Issue 14 / 18 April 2011

LIKE many patients, when faced with a difficult health decision I have occasionally asked a treating doctor, “What would you do in my position?”.

When a doctor has prefaced a recommendation with the words “If it were me …”, I have tended to give more weight to the advice than I might have if it had been couched in less personal terms.

But an American study on the differences between recommendations doctors give to patients and the treatments they would choose for themselves suggests it really isn’t that simple.

In fact, as research suggests, we may make better decisions on behalf of other people than we do for ourselves, so it might make more sense for doctors contemplating their own treatment options to ask, “What would I recommend to a patient?”.

In this latest study published in the Archives of Internal Medicine, the primary care physicians surveyed were more likely to choose a treatment option with a higher mortality risk but lower risk of complications for themselves than they would recommend to a patient.

Of course, it’s not necessarily irrational to prefer death to disability, but the researchers’ findings are at the very least thought-provoking.

In one scenario, the doctors had to choose between two treatments for colon cancer. Both treatments led to 80% of patients being cured with no complications, but the first led to a further 4% cured with complications while the remaining 16% died within two years.

The second treatment option carried no risk of complications but 20% of patients died within two years.

On the face of it, the first looks like the better option, with a lower risk of mortality and with complications that the researchers had established were preferable to death for more than 90% of respondents. The complications were a 1% risk each of colostomy, chronic diarrhoea, intermittent bowel obstruction and wound infection.

Yet 38% of doctors chose the second option – with its higher mortality risk – when asked to decide on a treatment for themselves, compared with 25% selecting that option when asked to make a recommendation to a patient.

There was a similar discrepancy when doctors were asked to choose between two treatment options for bird flu. This prompted researchers to suggest that some cognitive biases might wield less influence when we make decisions on behalf of others rather than for ourselves.

Two biases that they suggested might affect clinical decision-making were “betrayal aversion” (the feeling that harm is worse when caused by something designed to prevent harm) and “omission bias” (where harm caused by an action is seen as worse than harm caused by a failure to act).

So, where does that leave the “What would you do, doctor?” question?

An oncologist once told me it’s a question she chooses not to answer, explaining to patients instead that what is right for her might not be right for them.

But for doctors who do answer the question – and the many patients who ask it – it’s worth keeping in mind that the response may not be as straightforward as first thought.

Jane McCredie is a Sydney-based science and medicine writer.

Posted 18 April 2011

2 thoughts on “Jane McCredie: Answering the tricky question

  1. Martin Bailey says:

    A friend had significant morbidity following an attempted lung biopsy for suspicious nodules. He was advised by the treating specialist to re-present for another biopsy.
    My opinion of his local GP rose enormously when he responded “talking as your GP I must say that is very good advice- of course you must go for another biopsy- BUT talking as your friend I am telling you I would not be having it!”
    Two years later my friend is well , with a presumed diagnosis of sarcoidosis, and still hasn’t had further biopsy.

  2. EWB says:

    After the removal of a secondary melanoma metastasis in my parotid, I was told that radiotherapy would reduce the chance of recurrence from 30% to 15%, but the radiotherapist said, if it was him, he would not have the RT.
    I did, it was a very boring four weeks, but I am still here, with no evidence of recurrence, 5 1/2 years later. Was he correct? Did I have 4 rather trying weeks, or was it worthwhile?

Leave a Reply

Your email address will not be published. Required fields are marked *