Health

Sometimes, though, the balance of risks is less clear-cut. Doing nothing about a condition — a hernia, for example — may lead to problems: the hernia may become strangulated, leading to serious infection. Yet these complications may never happen, and that has to be balanced against the fact that the operation is not risk-free.

The question of hard judgment calls was highlighted recently with the withdrawal of Vioxx. This had been marketed as a “lower-risk” alternative to traditional, cheaper anti-inflammatories such as ibuprofen, as the risk of it causing bleeding from the bowel (a significant side-effect) was lower. However, further research showed an increase in heart attacks in patients using it — although the knowledge that led to the drug’s withdrawal was not entirely new. The non-profitmaking Cochrane Collaboration had already published a review showing that Vioxx decreased the amount of gastric bleeding but had an association with increased heart attacks, and the British Medical Journal said that this category of drugs had “serious potential harms that have been minimised in reports of research sponsored by drug companies”.

So should I have been telling patients routinely about this possibility when prescribing Vioxx? It would, after all, have kept the patient fully informed about what they were being prescribed.

This should be a good thing; for, as Richard Smith, then editor of the British Medical Journal, wrote in July: “Very few people attend a doctor thinking that they may come out worse than when they went in.” There is potential for harm in any medical intervention — even, as Smith writes, in “a throwaway comment or a test ‘just to be sure’. ”

But if patients had been informed about this potential side-effect, it would have “covered” the doctor — ie, if I explain all possible side-effects, the theory goes, I will make myself less likely to be sued successfully if a serious side-effect arises. And while getting better information about medication can only be good, in some instances, if the motivation arrives because of the threat of litigation, it comes at a price — the price of “cover-your-arse ” defensive medicine.

Let me explain. Headaches are common, but there are some rare and serious causes. If you offer a head scan to everyone with a headache, you will rarely pick up a treatable abnormality. But suppose the scan is reported as borderline, and is eventually proved normal after many tests and long months of anxiety. Has the doctor, in ordering a “just-to-be-sure” head scan, really done that patient a favour?

But in the case of deciding about treatments, maybe patients should be able to decide, as happens in several US states, what level of risk they wish to be informed about. Patients can choose the information they are given about the complications of elective surgical procedures — from every vanishingly rare complication to just “the bare bones”.

But even getting decent information to decide on can be hard. There is no compulsion for pharmaceutical firms to publish all their research, which can create bias. And even when studies are published, results can be framed in such a way that it is very difficult to know one’s personal risk of developing problems.

You might be warned that using a drug quadruples your risk of developing, say, kidney problems — this is your “relative risk”. However, the “absolute risk” is a more representative — and less frightening — statistic. For example, for the same drug, the absolute risk of getting kidney problems may be only 2 in 10,000 — a “quadruple” rise from 0.5 per 10,000. Less dramatic, but more informative. This, and the concept of “ number needed to treat”, deserves to be used much more.

Ultimately, it is for the patient to decide what level of risk he or she wants to know about and accept. I think there are fantastic opportunities to educate ourselves about the likely benefit of treatment via the internet. I recommend the Cochrane Collaboration site, www.cochrane.org, and Bandolier (www.jr2.ox.ac.uk/bandolier/), which distill statistics and research into unbiased, manageable chunks of knowledge that are, quite simply, pearls. If you have only ten minutes with your doctor, you have to use it effectively, and this helps even before you get there.

Doctors shouldn’t absolve themselves from the responsibility of giving advice, but patients should at least try to make up their minds as to what kind of risk they wish to take.