Of all the animals of prey, man is the only sociable one.
Every one of us preys upon his neighbour, and yet we herd together.
The Beggar's Opera: John Gay

Wednesday, 25 April 2012

Margin for error

A member of Clan Macheath had a blood test last week as part of a routine check-up. Returning to the surgery for the results, she was told she had 'a risk factor of 15%'.

The nurse seemed surprised to be asked for an explanation; "Look, it's here on the chart: your cholesterol gives you a risk of 15%, see?"

"Yes, but a risk of what, exactly?"

Turns out it's a risk of dying of a heart attack in the next ten years. Now it's not the easiest thing to express without giving the lady's age, but generally speaking this looked like not unreasonable odds of survival (her words, not mine); in any case, like most well-educated people, she knows that medical statistics - like boiled eggs - are best taken with a substantial pinch of salt.

But the nurse had not yet finished; under the circumstances, she said, she would recommend a course of statins. Not a good move; the patient is firmly opposed to blanket prescription and has a particular aversion to the idea of statins - and she's not the only one (see my post Statins for all and a death sentence).

The nurse was clearly disappointed; "Then we'll just have to try and manage it through diet". Manage what? The cholesterol reading was well within normal parameters; either the word 'risk' seems to have triggered a knee-jerk response or the 'statins-for-all' movement is alive and thriving in the hands of blinkered zealots.

One thing that interests me here is that our relative was not told whether to fast before the test. This is presumably because a study in 2009 found that 'cholesterol measurements are at least as good - and probably somewhat better - when made without fasting'.

But if that is so, why does the NHS website - reviewed in 2011 - still clearly advise 'Do not eat anything and only drink water for 10-12 hours before having blood cholesterol tests'? Are doctors, in fact, actively seeking raised readings by ignoring this advice?

This suggests at least some difference of opinion - and implies that the cholesterol test is rather more of a blunt instrument that those acting on its results would like us to believe. Given the very real possibility that statin side-effects will mask the symptoms of serious illness, I would question the ethics of prescribing on these terms.

The complexity of the human body means that diagnosis is not a exact science; to reduce it to box-ticking and percentages on a chart is to act under false pretences and with a dangerous complacency.


  1. One elderly lady I know went to the quack to have her prescription renewed to be told that the local hospital had declared her to be dead so she couldn't have it. If there is much of this kind of thing it could effect the statistics.

  2. Being as my new job means hanging about in GP surgeries , I'm picking up a bit of what nonsense is out there and what it is based on.

    This 'Risk Factor' is probably based on something called the Framingham Risk Score a study of a small American town , where diets and lifestyles bear no relationship to the UK at all or I would add, to any individual in particular.

    Much like a one size fits all BMI it is complete tosh (clinically obese toddlers?)

  3. I don't know what a "risk factor" is, but the accuracy of the cholesterol test is only about +/- 15% at the 95% confidence level for a single test.

    Even that sounds optimistic to me after reading the methodology. Not only that, but different labs will vary in the quality of their results both between themselves and internally on a daily basis.

  4. "The nurse was clearly disappointed; "Then we'll just have to try and manage it through diet"."

    I'm afraid that would have drawn the rejoinder 'We will, will we? You go right ahead and diet, then, ducky! I'll do as I please..'

  5. Demetrius, I'm sure there's plenty of that sort of thing going on all the time; I don't envy her the struggle to convince a bunch of NHS jobsworths that she is actually alive.

    PC/AKH, thank you; I suspect you - and now I - know rather more than some of the staff in this field who deal with patients every day.

    Julia, I think I can speak for everyone here when I say I would pay good money to watch what would happen if a nurse attempted to patronise you and get away with it.