The nurses obviously couldn’t respond to it each time, but nor could they switch it off altogether, and it didn’t reset after any period of time.
My siblings and I took turns to gently hold his arm down on the side of the bed… which became just holding his hand, which I still miss.
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.
Props to Airbus for proper UX and information prioritization.
I was once in the recovery room with my wife. For some reason the sensor was having a very hard time reading her pulse. The normal bips would frequently fail. Too many failures in a row and the alarm would start it's EEEEEE scream we've all seen from Hollywood. It would shut up as soon as it managed to pick up a beat.
Hers was definitely not the only one in the room occasionally screaming. The nurses were completely ignoring it. Quite understandably so as it was obviously doing false alarms. But in a flood of false alarms like that are the real ones going to be noticed??
The reality was we knew what was going on just by listening to the alarms. I could predict which alarm was going to go off before it did and so I could safely (appear to) ignore them. I would only panic if an unexpected alarm went off (or happened in an unexpected sequence). It is possible the same situation was going on in the hospital.
Like residents who are getting a few hours of sleep over days worth of high-stress / high-stakes work, poor hand-washing between patients, and not clearly printing one's handwriting on prescription forms - all things that kill patients - doctors and hospital administrators just don't care enough.
For a profession that is supposedly so pure morality-wise - do no harm, patient privacy, etc - doctors are remarkably careless.
I remember an accident report. It was about a container ship which had a bad flooding incident in their engineering spaces. One thing the report pointed out that the engineers had ways to fight the flooding, but they were not doing them because they were playing whack-a-mole with all the alarms caused by the flood. If i recall correctly the engineers kept ignoring the waist deep and rising water and prioritised silencing the alarms. (And not because they were stupid, but just because the many independent blaring alarms task-saturated them.)
Alarms with incessant false positives are inherently dangerous. Sure, there's some threshold of false positives, under which we should still expect people to investigate all alarms. But above that threshold, how can we continue to blame the people involved? The hardware is at fault.
If the cost of a actual negative is 100 and the cost of an actual positive is 1. You'd expect there to be approximately 100 times more false negatives, because we want to be 100 times more sensitive to the costly negative condition.
I'm this sense, the alarms in hospitals make sense. Actual negative are very costly.
But this is a cold mathematical analysis that doesn't consider alarm fatigue and the cost of people learning to ignore the alarm. I wonder how to best model human nature in this calculation?
An optimal solution would require considering all alarms, and modeling the fact that every alarm given is another alarm ignored (assuming the hospital is operating at capacity, if it's below capacity the solution is easy, just manually check all alarms). This system might realize that the 4th "no pulse" alarm of the night for Alice would detract from the 1st "no pulse" alarm for Bob, and that Bob's is more likely to need attention. I'd be terrified to program such a system though, and from what I've seen in corporate programming environments, I'm not confident any company could get this right.
They really do not want false negatives because that gets them sued. Thus the system will be set up to err on the side of false positives--the current liability climate does not blame them for alarm fatigue.
Consider a local case (although it's possible it was overturned on appeal): Yes, the doctor was unquestionably playing loose with standard safety precautions. His behavior transmitted blood-borne infections. He died in prison which was well deserved.
However, the lawyers went hunting for some deep pockets. The manufacturer of the drug involved in the cross contamination. They made various size vials, including some that were bigger than would be used on one patient. This permitted the doctor to contaminate between patients and got them hit with a $250M verdict. (Never mind that had they truly only used clean needles with them like they should have there never would have been an issue. They used a new needle but the old syringe.)
That's the sort of insane legal pressure driving the garbage.
All of those sound superficially plausible to me, although I have my ideas on which are more likely... Would you even do an, um, incident post mortem for something like that or would it just be a statistic?
I think they could substantially improve patient outcomes by taking some tips from the best modern birthing centers, and make a quiet, relaxing, dimly lit, and peaceful environment at hospitals. I'd also say add some plants, natural (wood) surfaces and natural light, but realize that might make it hard to keep things sterile and private. It would make sense to create a rough schedule for each patient also with a consistent "left alone unless there is an emergency" time for sleep, etc.
I would imagine a calm and quiet physical environment would also reduce stress, fatigue, and improve performance of the medical staff themselves.
https://www.statnews.com/2016/10/14/icu-delirium-hospitals/
But it's tough to make improvements. Regular hospital design is (roughly) optimized for staff productivity. They need to be able to treat and monitor many patients simultaneously which requires clear sight lines, good lighting, and a high level of automation. A more humane hospital design would also require more staff at a time when we already have a severe shortage. Where would the funding come from?
My ward even managed to have the (networked digitally controlled, and do presumably very expensive) lighting set up so the night lighting was inside the curtains and shining directly into the bed spaces, and the main ward lights would come up if you touched the wrong thing (even the nurses weren't quite sure exactly what the proximal causes of lighting changes was). With the pumps alarming the whole time (about once per night, per patient, up to 20 minutes until resolution each time) plus all the other regular medical checks preventing any extended quiet time, it was absolutely exhausting at a very deep level.
