<img src=”http://farm1.static.flickr.com/39/76765412_618a458105.jpg” width=410 alt=”heart monitor notification”>
When designing an interface, I always strive to make my notifications accurate, timely, relevant and helpful. Never is this is critical as when someone’s life is at stake.
NPR did a piece [original recording] on a Boston Globe study that indicated that in a single day at a small 15-bed unit at Johns Hopkins unit, there were 942 alarms. Approximately 80% of those were false alarms. Understandably, when you’re hearing an alarm every 90 seconds, your brain eventually tunes them out. You can predict where this story is going… what happens when the real alarm sounds and no one recognizes it.
This issues is so serious and difficult that some hospitals have a dedicated staff member whose sole job is to monitor alarms. This is a costly fix to something that should be able to be solved with smarter technology.
The obvious solution is to make the alarms work better. They shouldn’t sound when a patient rolls over and dislodges a sensor. I’m sure this hasn’t been lost on manufacturers or on hospital purchasing. Short of that, what other solutions are out there?