Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Medicine AI Technology

Hospitals' New Issue: A 'Glut' of Machines Making Alarm Sounds (fiercehealthcare.com) 77

"Tens of thousands of alarms shriek, beep and buzz every day in every U.S. hospital," reports Fierce Healthcare -- even though most of them aren't urgent, disturb the patients, and won't get immediate attention anyways: The glut of noise means that the medical staff is less likely to respond. Alarms have ranked as one of the top 10 health technological hazards every year since 2007, according to the research firm ECRI Institute. That could mean staffs were too swamped with alarms to notice a patient in distress or that the alarms were misconfigured. The Joint Commission, which accredits hospitals, warned the nation about the "frequent and persistent" problem of alarm safety in 2013. It now requires hospitals to create formal processes to tackle alarm system safety...

The commission has estimated that of the thousands of alarms going off throughout a hospital every day, an estimated 85% to 99% do not require clinical intervention. Staff, facing widespread "alarm fatigue" can miss critical alerts, leading to patient deaths. Patients may get anxious about fluctuations in heart rate or blood pressure that are perfectly normal, the commission said....

In the past 30 years, the number of medical devices that generate alarms has risen from about 10 to nearly 40, said Priyanka Shah, a senior project engineer at ECRI Institute. A breathing ventilator alone can emit 30 to 40 different noises, she said... Maria Cvach, an alarm expert and director of policy management and integration for Johns Hopkins Health System, found that on one step-down unit (a level below intensive care) in the hospital in 2006, an average of 350 alarms went off per patient per day -- from the cardiac monitor alone.... By customizing alarm settings and converting some audible alerts to visual displays at nurses' stations, Cvach's team at Johns Hopkins reduced the average number of alarms from each patient's cardiac monitor from 350 to about 40 per day, she said.

Hospitals are also installing sophisticated software to analyze and prioritize the constant stream of alerts before relaying the information to staff members.

This discussion has been archived. No new comments can be posted.

Hospitals' New Issue: A 'Glut' of Machines Making Alarm Sounds

Comments Filter:
  • Honestly... (Score:5, Funny)

    by dex22 ( 239643 ) <plasticuser@NosPaM.gmail.com> on Saturday December 07, 2019 @12:43PM (#59495294) Homepage

    Honestly, if we can't solve the problem of nag screens on car radios telling us not to use them when moving EVERY SINGLE TIME WE START THE CAR then solving a beeper problem in a hospital might be a bit beyond us.

    • by burtosis ( 1124179 ) on Saturday December 07, 2019 @12:49PM (#59495304)
      I usually get distracted trying to clear that message by a “keep your eyes on the road” sign.
    • If you drive a Toyota, after three hours on the road the glass cockpit status display will be replaced by a coffee cup, which is supposed to prompt you to take a rest.

    • YES! I just got a new car and the nag screen covers the backup camera for 10-15 seconds.
    • Those are not warnings for the drivers, those are merely protection against lawsuits.

    • And we couldn't managed to throw an error in the early days of smartphone cameras that tried to record in 9:16. Now all my real world disaster porn is forever cropped...

    • Cars actually have the same problem these hospitals have, but at a smaller scale. There's so many damn things beeping and lights blinking and crap trying to get your attention that after a while you just start to tune it all out. That's one of things I like about my older car, which generally doesn't make noise or light up something on the dashboard unless there really is something that does need immediate attention.

      • Re: (Score:3, Funny)

        by therealbev ( 679533 )

        My 69 Ford obligingly lit up the HOT light maybe 5 seconds after I saw steam issuing from under the hood.

      • by tlhIngan ( 30335 )

        Cars actually have the same problem these hospitals have, but at a smaller scale. There's so many damn things beeping and lights blinking and crap trying to get your attention that after a while you just start to tune it all out. That's one of things I like about my older car, which generally doesn't make noise or light up something on the dashboard unless there really is something that does need immediate attention.

        And yet, people still left their blinker on back then too. Heck, blinker left on alarms are

    • I'm a lot late to this discussion but I need to vent just the same:

      If that damned startup nag screen telling me to drive safely would actually go away when I press the Continue button I wouldn't get so pissed at it.

