Skip to content

Confounding factors

The death rate for female patients was 8.15 per cent when treated by female physicians compared to 8.38 per cent when the physician was male.

It means that in every 1,000 patients treated, an extra two would be expected to survive if they were treated by a woman.

While the difference for male patients was smaller, female physicians still had the edge with a 10.15 per cent mortality rate compared with male doctors’ 10.23 per cent, a difference of around one death in a thousand.

There will be some – even many – and the researches validity will depend upon whether they deal with them properly or not.

A study in February found there is still “considerable gender inequality” in surgical fields in England and Wales and that “surgery is still a specialty that struggles to recruit women”.

Quite so, there are distinct gender inequalities across fields of medicine. Maybe the birds are more likely to be treating the easier stuff?

18 thoughts on “Confounding factors”

  1. Or maybe a woman is more likely to ask for a second opinion, so that a bloke gets blamed when the patient dies.
    Just one of a thousand variables to explain a 0.23% difference. How many millions would have to be in the study to make that statistically significant?

  2. If I put myself in the hands of my female GP for a sprained ankle, my leg will be off within a week.

  3. Funny how they make a fuss of a tiny difference in outcome between males and females, while completely ignoring the statistical elephant in the room:

    Visiting a doctor, regardless of gender, makes you almost 1 in 10 likely to die within 30 days.

    (Note that I’m using the same linguistic trick as the Journo who has pounded out this particular bit of Narrative.)

  4. It’s just possible that because they have to overcome structural inequality etc, the female surgeons who do make it through actually are very slightly ‘better’.

    But the difference is so small that it’s clearly not worth spending large amounts of time and money to find out.

  5. The gender story here relates to patient survival, not the surgeon. A difference of 200, not 20, basis points.

  6. This isn’t even confounding factors, it’s pure statistical noise.

    “Visiting a doctor, regardless of gender, makes you almost 1 in 10 likely to die within 30 days.”

    Ssshhhhhhh, don’t tell that to the hundreds of epidemiologists who churn out thousands of case-control “studies”, taking control patients out of the register rather than “on the register and bothered a doctor for anything other than what gets you in the case arm within [insert relevant timeframe]”.

    You will destroy literally hundreds of promising careers if you let that one out in the open!

  7. phillip,

    It could be highly statistically significant but some differences are so small they are totally irrelevant.

    This lesson was brought home for me when doing some training for a vet pharma company, that made vaccines and such for fish farming. In human clinical research, sample size (the number of patients) is a huge problem. Getting enough patients can be really hard, and more patients equals more cost, and more time until you get a result (and less time to make money from your medicine).

    I asked the fish farmers how they dealt with it, the answer being “our smallest phase 1 study includes a million animals. Everything we measure is statistically significant”.

    Their problem was not demonstrating statistical significance, but determining what differences actually meant anything. In the fish farming world, where yield, death rates, quality of meat and such, are what counts, less difficult than in human medicine, where mortality (while really, quite overwhelmingly, important) is not the only variable, and statistical noise needs to be ignored. Try repeating this experiment with the race of the surgeon, or patient, and see what happens….

  8. Anecdata – when I need my mental health medication, the female GPs in my surgery palpably don’t give a toss and usually ignore my requests or sometimes decide to randomly fuck me around on prescriptions and dosage for shits and gigs.

    Doing that to somebody with suicidal depression is not cool, by the way.

    If I’m lucky enough to get through to one of the male GP’s, I can get some medical treatment. But they’re all hiding behind their unhelpful receptionists and the shittiest online systems you’ve ever seen.

    Actual NHS mental health professionals, including the women, are great. But the overall system of how we provide healthcare is shit. It needlessly tortures people who are in terrible pain.

  9. BiG: “You will destroy literally hundreds of promising careers if you let that one out in the open!”

    It will no doubt not surprise you that I consider that a Good Thing.

  10. My guesses will be that the confounding factors are age and over-work.

    I have always preferred young professionals (doctors, dentists, lawyers etc).

    I would never use an overworked one, as they have every inclination to rush.

  11. Like female veterinarians who tend toward goldfish toenails and kitten sleep problems while the guys are wrestling 1,200 lb angry bulls and herds of ruminants doctors somewhat separate with females doing pediatrics, allergies and skin while orthopedics, cardiologists and the guy messing in your brain are, well, guys. Plenty of exceptions but certainly true within the margins cited n the article .

  12. Steve,
    Re GPs using “the shittiest online systems you’ve ever seen.”

    Just had my very first experience of one of these. Now know what Steve is on about. Had to specifiy the pharmacy I wanted my scrip sent to. Took me 24 hours to find something acceptable to the system and that only when I started providing non-self-evident answers!

  13. Dr K.A. Rodgers – shite, intit?

    The one my doctors use also only lets you raise a request for a few hours during the day. If they don’t come back to you by about lunchtime, you have to phone reception and wait in a queue for half an hour before you can speak to a receptionist who won’t help you. Computer says no if you try using their portal in the afternoon.

    Imagine if Tim turned his blog off for 18 hours a day. Sort of defeats the purpose of this internet thing, eh?

  14. “Imagine if Tim turned his blog off for 18 hours a day. Sort of defeats the purpose of this internet thing, eh?”

    Or if the only way to get an appointment to visit your proscribed supermarket was to ring in at 8am and play “fastest finger”. If all the slots had gone by the time you got through, no problem, just try again the following day.

    That truly is just how utter fucking SHITE our NHS is……

Leave a Reply

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