[ijustlovemath]

We're building a life critical medical device, and I haven't seen this mentioned, so I thought it was worth contributing:

The use of these alarms is not something imposed by the manufacturers, but by the standards, eg 60601, 62304 etc. For devices involved in diagnostic, or more importantly interventional care, you are required to have alarms within certain auditory and visual thresholds, and a lot of them have mandated silence times (in a life critical system, you can only silence a true alarm for 120 seconds at a time).

Then again, "ALARM" as dictated by the standards means something truly emergent, though the wording can feel a bit fuzzy at times. Trust me, alarm fatigue is a known phenomenon to these manufacturers, and theres been a recent trend (with, eg, the Dexcom G7) of giving users more control over delaying alarms, silencing them until you can respond etc etc, which has its benefits, especially as quality of life is concerned.

You'll have a hard time convincing the FDA of this for critical devices like those found in hospitals though.

[jimmySixDOF]

The airline industry went through this too and have moderated requirements to be more understanding of who it's consumers are and when. One of the big near miss cases was QF32 out of Singapore where they had over 50 alarms to deal with in addition the the emergency at hand. Alarm pollution is a real UI/UX dilemma.

[hejdufufjrj]

In an airplane at least all the alarms are integrated, but in a hospital room you'll have 15 devices from 7 manufacturers spanning 5 generations.

[hi-v-rocknroll]

58 faults on the ECAM. Source: https://youtu.be/a-4FBN8OTkk

Props to Airbus for proper UX and information prioritization.

[netsharc]

Huh, that title smells of tabloidization. I know about this incident, the "mid-air explosion" have to do with an uncontained disintegration of 1 turbine (with shrapnel flying that breached the wing, luckily not the fuselage), but the title makes it sound the whole plane exploded...

[amelius]

Maybe someone can train an AI to decide which alarms need immediate attention, given N staff members available.

[ijustlovemath]

The FDA would not let this fly. To get a device in the hospital, you have to enumerate EVERY failure mode that you can reasonably protect against, as well as the ones you can't. Some of these failures are crucial enough that they qualify to be required to implement an alarm for.

There's a reason everyone is so loud in the hospital, it's because we have to be to be there in the first place.

[wolrah]

> Maybe someone can train an AI to decide which alarms need immediate attention, given N staff members available.

The words you've used could hypothetically mean some future artificial general intelligence that does not currently exist and there is no guarantee will ever exist, especially within the lifetimes of those participating in this thread. That could obviously be quite good.

"AI" as currently defined by marketing and pop culture to mean machine learning, large language models, etc. should never be allowed to make a medically important decision. We've already seen beyond any reasonable doubt how risky it is to even treat them as a natural language search engine, the idea of handing over life-or-death decisions to them is literally insane.

[throwaway765123]

Yikes I hope this is tongue-in-cheek, I definitely don't want a statistical process deciding whether to surface a life-critical alarm to healthcare staff

[amelius]

If it statistically saves lives?

It's the same as allowing full self driving cars which on average are safer than human drivers but sometimes accidentally drive into a fire truck because they couldn't train an image classifier to more accuracy than 99%.

[hydrolox]

this is very true with diabetes equipment since there is constant alerts (for example from insulin pumps) of low battery, undelivered insulin, etc. I think it definitely helps to give users the right amount of control if it's non life critical like a CGM or for insulin delivery.

[ijustlovemath]

Insulin delivery is considered life critical by the FDA, because the failure modes of those devices can involve coma, brain damage, and even death. Some of those alarms will still be hardcoded, and for good reason!

[hydrolox]

I agree, but something like "incomplete bolus" doesn't really make sense to me and I think those types contribute to alarm fatigue. The key issue with T1 diabetes is hypoglycemia, which can cause acute damage and death, so in theory not having insulin on isn't as big of an issue (assuming the patient is actually trying to control their disease and checks blood sugar etc, in the alternate case the alarms probably don't help much). Of course, I agree something like a hypoglycemia alarm is important