      The damned DING it makes anytime the temp goes below 37F where it says "IT MIGHT BE ICY BE CAREFUL" is also fucking irritating.

      OK I'm done now...

  • Issues (Score:5, Insightful)

    by ceoyoyo ( 59147 ) on Saturday December 07, 2019 @12:46PM (#59495300)

    We do love causing problems for ourselves in creative ways. Which we then get to solve. But the solutions cause some more problems so....

    That way almost nobody actually has to do any work.

    • We do love causing problems for ourselves in creative ways. Which we then get to solve. But the solutions cause some more problems so...

      Perhaps a slight change would be in order: "We do love causing problems for other people in creative ways".

      And then there is Sevareid's Law: "The chief cause of problems is solutions".

    • We do love causing problems for ourselves in creative ways. Which we then get to solve. But the solutions cause some more problems so....

      That way almost nobody actually has to do any work.

      Actually, it does cause work. If you can't find a solution, there's money to be made in prolonging the problem.

  • by Dunbal ( 464142 ) * on Saturday December 07, 2019 @01:06PM (#59495328)
    And of course there's the story of the overworked intern (yours truly) who was on call at the ER and was called to the ward by nursing to confirm that a 90 patient had indeed died. As per procedure said intern hooked the patient up to a monitor to confirm the lack of electrical activity of the heart. Then said intern returned to the ER to continue with the rest of the patients. 2 hours later I get another call from the nurses on the ward saying "Doc, are you ever going to come back and disconnect that monitor? It's just the other patients are getting really spooked by staring at that flat line...."
    • by malkavian ( 9512 )

      Hehe.. Task overload is a pain! Is is standard protocol that only a doc can remove the monitors (I honestly don't know, and it's an interesting bit of info)?

      • by Dunbal ( 464142 ) *
        Nah, the nurses are lazy here (Costa Rica) especially in government run hospitals and especially at 4am. Why should a nurse do it when she can just call the intern...
    • Re:Sometimes... (Score:4, Interesting)

      by aaarrrgggh ( 9205 ) on Saturday December 07, 2019 @01:57PM (#59495458)

      It is kind of odd that this problem has not been addressed properly; for datacenters this issue was identified, largely addressed, and eliminated as an issue. The automation is not that complicated, it just requires proper integration of devices.

      Maybe this is why the medical industry is such a mess...

      • by HiThere ( 15173 )

        No, it's not surprising. Different machines are made by different companies, and they all think their machine is exceptionally important. And they have failure modes that need to be addressed. (E.g., the IV drip has run out. This may be critical, or may just be "see to this soon", or even "get to this when you can" depending on what's in the drip. The drip machine doesn't know what's being dripped through it [well, it's got a code, but it doesn't really know what that code means], and it doesn't want t

        • Comment removed based on user account deletion
          • Seems the medical industry should have a similar department that centrally monitors and alerts the doctors and nurses directly based on urgency and importance.

            Did your MSP have to get FDA approval for your management scripts? Update the approval each time you changed them?

            That's the environment medical companies operate in, including the hospitals. Seemingly trivial adjustments need to be documented extensively, and perhaps recertified.

            In some cases, the actually federal requirements aren't as strict as people think, but... all it takes is a bureaucrat with a burr under their saddle to make your life miserable trying to prove that regulation X really doesn't appl

    • It's just the other patients are getting really spooked by staring at that flat line

      You missed a great chance to pull a funny gag.

      Put a hospital bill in the deceased's hands . . . and tell the other patients that he flat-lined after reading the bill.

      You wouldn't have been overworked anymore, as all the other patients would have skedaddled out of there.

      • by Dunbal ( 464142 ) *
        Yeah, only the country I trained in has socialized health care. No bills. But there's always the one for family members: "He said he wanted to tell you something really important" "What did he want to tell me?" "I don't know, he died".
  • I had to shout really loud to get help when my beeper was ignored for over half an hour once.
  • They'll need a ticketing system to help their 24/7 SOC full of analysts (nurses) to decide which events are medium or full escalation and depend immediate interaction with the client.

    Soon they'll have a Machine Learning system in the cloud for event correlation, enrichment and automatic analysis.

