Latest suicide pub (w/ @drkatebentley@TaylorABurkePhD@jorsmo@DrJBStephens) provides strong evidence to 1) counter the common claim that clinicians can't predict suicide risk, demonstrating the value of clinical judgment, yet also 2) massive improvement through ML on SRA items
While clinicians can stratify suicide risk above chance levels, predictive accuracy for future suicide attempts significantly improves when using machine learning to incorporate comprehensive clinical assessment data.
https://t.co/QvbBKljIa8 @drkatebentley
@evavivalt In my experience, if you ask an LLM before asking a presenter, you can sometimes get an answer to help you think about a paper without needing to take up everyone else’s time. And this scales to multiple personal questions, which can lead to a better Q for a time-limited session
@YunyuXiaoPhD The thing is, rural clinics or safety-net hospitals by their nature do not have the sample size to differentiate the likely performance differences from multiple validated models. Thus the broader strategic question is: why focus on small sites before large ones?
Realized I wrote this exactly 8 years ago
When you go about testing everyone for everything, you don’t create a world of healthy people; you create a nightmare where everyone is sick.
https://t.co/viTin3hT6O
Here’s what happened when Korea promoted widespread thyroid ultrasounds for the general population.
Of course those idiots didn’t have Silicon Valley so I’m sure this time will be different
@bself@midjourney Unclear. False positives may lead to procedures with little average benefit and real harms in some %, sufficient to outweigh the rare asymptomatic patient who finds a needle in a haystack. You at least need a net benefit curve analysis to establish that - and ideally an RCT