3 ms·
Yeah, survivor bias was the first thing that leapt to my mind when she mentioned the 'zebra effect'. The doc is 100% correct in saying that this patient almost
by taneq 8y ago
Yeah, survivor bias was the first thing that leapt to my mind when she mentioned the 'zebra effect'. The doc is 100% correct in saying that this patient almost certainly doesn't have the super rare disease (although maybe he should have revisited that heuristic when the other, more likely diagnoses fell through). He was right the other 9 times he told people their odd-sounding symptoms were just a common disease presenting oddly... but people don't write news articles about horses.
- sterlind 8y agoSpeaking as an officially-diagnosed zebra (Ehlers-Danlos syndrome), it took a year of doctor searching until I found a specialist who knew what to look for.. and it started with someone in a powerchair I met at a party remarking that my skin felt velvety like his. Many doctors are bad at conditional probability. [0] 2% of the healthy population is hypermobile, and hEDS is ~1:5000. But P(hEDS|atrophic scars, spontaneous dislocations, 7/9 Beighton score, paralysis attacks) is much higher. There's a weird stigma against zebras seeking diagnoses because you almost have to do your own research to be diagnosed properly. 0. https://opinionator.blogs.nytimes.com/2010/04/25/chances-are/ https://opinionator.blogs.nytimes.com/2010/04/25/chances-are...
- taneq 8y agoYeah, I remember reading something about that in some introductory Bayesian statistics article. They asked a whole bunch of doctors the P(cancer | positive screening test) question and their answers were way off. Which is fair because it's pretty unintuitive, but also not good because they're using that intuition to make life-and-death decisions. By weird coincidence, I think one of my friends was recently diagnosed with EDS after a long, long search to find out what was actually going on with her.