Reducing Alarm Fatigue: A Modest Plan

Alarm fatigue and its potential dangers have gained prominence in the last few years thanks largely to the dogged work of front-line healthcare professionals and some solid reporting in the media, particularly the Boston Globe.     

This increased awareness has brought with it the acknowledgement that while alarm fatigue may be simple to understand, it also raises challenging questions. For example, if there are too many alarms, which beeps and bells are prioritized? Which can be ignored? Which are redundant? And most worrisome: What if a clinician ignores the wrong alarm?  And who gets the blame?

Maria Cvach, RN, the assistant director of nursing clinical standards at The Johns Hopkins Hospital, has led the acclaimed Baltimore hospital's award-winning alarm improvement efforts since 2006. She says alarm fatigue has grown because medical devices and monitors found at hospital bedsides are not interoperable.

"Each device functions as its own little alerting system," Cvach tells me. "We work in a world where equipment is isolated, and as a nurse you have to rely on getting rid of the nuisance alarms and trying to extrapolate out of that the important alarms. It is hard to do when you have thousands of alarms a day from every device in a patient's room. In an ICU you may have seven or eight pieces of equipment that alarm. How does a nurse figure that out?"

Cvach believes most nurses do an admirable job sorting out alarms by relying on critical judgments that are based on intuition, experience, and knowledge of the particular patient.

"As a clinician, it is really hard for us to sort out all the noise that is in a room and dig out the important information," she says. "The average nurse doesn't just rely on alarms. They look at the perceived urgency of the alarm and they try to decide, based on the patient, whether or not it is important. That is why you need the critical thinking piece."

But relying on critical thinking and experience in lieu of exact science and best practices also unfairly exposes nurses and other clinicians if things go awry.

"Whenever I hear ‘someone missed an alarm,' people fail to realize that there are thousands of alarms," Cvach says. "It's like Aesop's fable about crying wolf. Where do we go to prevent any kind of events happening? Because they are going to continue to happen until we have a way to predict negative outcomes as opposed to waiting for an alarm to occur."

Cvach and The Johns Hopkins Hospital have been reviewing alarm fatigue since 2006. In that time, they've reduced audio alarms dramatically. "On the unit that I staff we started at over 500 alarms per patient per day, but we were able to reduce the number to 100 alarms per patient per day. We were able to take it down to a fifth of the amount, but that is still a lot."

It's an impressive achievement considering that The Johns Hopkins Hospital team didn't have a map when it started the journey.

"One big thing we didn't understand when we started was we didn't know how big the problem was. We didn't have any data. We used a fault tree analysis approach to try to figure out why alarms get missed," Cvach says.

"We analyzed the problem and started an intervention. We got rid of the duplication. We didn't turn things off. We turned some things to message or visible alarm instead of audible alarm. It's a visual versus an audible representation. We went through every unit and asked the leadership on that unit what do you want this to be and when is this actionable and we worked out default parameters."

The hospital also installed software that triages alarms to nurses' pagers. "We don't send everything to the nurse. They don't want 100 alarms per patient per day, so we came up with these delays," Cvach says.

"We looked at the crisis alarms and the non-crisis alarms. For the crisis alarms, which are very infrequent, we sent those to the nurse through this software. For the non-crisis alarms we set up a delay that allows the alarm to correct itself before it sends a signal to the nurse."

"If it corrects within the timeframe, Cvach explains, "we have established it doesn't send a signal to the nurse. But if it doesn't correct the nurse gets an alarm and acts on it. This really decreases the number of alarms the nurse gets. She is only getting things that are persistent as opposed to those that go away on their own."

Even though it's quieter in the wards of The Johns Hopkins Hospital these days, Cvach says staff still can't measure the effect that reducing alarms has on quality of care.

"We know that we have decreased alarms. What we don't know and what we need to determine is if we done anything to patient outcomes," she says.

"We have the number of response teams called, the number of codes that are called on the units. We can do a before and after. What we don't have is hardcore data on outcomes that are multi-site that also look at specific changes we made. That is a study that is going to happen but it takes many thousands of dollars for these randomized control trials."

Without that data, it may prove difficult for many community hospitals to take up alarm fatigue.

Not so, says Cvach. "Actually it is easier in community hospitals because they can make decisions quicker than you can in a complex place like this," she says.

For instance, reducing alarms doesn't necessarily require expensive software. "Even just changing electrodes daily, we saw a significant drop in alarms and we got good data on the monitors. If you get good data in you are going to get good data out," Cvach says.

Cvach offers a few quick steps to get things rolling.

1. "The first thing is you need to get an alarm committee, an interdisciplinary collaborative group that wants to solve the problem for the hospital," she says.
2. "No. 2 is you have to figure out a way to measure what is happening. What is your baseline?"
3. "The third thing is don't look at solving the problem as a whole hospital. Look at solving it as population-specific," she says. "What happens in an ICU is going to be different from what happens in a telemetry unit and that is different from a pediatric unit."
4. "No. 4 is to establish an alarm protocol that works for your institution not just on paper, but actually in practice.
5. The fifth thing is to try to establish default parameters and train the nurses so they know how to go in and customize alarms so they are actionable alarms for patients," she says.

In the short term it looks like alarm fatigue will have to be taken up by hospitals and health systems on an individual basis.

Broader efforts in the United States to solve alarm fatigue will require coordinated input from patients, device makers, clinicians, hospital administrators, professional associations, and the government.

"Part of that is to get a common language and standards between equipment. Your monitors may not speak the same language or work on the same systems," Cvach says. "The problem is we have not achieved interoperability and we are a long way from that. Once we can achieve interoperatbilty with our medical devices we are going to take a closer step to solving this problem."