Sick people die more often: not very surprising

ASSOCIATION IS NOT CAUSATION.

Now that I’ve cleared that up, let us examine the paper causing the headlines today . “Over 65s who take more than one medicine should consult their doctors. Taking some commonly used medications in combination leads to an increased risk of death or brain impairments in over 65s” says the Guardian. The Telegraph goes a bit further. “Combinations of popular drugs could prove fatal in elderly.” The actual paper, which is in the Journal of the American Geriatrics Society, is apparently about to go up online (as I write, it does not even appear as an abstract there or on the University of East Anglia website; it’s just a press release and a list of the offending drugs.)

Basically, through looking at a database of 13,400 older people living at home or in carehomes, they found a higher risk of cognitive impairment or death the more anticholinergic drugs were taken.

Does this mean that the medications caused the cognitive decline or death?

No. It means that if you have high blood pressure, dementia, chronic pain, depression, heart failure or angina (which the medications were prescribed for in the first place) then you have a higher risk of death. This isn’t surprising. What we really need to know is, if you have heart failure for example, you are better off taking medication than not. There are many studies looking at these kinds of questions and research which aims to answer this. But in this study, you can’t conclude that the prescriptions are causing more deaths; this study cannot answer than question.

The ‘cognitive decline’ the researchers found relating to anticholinergic drug use was actually small, between half a point and a point of the MMSE examination, a 30 point tool used to measure cognitive function. In fact, a score of above 25 is generally considered normal, and the lowest score in the study was 24.9.

Of course, people should not be on drugs they don’t need, or where the balance of risk is unfavourable. It isn’t usually that straightforward, which is another reason for thoughtful, evidence based, discursive prescribing; and not reflex protocols.

2 Responses to “Sick people die more often: not very surprising”

  1. ptl June 24, 2011 at 1:48 pm #

    Yeah. But the side-effects of anti-muscarinics and their (related?) cardio-toxic potential are known — OK the side-effects are… . More than one such medicine= higher anti-muscarinic dose. The cognitive decline finding was surprising to me, but, I’m not a doctor/pharmacist/etc.. A cardiac effect would not surprise me in the least.

    What I do – as a patient who has had a bad reaction to a potent anti-muscarinic — is note what doctors say about side-effects. If those sound like anti-cholinergic ones, I ask if the drugs are anti-cholinergics. (GPs have had to look them up to check.) Then, well, I decide whether the benefit’s worse a possible risk.

    (I check everything medical. I was on MAOIs. Phenomenally safe drugs for a patient prepared to check each and every drug doctors prescribe.)

    I do agree that restricting such drugs would be foolish. Urging all doctors to check these matters is worth it even though it will annoy some of the good ones. Sorry.

  2. Margaret McCartney
    Margaret mccartney June 24, 2011 at 2:28 pm #

    No disagreement that less side effects are better for patients. However this study has many limitations – we don’t know that changing the drugs will lead to less adverse events. And I also think we use too much medicine in general, courtesy of a pharmaceutical industy that advertises directly and indirectly. What we shouldn’t do is read more into retrospective data which is safe, and which is not designed to tell us about causation.
    there is a bigger debate to be had about increasing polypharmacy and the lack o data in the longer term: we are unlikely to get good unbiased data about unintended effects of statins post 15 or 20 years use.
    On the MAOIs , they do seem to work well for some people but seem hardly ever prescribed now.

Leave a Reply