Medicine’s Uncomfortable Relationship with Math

Today is my 52nd birthday, an unpleasant reminder that at some point in the (nearer) future I can expect to become more reliant on the medical profession.

If we consider doctors as information processing systems (data in, diagnosis out), how are they improving compared to other information processing systems? “Medicine’s Uncomfortable Relationship With Math,” a 2014 paper from JAMA Internal Medicine shows no improvement over the past 30 years. 75 percent of the docs surveyed couldn’t figure out how to use data from a test for a rare condition: “Of 61 respondents, 14 provided the correct answer of 2% [likelihood that a patient was actually suffering from the condition, based on having tested positive]. The most common answer was 95%, provided by 27 of 61 respondents. The median answer was 66%, which is 33 times larger than the true answer.”


5 thoughts on “Medicine’s Uncomfortable Relationship with Math

  1. dguilder: One of my undergrad math/computer science profs spent a lot of time consulting with medical establishments. He commented that he saw more than a few PhD projects which, stripped of the medical terminology, were problems of basic statistics….

  2. philg, Happy Birthday! and, re your first paragraph

    My health plan for many years was to live a healthy lifestyle and let doctors starve in their office waiting for the phone to ring. It worked until about age 65, coincidentally the age Medicare became available. (I was fortunate to have good insurance and even more fortunate not to use it much.)

    Since then, life has revealed it’s great hidden plan: We accumulate whatever we can for 75 years or so, then we disburse a LOT of it to the sick care industry over very few years. And that’s the happy version – I’m actually much healthier than my cohort and thus able to see what’s happening. If the pitchforks ever come out, it will be over health care. OTOH, 75+ yo’s are not much of a threat even with pitchforks. And we may be the targets if young people ever catch on to the subsidies we get.

  3. Recently I read somewhere that goal of making predictions (and a diagnosis is a type of prediction) is not so much to come up with the right answer but to serve the interest of the person making the prediction. So in the ’70s, the “experts” that were predicting a nuclear winter or global cooling or whatever were objectively wrong but their predictions served whatever they were pushing at the time (nuclear disarmament or the removal of US missiles from Europe,etc. – a lot of these people were knowing or unknowing Soviet stooges) and that was more important than actually being right.

    So how does it serve the interest of doctors to admit that 49 out of 50 times the results of a particular test are just random noise rather than an accurate signal? This makes medical “science” seem pretty weak, doesn’t it.

  4. OTOH, 75+ yo’s are not much of a threat even with pitchforks.

    Oh yeah? 75+ yo’s vote. The baby boomers are passing 65, and they will dominate the vote. They are in the perfect position to screw Gen X, Gen Y, and Gen Z.

Comments are closed.