Abstract
This perspective paper describes the building elements for realizing a precise patient monitoring algorithm to fundamentally address the alarm fatigue problem. Alarm fatigue is well recognized but no solution has been widely successful. Physiologic patient monitors are responsible for the lion's share of alarms at the bedside, most of which are either false or non-actionable. Algorithms on patient monitors lack precision because they fail to leverage multivariate relationship among variables monitored, to integrate rich patient clinical information from electronic health record system, and to utilize temporal patterns in data streams. Therefore, a solution to patient monitor alarm fatigue is to open the black-box of patient monitors to integrate physiologic data with clinical data from EHR under a four-element algorithm strategy to be described in this paper. This strategy will be presented in this paper in the context of its current status as described in our prior publications.