Why We Ignore Notifications: What Hospital Alarm Research Shows

Be honest: how many notifications have you cleared today without reading? For most people the answer is "most of them", and the usual response is a vague sense of personal failure.

It is not a personal failure. The best evidence for that comes from a setting where ignoring an alert can kill someone — and where trained professionals do it anyway.

What happens when experts get too many alerts?

They stop responding, and people die. In 2013 the Joint Commission, which accredits American hospitals, issued a Sentinel Event Alert on medical device alarms. Its database held "reports of 98 alarm-related events between January 2009 and June 2012. Of the 98 reported events, 80 resulted in death, 13 in permanent loss of function, and five in unexpected additional care or extended stay."

The cause was not incompetence. It was volume. The alert describes "the number of alarm signals per patient per day" reaching "several hundred depending on the unit within the hospital, translating to thousands of alarm signals on every unit and tens of thousands of alarm signals throughout the hospital every day."

And crucially, most of them mean nothing: "it is estimated that between 85 and 99 percent of alarm signals do not require clinical intervention."

The Joint Commission's description of what follows is worth reading slowly, because it is exactly what your phone has done to you: "clinicians become desensitized or immune to the sounds, and are overwhelmed by information – in short, they suffer from 'alarm fatigue.'"

Does that really apply to a phone?

The mechanism is the same, and the stakes being lower make it more likely, not less. Intensive care nurses, professionally trained, responsible for people's lives, stop responding to alerts when the overwhelming majority of them turn out to be noise.

You are not more disciplined than an ICU nurse. Nobody is. If a 90%-noise alert stream defeats them at work, a 99%-noise alert stream is going to defeat you on a Tuesday afternoon — and with no consequence for missing a marketing push, there is nothing to train the habit back.

This is habituation, and it is automatic rather than chosen. The brain demotes signals that are frequent, repetitive and usually irrelevant, which is the same machinery that stops you hearing the fridge. You do not decide to ignore notifications. Your attention system files them below the threshold at which you would get a say.

What did the professionals actually do about it?

Something worth knowing, because it is the failure mode the alert warns about. Faced with constant noise, the Joint Commission reports that "clinicians may turn down the volume of the alarm, turn it off, or adjust the alarm settings outside the limits that are safe and appropriate for the patient – all of which can have serious, often fatal, consequences."

That is the endgame of alert overload, and it is what people do with phones too: notifications off for the noisy app, then off for most apps, then Do Not Disturb permanently, and eventually the one genuinely important thing is silenced along with everything else.

The lesson the hospitals drew was not "alarm harder". It was to cut the number of alarms so that the remaining ones meant something.

What does break through?

Signals that are rare, that are different in kind from the background, and that require an action rather than a dismissal. A phone call qualifies on all three — for now — which is why it still works where a banner does not.

But notice that "rare" is doing as much work as "phone call". The reason a ringing phone gets answered is that almost nothing rings any more. It is a high-signal channel because it is an empty one, and that property is not permanent. It is a resource you can spend.

Three practical consequences:

• Match the channel to the cost of missing it. If nothing bad happens when it is missed, it does not warrant an interruption that cannot be ignored.
• Fewer, not louder. The hospital answer was to reduce alarm volume in the sense of count, not decibels. The same applies to a phone.
• Prune as habits form. A reminder that has done its job is a reminder that should stop, before it becomes background.

The part that applies to us as well

Everything above is an argument against overusing our own product, and it would be dishonest to leave that out.

A reminder call cuts through because calls are rare. Set one for every small task and you will do to your phone exactly what the hospitals did to their wards: a stream of alerts that are mostly non-actionable, answered less each week, until the call about your mother's medication arrives in the same mental category as the rest.

So the useful advice is not "set more reminders". It is to reserve the interrupting channel for the things that genuinely hurt if missed, and to let everything else be a list you check when you choose to. A product that gets this wrong stops working, and it stops working quietly — you will not notice the calls you started ignoring.

If you take one thing from the research: 85 to 99 percent noise is enough to defeat trained professionals with lives at stake. Whatever your own alert stream looks like, it is probably worse than that, and it is beating you for the same reason.

Source: The Joint Commission, Sentinel Event Alert, Issue 50: Medical device alarm safety in hospitals (April 2013) — event counts, alarm volumes, the 85–99 percent estimate, and the description of alarm fatigue and clinician workarounds.