If you're a multibillionaire then obviously you can just hire and equip your own private medical team that will focus 100% of their attention and care exclusively on you and your needs. The vast majority of the humans will never have that luxury. Normal people enter the system and are processed like everyone else.
Contrast with the GPWS warnings in aviation, which tells you what the problem is (TERRAIN TERRAIN) and what to do (PULL UP) in a progressively more alarmed voice as things get worse.
(Well.. Sometimes you hear of some particularly bright individuals who think the bank angle warning is a checklist item, but it's generally hard to get these wrong, compared to many other beeping warnings)
Aircraft systems are developed independently and added as options to planes. Which means they get swapped out, there are variants in capabilities, and multiple manufacturers involved.
> This can lead to cognitive overload when multiple systems issue verbal warnings simultaneously.
This is a known phenomenon on flights as well. There is some speculation it played a part in Air France 447. The plane technically _was_ telling the pilots the _precise_ problem they faced, but in the sea of other warnings they were entirely lost.
> tone alarms might be easier to manage and differentiate than multiple overlapping verbal warnings.
If you're a nurse, is the fact you have a ventilation alarm in one room and a temperature alarm in a different room that can be discerned without visual confirmation a useful feature in a health care setting?
I think the big difference is your flight has 2 people responsible for hundreds of lives. In the hospital you would hope the ratio would be more favorable.
I could imagine
ventilation? arrhyth-*C-chord*-ARRHYTHMIA! CHECK PUMP! HEART RATE!
I think GPWS can set windows of cases where an alert is given. Like, a terrain warning isn't much help when landing. Maybe there's something like that already for medicine, but a device who's job is to consume information from other devices, and only provide alerts based on rules the staff can configure before an operation, could be a thing that's useful.
https://looptube.io/?videoId=W5Z-d1Zx02o&start=77.1286764705...
buh-bump is cardiac stuff. wiSShhh... wooosSH is respiratory stuff.
Only thing is, I bet you can hear sounds similar to those in a hospital. The "beep beep" they put over it might not be enough. Still a really interesting research topic!
My take - the medical industry has too many barriers to competition, and it is too difficult for people who work with these things to do anything about it as well. It’s unclear who the buyers are at a hospital or how a startup could reach them. It’s also unclear what sort of interoperability (for example with Epic for charting) is needed. Regulations also make it difficult to get devices approved and investors are less likely to support a startup in this space.
If you push the button once, it would stop infusing drug into the patient.
If you push the button twice, it would EMPTY THE SYSTEM - as in, run the pump continuously, infusing all remaining drug into the system, at high speed.
We ran usability tests where we'd say to the nurse "wrong drug! stop! you're giving the patient the wrong drug!"
90+ percent of them did what any human would do - jab STOP over and over. Whoops, patient's dead.
In part because of our report Baxter was forced to recall[0] hundreds of thousands of the pumps and pay for their replacements with competitors' products. The stock dropped by 30% in a day. Sadly I didn't short it, or I'd be [checks notes] in jail.
[1] like drug libraries where sometimes the units were displayed, sometimes they weren't, and sometimes they were displayed in your "preferred" units even though the number being shown was in a DIFFERENT unit and the system didn't translate it, just showed the wrong value.
Wow this sounds so dangerous and so easy to predict.
Ideally you'd have a 1:1 (or better!) assignment between a single patient to a single nurse in critical care, 1:3 for patients that can't move around on their own (and thus need more assistance, even if it's just helping them to eat or go to the loo), and 1:5 to 1:10 for everyone else. The sad reality is that even in Germany, you have care home staff calling in the fire department to assist because there were just three staff in a night shift, having to deal with 170 patients.
[1] https://www.morgenpost.de/berlin/article242110812/Kurioser-G...
https://open.epic.com/Interface/
The FDA has a whole program office to assist startups with medical device innovation. They can help you a lot if you engage with them early in the development process and explain what you're trying to accomplish. Think of them as partners, not obstacles.
https://www.fda.gov/about-fda/cdrh-innovation/activities-sup...
This doesn't sound like the equipment's fault.
It’s a low volume but high margin business. Some of the issues were the constant fight against the factory not following design requirements to cut costs, knockoffs etc.
It seems to me that clear verbal alerts like "BLOOD PRESSURE VERY HIGH" could be more immediately understandable than tones. A hybrid system combining verbal alerts with alarm tones might be a good compromise for clarity and international usability.
If even 2 verbal alarms are going at the same time, it’s going to create a chaotic environment.
In a decentralized system, I think tones have less of an overlapping problem.
Unfortunately, I cannot find the article anymore.
Though a bit disappointing that there is no machine that goes PING! [1]
The alarm waveforms described are within the scope of the hardware standard guidelines, sufficiently common that application notes such as this exist. https://www.ti.com/lit/pdf/slaaec3 [ti.com]
A bit more than that. Certification is required in order to put your product on the market. Whether or not customers require it is irrelevant.