Improving alarm management efficacy through predictive modeling and trending

Alarm fatigue, the progressive desensitization of clinical staff to audible alarms in their environment, has been reestablished as a National Patient Safety Goal by The Joint Commission as of January 2014. In order to manage the number of alarms experienced by hospital staff, facilities are charged with finding a way to measure their existing alarm load and then develop methods and policies by which to reduce or mitigate the threat of alarm fatigue. This study used access to a large archive of patient monitoring alarms in order to develop a process by which patient monitoring alarms could be analyzed with high specificity across multiple units of differing specialties and acuity levels. Trends in the distributions of patient monitoring alarm data were identified, with the greatest potential contributors to alarm fatigue being confirmed as the medium and low priority alarms, comprising 70+% of patient monitoring alarms, and condition-based distributions showing strong correlation (with most R² values exceeding 0.990) to average distributions across patient care units of differing specialties and acuity levels. A trend of spikes in heart rate alarms was also observed, suggesting that heart rate limits should be considered differently than those for other vital signs.