Alarm Fatigue

Alarm fatigue describes a problem in UX design where a small number of alerts can be effective, but a large number begin to be counter-productive due to limits of human attention and perception. One wonders if there are similar patterns in educational feedback. Original article follows.


Hospitals are overwhelmed with alarms, and it's a problem with severe consequences: "Current clinical alarm technology is generally based on Data Threshold Science, which detects when a specific data threshold has been crossed and activates an alerting mechanism (usually an audible alarm). Unfortunately, this creates a concept known as “alarm fatigue,” which has truly frightening consequences when encountered in complex clinical environments. In such environments, critical alarms are often either ignored or even turned off." - html

The results, predictably, are injury and death.

Some causes of alarm fatigue include

False Alarms, which "cry wolf" and add noise to the system.

Notification Wars where every device must compete with the others for competition.

The sheer number of alarm conditions. In an ICU the average number of alarm conditions per day for a *single bed* was 771.

In chemotherapy infusion centres, each patient is attached to a device that gives off an alarm if the infusion either stops unexpectedly or finishes as planned. The nurse to patient ratio means that nurses managing several patients on different timers need to be able to hear one machine above another when all the alarms go off at once: aural multitasking.

Something similar happens with penguins and their chicks.

Jeff Rimland: See sonification, in which data is converted to an "auditory display" more holistically and with greater understanding of context and human perceptual factors.