  • I spent a week and a half at the intensive care ward and in the end I learned to pause the alarms when I disconnected myself to visit the restroom(after having the catheter removed). :D The pump for the drip made a noise when it was done, the water level for the oxygen machine made a beep, but I could smell it before it was all empty, the machine monitoring heart, lung, pulse Vo2 etc had it's noise and I think it was only the last one that was connected to the room where they monitored everyone which were t

    • The connected 'integration' of these remote devices is definitely the huge opportunity.

      In the intensive care unit, the nurse is in the hallway monitoring some computer screens and doing some other administrative work (reports). When the drip monitor goes off, the alarm is sounding in the room with the patient, which is not where the nurse is located. It's just mess. Dell just built a brand-new hospital in Austin in 2015 and could have worked with all the device vendors to integrate this stuff. They didn't
  • I can't help but think this is very analogous to the metric checks and alerts in the monitoring systems in IT enterprises (Prometheus,ELK stack, PRTG etc.).
    Pretty much all medical devices now are either network connected, or connectable, which bring up the potential to monitor them remotely on dashboards, using the same alerting criteria (which would have to be set by clinicians) as is used in those tools.
    This way, you could have a lot less scary stuff blatting at the patient (but available on demand), and

  • It's an old issue that was probably a solution to potential liability issues. In this day of the IoT the need for audio is likely outdated, but things won't change unless the incessant audio becomes a new liability. First thing patient visitors learn is how to disable the audio alarms.
    • by Falos ( 2905315 )

      Long as my machine design is indemnified, what do I care about patients? About a building drowned in noise? I don't even care if my alerts can even be heard, as long as I can legally claim it did one.

      Because it's 2019 and every incident is a fingerpoint circus, "why wasn't there a safety feature" "the school should have done something" "the forklifts need human-detecting cameras". Everyone wants a payout, everyone wants to pass the buck, everyone is holding up their hands because steps 2 through 50 of any i

  • ... refinery in Port Arthur (Texas, not Australia).

    I was an instrument man and the units would have a big red megaphone-like alarm horn for evac if a connected instrument had a cow.

    The operators were not allowed to disable, so they stuffed the alarms with red rags.

    It did bite them in the ass a few times and some people were killed.

    • ... refinery in Port Arthur (Texas, not Australia).

      God your sentence confused the hell out of me given I have a friend from here: https://www.google.nl/maps/pla... [google.nl]

      Your problem was not unique to Texaco, or even refining, or the oil industry .... or the process industry for that matter. The fundamental problem is alarms were regarded as a great means of alerting for an abnormal condition. But in the days of piped 0.2-1 bar to the control room alarms cost actual dollars to implement so people put a lot of thought into what to apply them to. Fast forward to th

  • This is the inevitable result of the medical instrument industry designing products in isolation and adding "alarms" just in case, perhaps from a perception of defensive design (from their perspective) - the audible equivalent of those boiler-plate warning booklets distributed with every consumer product these days.

    But these alarms are dangerous is patients in at least two ways.

    I'll start with the most direct way, which is the one most people are least aware of -- ICU syndrome. When someone stays in an envi

  • "My product is most important product" attitude in product design is what leads to this. the products are not designed into the environment they are to be used with which have other products.

    it's just the same thing as notifications on your smartphone. you've probably already started using your smartphone in muted mode 24/7 due to them, haven't you?

    notifications from most single app would be okay, but having them from 20 sucks. and having them from 30 makes you ignore the notifications tray completely. does

    • "My product is most important product" attitude in product design is what leads to this.

      No it's not. What leads to this is the realisation that alarms are essentially free along with just a general marketing push for every product to be capable of generating an alarm (if the customer desires) and the fact that hospitals have only just gotten to the point where they are annoying.

