On noisy hospitals and ‘alarm fatigue’

Several patient deaths in the U.S. have been attributed to “alarm fatigue,” after overworked medical staff failed to respond to an alarm in time. Acoustics expert , provost of McMaster University and a mechanical engineer by training, Ilene Busch-Vishniac spoke to Maclean’s from Washington, where she was preparing to attend a conference addressing the topic.

Q: What is alarm fatigue?

A: Alarms are meant to alert you, and if they’re constantly going off [the idea is that] you could become somewhat blasé. Many of my peers suspect that medical staff are tuning them out, but I’ve seen no evidence of that in the literature. What is certainly true is that there are so many alarms going off, it is not physically possible for medical staff to respond to all of them as quickly as we would like. Alarms will occasionally go unanswered, but that could simply be because noise in the hospital is so loud you literally couldn’t hear the alarm—or it could be that at the moment the alarm was going off, there were 70 others going off.

Q: How are all these alarms affecting hospital workers and patients?

A: Alarms have been identified as the number one annoyance for patients and staff. There is a suggestion that it’s actually causing damage to patients because it influences their ability to sleep, which influences their ability to heal. One study of nursing staff correlated the presence of [the hormone] cortisol, which is an indicator of stress, with noise on the unit. For patients, I do not know of any studies that have been done, but if you’re lying there in hospital worried about how you’re doing, and an alarm by your bed goes off and you don’t know what it means, you are more or less guaranteed to be startled.

Q: Is alarm fatigue a problem in other work environments, too?

A: There are certainly very noisy work environments and you could probably list lots of them, like printing presses: you can’t hear even when you scream, so those are awful. The question of alarms comes up sometimes in nuclear plants and other places where you have to worry about paying attention where any sort of adverse event could happen, but I think the enormity of this problem is quite specific to hospitals. The only context in which I have heard alarm fatigue mentioned is in the medical context, and it’s a relatively new discussion topic.

Q: Are alarms increasingly being used as a stopgap measure because hospitals lack the staff to monitor patients and follow up in person?

A: I don’t know whether staff have got more patients, but we’ve got more and more equipment for every patient, and more and more alarms going off. It seems to me that a lot of what has driven this, at least in the U.S., is not necessarily medical outcomes, but legal outcomes. God forbid something should happen and you didn’t have an alarm go off: the hospital and equipment manufacturers might be seen as negligent. From a purely acoustic point of view, what we want to say is, take a breath, and let’s talk about when should we have alarms, and when should we not. We no longer make any devices that don’t have alarms, but we haven’t been very smart about those alarms.

Q: And hospitals would be noisy places even without alarms going off.

A: The background noise in hospitals has been increasing year after year after year forever, as far back as we could go. Hospitals have a huge amount of low-frequency noise, almost certainly [caused by] heating, ventilating and air conditioning systems. As years have moved on, we’ve required greater and greater airflow in hospital environments, to filter organisms and pathogens out of the air. You design for a low airflow, and you force the same system to have a much greater airflow, you can produce a system that roars. The other thing is that over the years, the density of people in hospitals has risen dramatically, and all of them are talking. There is a rich oral tradition in hospitals, when you think of the concept of doctors doing rounds in the morning. So you have a huge number of people who are mobile and who are speaking—and who are, by the way, doing that 24-7. Add to that the overhead paging, and you understand why there’s so much speech noise.

Q: Why haven’t we done a better job of controlling all this noise?

A: We’ve known the hospitals are noisy, we’ve known it’s a big complaint of patients, staff and visitors, but we haven’t known what to do about it. We haven’t had much in the way of materials available to us. For cleanliness purposes, there is a strict limitation on traditional sound-absorbing materials at hospitals. All of the walls are hard, the floors are hard, the ceiling is hard. In a typical office building, you might have acoustical tiles. Carpets would prevent you from having the click-click-click of heels, and might dampen some cart noise, but they don’t actually absorb sound energy well. The best sound-absorbing material is those panels [made of] a fibrous material, with holes in them. Panels like that are traditionally used in a dropped ceiling. The problem is that it’s not cleanable, and those little holes are good places for bacteria to hide. So there needs to be new materials made that are effective, but can be cleaned and appropriate for hospitals. There is some progress there.

Q: Would it be expensive to retrofit hospitals and make them more sound absorbent?

A: Hospitals are probably the most expensive buildings you can imagine, and the cost of sound-absorbing materials would be trivial. It’s pennies. While I understand there is a cost there, I also know that you could be clever about how they got put up. We did a demonstration where we put new sound-absorbing panels up with Velcro. And the day we put it up, [hospital staff] turned down the volume of the telephones, they turned down the volume of their paging on the unit, and it immediately got much quieter.

Q: Many people view alarms as a necessary evil. Is it difficult to get people to view alarm fatigue as a problem?

A: We have precious few dollars to spend on doing research and making hospitals better. It’s hard to convince people that it would be good to make them quieter when the funding we would need to study that problem would compete, say, with funding for stem cells that would let us do something extraordinary. It is very hard to get people’s attention to this. It’s a problem that I can’t tell you, “Please don’t go to the hospital because it will kill you because of the noise.” It is at a different level from what we tend to focus on when we look at health care outcomes. On the other hand, it affects virtually everybody who has to go into the hospital. It really would be good to know if we are already at the point where the noise is so high, and people are known to be sleeping so poorly, that we are causing medical declines that we could avoid if we could simply make hospitals quieter.

Q: What needs to be done to address this?

A: The literature tells us that over 90 per cent of alarms result in no actions being taken. That’s different from a false positive. The machines are working just fine, the alarm goes off, you check and say, “Okay, there’s something with this patient but we’re not at the point where we want to do anything.” So, 90 per cent of the time, no action is taken. What that says to me is that if we could reduce the number of alarms, say by 70 or 75 per cent, we might get down to few enough alarms that we could start sending them to just the medical staff—whether it’s the nursing station, or by a device that nurses wear—instead of having them all going off and audible to everybody.

Alarms on medical devices tend to be very good about catching your attention, but we haven’t made sure they’re providing a lot of information. I read a study where they played 100 of the most common alarms in hospitals, and they had medical staff try to identify what they were. The group that did the best, not surprisingly, was nurses, but they just barely got half of them right. Alarms are not distinguished well enough from one another, and don’t tend to tell us whether or not this is an urgent matter.

Q: What else?

A: The other thing we haven’t done is integrate our knowledge from multiple devices. When you have a critically ill patient, it’s likely they have not one, but two or three sensors hooked up to them, and therefore lots of alarms might go off. I don’t know if you’ve been in hospital and had an oxygen sensor that’s like a clothespin on your finger—you would remember because they feel miserable, they squish your finger and hurt, so it’s common for patients to pull them off. Then the oxygen level goes to zero, and an alarm starts sounding. Imagine the patient who’s really in danger not only has that on, but is likely to have other measurement tools. For instance, if a person is on a ventilator, and the ventilator is working and no alarm is sounding there, but the alarm is sounding from an oxygen meter on a finger, then the patient [probably] just took off one of the sensors.

I think there is some progress in dealing with these problems. I am worried about some of the baseline assumptions people are making. Simply the use of the phrase “alarm fatigue” suggests a tendency to blame this on medical staff not responding, and I think that misses the bigger point: to pare down [the number of alarms], prioritize them, reduce noise in hospitals by having done that, and therefore improve the probability that patients will heal.