for.

The theory (and evidence) is reasonably good. Whether this recommendation will be adopted by all the experts remains to be seen. And sometime in the next week or two, you can expect a new set of guidelines to be released…..