      Many other fields have been through this. Alarms become easy, customers just implement them everywhere, and eventually the industry

    • by AHuxley ( 892839 )
      Re "not designed into the environment they are to be used with which have other products."
      The service that costs a gov/charity/private company less..
      Falls off in a min after one test due to not sticking... Good enough for one test and it fails.
      The other product that stays on the human for days due to good design and actually been tested.. but cost to gov... the owners... the charity ...
      So buy the junk product in bulk and set the one test done quickly as a new standard ... work around the junk hardware.
  • I was hospitalized briefly after a bicycle crash. The cacophony of noise was unbearable. The worst was a loud shrill alarm that went off every 5-10 minutes and sounded until silenced. I asked about it and was told that it was an alarm from a CPAP that wasn't adjusted properly. Why don't you changed the settings, I asked. Because only a licensed something-something-something can change it and they aren't due on until sometime later. I couldn't imagine being the patient in that room. After a couple of hours o

    • by jimbo ( 1370 )

      That's the thing; with all that noise it's impossible to rest, adding to the personal stress of the already exhausted patient.

  • This reminds of the story of Leilani Schweitzer and her baby Gabriel. This story was covered in the Radiolab episode "Apologetical", which I recommend you listen to. The basics are that Leilani had to take her son to the hospital late at night, and they put Gabriel under observation, hooked up to several monitors. The monitors would sound off warnings frequently that did not let either of them sleep, so a nurse volunteered to silence the alarms. Unfortunately, she mistakenly turned off all warnings (inc

  • by Applehu Akbar ( 2968043 ) on Saturday December 07, 2019 @02:00PM (#59495468)

    My wife had two six-day stretches of hospitalization during November. Most of this time was in a regular room, with about one-third of the time in the ICU. Both locations were crammed with electronic gear, some element of which found a reason to beep loudly every few minutes. There was even a loud alarm built into the bedside chair that sounded off whenever I had been sitting in it and got up, the rationale being that at times when the patient sits in that chair, staff needs to know if they might have fallen off it.

    If a loud repetitive beep is supposed to summon help, why doesn't it bring an immediate response? And if the problem isn't immediate, why does the device need a loud warning? Why can't error conditions be quietly reported to the charge nurse desk, where screens are being continuously monitored?

    My impression of the slow response to machine alarms was that because the same cacophany is coming from every other room in the pod, nurses have to prioritize their responses.

    • by k6mfw ( 1182893 )
      Yes, I have seen this when visiting a friend in the hospital and one of the machines alarms but nobody responds. If nobody will respond to an alarm then don't need an alarm at all. So if alarms are not needed, then how does staff know when to respond when something real goes wrong?
      • by ufgrat ( 6245202 )
        I asked a nurse this question... her response was that they frequently have no idea where the alarms are even coming from, because they're so pervasive.
        • It's getting to be like when some cheap-ass sensor in your house chirps every few minutes because of a failing battery, and it's impossible to tell where the chirp is coming from. I want a Constitutional amendment requiring all such sensors to be connectable to a smartphone app, which relays a useful message like BEDROOM SMOKE DETECTRr NEEDS BATTERY.

    • by geek ( 5680 )

      Its worse than that. A big part of recovery is rest. My mom had a knee replacement surgery and didn't sleep at all the following days because of all the machines making noises, not just in her room but all the other ones. It sounded like a freaking fair ground with all the crazy going on. It's amazing to me this isn't being addressed. My mom slept several days solidly once at home, the experience was horrifying to her.

    • Alarms should be reported to a central station with a person assigned to monitor them continuously and make sure important ones are not ignored. This would require all devices to be networked though, which raises costs and complexity. Many devices are already networked (pumps of many kinds) though. Liability may be one reason the alarms have not been centralized. If a patient is harmed those responsible can say they didn't hear the alarm amongst the many others. If there is a central station that logs
    • I read an interesting article about medical startups here. I think this MedTech is now a relevant and useful idea. bookimed.com [trendingtopics.at]
  • Yes, the alarm fatigue problem is very real
    https://en.wikipedia.org/wiki/... [wikipedia.org]

    It has been well known for years.

  • After surgery I was in the IC ward and had been mouth tubed because I had stopped breathing under the anesthetic. At one point, while no one was around of course, some mucus in my throat formed a seal and I couldn't breath. Groggy and panicked I tried to find any panic button and even started banging on the panel to try to attract attention. Finally I just yanked the tube out. It cost me an extra week in the hospital due to pneumonia. Talking to my father later I found out that alarms had been blaring but n
  • was a health IoT network specification to get the devices on their own subnets and connected comms systems (a shared API).
    But then I have also wondered why the guy at the nuke plant control panel is on the public internet playing games @ MalwareGamesRUs.com.

