News for hackers. Stuff that matters.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by qsort on Wednesday September 18, 2024 @10:16AM (#41580019)
"The deterioration prediction model was a time-aware multivariate adaptive regression spline (MARS) model"

https://doi.org/10.1503/cmaj.240132

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by tantalor on Wednesday September 18, 2024 @10:20AM (#41580058)
Thanks for posting this! Much better source than CBC article.

I found this interesting:

> 1 truly alerted patient for every 2 falsely alerted patients was deemed an acceptable number of false alarms

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by shadowgovt on Wednesday September 18, 2024 @10:34AM (#41580200)
Interesting, and I think it makes sense.

In an ideal world, the nurse to patient ratio would be high enough that patients could be seen on regular rotation frequently. I've never been in a hospital where this was the case. So a system that can correctly prioritize resources for critical cases even if it's pulling resources away from non-critical cases will probably result in a net improved outcome.

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by rscho on Wednesday September 18, 2024 @11:14AM (#41580770)
With such a false positive rate, I expect the staff on site to find ways to negate the effect pretty quickly.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by Tostino on Wednesday September 18, 2024 @11:48AM (#41581331)
I don't know...If every 3rd time I was alerted it was some relatively serious issue, vs how often there is a serious issue when just doing rounds that you stumble upon, I'd think that would be a pretty good alert rate compared to the norm. But then again, I'm not in healthcare.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by rscho on Wednesday September 18, 2024 @12:00PM (#41581559)
Essentially, it depends on workplace integration, i.e. how much effort it takes to discover the alert trigger. From personal experience, I'd say the upper limit of inconvenience is 'click the alert' on mobile, and 'move mouse on alert label to see tooltip' on desktop. Anything more will be quickly discarded, especially if it involves a popup or opens a new window.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by firejake308 on Wednesday September 18, 2024 @02:02PM (#41583334)
While alarm fatigue is a real thing, the finding from this study is that they didn't, which is what matters.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by inglor_cz on Wednesday September 18, 2024 @03:49PM (#41584624)
2 in 3 isn't such a terrible false positive rate.

If your home alarm caught one real burglar per each three occassions it triggered, I bet you wouldn't develop alarm fatigue. I certainly wouldn't.

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by rscho on Wednesday September 18, 2024 @03:54PM (#41584692)
Do you get burglars multiple times a day? I bet not...
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by inglor_cz on Wednesday September 18, 2024 @03:55PM (#41584699)
Yeah, but potential death of a patient is on a similar level of seriousness.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by rscho on Wednesday September 18, 2024 @04:21PM (#41584992)
Most of those alarms will warn about trivial everyday results, that may (with rather low probability) cause death down the line. My bet is they'd get mostly ignored very quickly.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by jampekka on Wednesday September 18, 2024 @10:51AM (#41580433)
GOFAI from 1991! Wasn't AI back then though.

https://en.m.wikipedia.org/wiki/Multivariate_adaptive_regres...

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by 0cf8612b2e1e on Wednesday September 18, 2024 @12:04PM (#41581636)
Fun bit of trivia (though depressing) from the wiki

  The term "MARS" is trademarked and licensed to Salford Systems. In order to avoid trademark infringements, many open-source implementations of MARS are called "Earth".
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by blitzar on Wednesday September 18, 2024 @12:12PM (#41581748)
Multivariate adaptive regression spline saw the hype and pivoted to AI
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by lukeinator42 on Wednesday September 18, 2024 @11:23AM (#41580893)
I'm also shocked at how readable this wikipedia article is relative to most articles about statistical methods.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by IshKebab on Wednesday September 18, 2024 @11:52AM (#41581400)
Wow you're right. I mean it's all maths articles on Wikipedia, not just statistics.

I think there are two causes of Wikipedia maths articles' general awfulness:

1. They're probably written by people that just learnt about them and want to show off their superior knowledge rather than explain the concept.

2. The people writing them think it's supposed to be a precise mathematical definition of the concept, rather than an easy to understand introduction. It's like they're writing a formal model instead of a tutorial.

Often the Mathworld articles are a lot better than Wikipedia, when they exist at least.

Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by pj808 on Wednesday September 18, 2024 @11:00AM (#41580580)
MARS is one of the smartest methods: intuitive, hard to break, easy to quantify. Friedman was a savant.
Re:AI tool cuts unexpected deaths in hospital by 26%, Canadi... by meindnoch on Wednesday September 18, 2024 @12:14PM (#41581790)
Lol. "AI" strikes again.