Alarm Specificity: We Need it Now!

Alarm fatigue – it’s a well-known issue in nursing.  A recent search of CINAHL Complete yielded 153 results with the key words “alarm fatigue,” and that was with results limited to full-text articles.  Alarms are a critical piece of patient safety (AHC Media, 2018); they alert a nurse to a patient’s changing condition.  However, when nurses become tired of hearing those alarms, problems can quickly result.

What is Alarm Fatigue?

Alarm fatigue results when a healthcare provider delays response to an alarm – or fails to respond at all – due to the ever-present nature of alarms on a unit (Carcerei de Oliveira, Barbosa Machado, Duque dos Santos, & Bicalho de Almeida, 2018).  This delay is not due to provider burnout or maleficence; in cases of alarm fatigue, the provider felt the alarm was probably false and thus did not hurry to respond.  AHC Media (2018) notes that nuisance alarms – those where there is no problem with the patient – remain a common problem and decreasing them is easier said than done.  Alarms are set with specific parameters, and alarm technology has not advanced to the point where, for example, the heart rate monitor can determine that an infant patient is crying and therefore the tachycardia is not clinically significant.  Alarm fatigue research focuses on healthcare providers, but it is not limited to them – as a pediatric home care nurse, I started my shift many times to find a pulse oximeter monitor silenced or a ventilator’s alarm speaker taped over by a caregiver.  In fact, Respironic’s SmartMonitor manual warns caregivers not to block the speaker (Respironics, n.d.).  One respiratory therapist from a Chicago-area company for durable medical equipment (DME) even warns nurses to check for alarms that have been modified for “silencers” when receiving a ventilated child from home.

So what can be done about alarm fatigue?  Clearly, getting rid of alarms is not the answer.  While nuisance alarms are a problem – it’s speculated that around 95% of alarms are not clinically significant (AHC Media, 2018), in critically ill and medically complex patients, every alarm could be a real event.  AHC Media (2018) notes that solutions are slow in coming; even concentrated alarm fatigue efforts have only reduced that to about 50%.  Clearly, nurses have to respond to every alarm, every time…but when there is a feeding pump alarming at the same time as two pulse oximetry alarms and a ventilator, it can be a bit of a choice – where do I go first?

This brings up a second issue, one that corresponds with alarm fatigue.  While nurses are often quite good at identifying exactly which device is alarming because of the sound, how does one know what patient to go to when hearing a single alarm down a hallway of patients?  If a patient is in crisis, every second counts – and the time spent “following the sound” can delay a response even by a nurse not suffering from alarm fatigue.  Furthermore, an informal survey of nurses on one pediatric unit revealed that, sometimes their alarm fatigue led them to thinking of an alarm as a device causing a problem, rather than a patient potentially in crisis.  And, in support of addressing the larger and oft-studied issue of alarm fatigue, I present a potential solution to this second issue: alarm specificity.

Alarm…Specificity?

What started as a joke among nurses on one transitional care unit now may be one piece of the puzzle in solving the problem of blaming an alarm on a device rather than immediately being concerned for a patient.  Wireless alerts do exist that can send an alert to a nurse’s mobile device (AHC Media, 2018), but what of units in which the nurses do not carry work phones?  What about transitional and teaching units where caregivers are being taught proper emergency response? 

“Wouldn’t it be great if this thing just said the patient’s name?” one nurse asked.

Another nurse, caring for a patient whose monitors had a different manufacturer, commented that she appreciated the different tone of those alarms.  “I know it’s my patient right away.”

This technology already exists.  The website ReadSpeaker.com explores countless uses of text-to-speech technology, and digital assistants like Siri prove that a device can certainly be taught to talk.  A simple name would not be enough of an alarm, but coupled with the alarm’s distinctive pitch, HIPAA concerns, of course, could be a problem for announcing a patient’s name.  As could the technological challenge of reprogramming a device each time a new patient is using it.  The solution to that is also simple: a code word or even a room number could replace a name. 

Imagine that you are a nurse on a busy unit, with two ventilated patients in adjacent rooms.  The ventilator alarms with a piercing shriek, as usual, interrupted by – Alert, 104, Alert, 104 – and now you know that you should respond to room 104 rather than 102.  It’s a few seconds and steps saved, but one never knows when a patient may desperately need those seconds.

This Sounds Expensive

The biggest question, of course, then becomes the cost.  Text-to-speech software varies in cost, from hundreds to thousands of dollars (Fearn & Turner, 2019).  Other measures have focuses on cutting costs by limiting the duration of time a patient remains on monitors with alarms – as clinically appropriate (Chen et al., 2017).  This would of course, continue, along with the importance of maintaining patient-specific parameters rather than generally accepted alert levels (Williams, 2018).  Angel et al. (2018) noted the high costs of resuscitation of patients in sepsis – imagine not only the financial cost of a resuscitation, but potential litigation if an alarm was missed due to fatigue or even searching for the source. 

Maybe We Can Make Alarms Sound Nicer Instead?

Alarms, unfortunately, cannot be “gentle.”  They might be ignored entirely if that were the case!  Nurstoons artist Carl Elbing (2000) humorously speculated what that might be like…

A patient I once cared for had a bed alarm that played a merry electronic version of Yankee Doodle, to be “patient-friendly.”  It soon had to be replaced with a standard, screeching bed alarm, because the musical alarm quickly resulted in the patient getting up to hear the music!

Humor aside, there are very real implications to be considered in making alarms less “annoying.”  Patient specificity and nurses silencing alarms in advance of events that would set them off, like suctioning (Williams, 2018) can help to reduce the number of alarms.  However, Williams (2018) notes that reducing nuisance alarms is only one part of the alarm fatigue issue.  Caregivers cannot be educated that it is vital to respond to all alarms and then see nurses failing to do the same.  For the home care environment, where there is only one patient, user-specific alerts are admittedly not critical; the technology could be adapted to allow the option of a “standard” alarm or a specific one when the device parameters are set.  However, on a teaching or transitional unit, this technology could also aid caregivers – should they respond to that alarm or can they carry on with what they were doing, knowing it is not their loved one?

Do We Have to Replace All Our Equipment?

No!  This is not a scorched-earth solution.

For the good of our patients, we need this adaptation to be made.  When the American Heart Association mandated that CPR mannequins have feedback devices to monitor compression performance, there was an immediate response to make new mannequins with the technology included (Channing-Bete, 2019).  However, existing mannequins were not obsolete – these same companies developed after-market adapters to make the mannequins compliant.  The same could be done with devices already on the market if alarm specificity is demanded.  New devices could have the feature built in, with adaptations made to existing equipment.  Once we have it, we can study it and provide evidence that this will, as predicted, lead to quicker response times and reduced alarm fatigue.  But the first step is getting it out there.  Join me in the quest for more specific alarms!