    What could possibly happen.

    Just my 2 cents ;)
    • by ufgrat ( 6245202 )
      Well, every manufacturer believes their standard is better, and that everyone else should adopt their standard, so that's one problem. In fact one of the biggest businesses in modern health care is HL7 gateways between various systems, usually requiring a fair amount of custom coding.

      I work at a major teaching hospital, and medical devices do *not* have their own vlans, even though a number of us have been suggesting it for years.
      • by kevinbr ( 689680 )

        https://www.nccoe.nist.gov/sit... [nist.gov]

        Shove this document into the inbox of the right decision makers ....... :-)

        "Securing Wireless
        Infusion Pumps
        in Healthcare Delivery Organizations
        Includes Executive Summary (A); Approach, Architecture, and Security Characteristics (B), and How-To Guides"

    • by kevinbr ( 689680 )

      Most infusion pump manufacturers have all released integration gateways to take pump messages upstream and push things like drug libraries downstream, and also push infusion events upstream for hospital systems and billing records etc.

      AT the point of this integration gateways, it is not deeply difficult to push alarm messages to a central console ( i.e. at a nurses station) [this company I worked with had developed a Nurse console application to display ward level infusion pump data and alarms]

      HL7 is probab

  • Don't you ever forget those before entering a hospital.
    You might not survive without because you'll never sleep otherwise.

    • by rapjr ( 732628 )
      A small rechargeable quiet electric fan is useful too, so you can regulate your own temperature. The rooms are usually way too hot.
  • But please don't remove the machine that goes "PING!"

  • Sadly, hospitals know nothing about the application of alarm management.
    https://en.wikipedia.org/wiki/... [wikipedia.org]

  • Then you have to get the fire alarm system beeps, security system beeps, various manufacturers of medical gear to all be able to interface their alarms together, from an administration staff trying to cut costs as much as possible.

  • The last time I was in the hospital, the bedside monitor would sound red alert every time my O2 sat would briefly dip as I fell asleep, making sure that I did not sleep at all that night. The nurse said there was no way to silence the alarm. And then they wonder why hospital patients who stay more than a day or two start hallucinating and become delirious.

    It's not just hospital equipment. Coming home with a Christmas tree the other day, we had the trunk lid tied down since the tree was a little too big to f

  • That's JUST what we need! A machine to alert us whenever the room decibels raise above some specific level. Maybe it could brightly and quickly flash for hours to inform us. (And if that affects you, well, you're already IN a hospital...)
  • Better staff. Test and peer review staff so they connect humans in the correct way.
    Buy better hardware. If it keep failing as hardware, try a much better product/service...
    Make sure the products that connect to the human work, are usable and keep working...before buying into years of junk products that fail every day.

    Dont allow staff who cant connect humans to hardware near the humans...

    When the hardware fails do a "review"... was it a staff problem, the human removing hardware, failed junk hardware
  • When I was in the hospital after a surgery I couldn't sleep because so many machines were beeping throughout the ward that it sounded like some crappy 8-bit symphony, and that's not an exaggeration.
  • Hospital policies re. equipment alarms will skew to the most conservative (i.e., loud) settings owing to the fact that any litigator worth a damn would always suggest that a given alarm was muted per policy and consequently resulted in a patient's death/injury.
    • by rapjr ( 732628 )
      True, liability is a fear. Using sound as an alert is probably the cheapest way to raise an alert though so cost is probably a factor, and why is the alarm sounded next to the patient where it does the least good? The patient is not the one who needs to know an error condition has occurred and it tends to wake and frighten them. Networking alarms to a central station would cost more and make the system more prone to failure because we don't yet know how to make truly reliable systems (entropy is a mean a
  • I have a feeling all of that is due to an ambulance chaser case. Well if there was a simple alarm on this our dearly departed Alice would still be here. Now Bob doesn't have a wife any more. We're asking for 2 million for this oversight.

    Stupid juries award it.

Waste not, get your budget cut next year.

Working...