Introduction: The burden of bedside alarms is a well-established patient safety hazard with >85% of alarms in ICU settings considered non-actionable or “false alarms”. Healthcare providers risk developing alarm fatigue with potential for prolonged response times and silencing of clinically meaningful alarms. Additionally, alarms disrupt care and cause stress for patients and families. We hypothesize that creating conditional alarm triggers for bedside monitors can decrease the frequency of non-actionable alarms without compromising patient safety. The decreased burden of alarms on nurses, patients, and families may promote increased satisfaction and wellness.

Methods: Single center, quality improvement initiative in a 26-bed cardiac acute care and 36-bed pediatric ICU, modeled after successful iterative effort at another tertiary children’s hospital. After 1-month pre-intervention observation of bedside alarm burden (August 2019), monitors were programmed with conditional pulse oximetry (SpO2), respiratory rate (RR), and premature ventricular contractions (PVC) alarm triggers. Hierarchical time delays were established for saturation alarms (>90, 80-89, 70-79, and <70%) with different parameters for single ventricle versus septated patients. Alarms were tallied for 1-month follow-up (October 2020). Primary outcome was alarms per monitored patient day. Nurses and families were surveyed pre- and post-intervention.

Results: A total of 2034 and 1968 monitored patient days (MPD) were evaluated in the pre- and post-intervention periods, respectively. The median number of cardiorespiratory monitor alarms (PVC, SpO2, RR) per MPD decreased from 5 to 0 (p < 0.001) in acute care and 6 to 0 (p < 0.001) in the ICU. When evaluating patient days with at least 1 alarm, median alarms decreased by 73% (15 to 4, p < 0.001) in both acute care and ICU. Alarm burden in RR, SpO2, and PVC categories also decreased. No increase in frequency of rapid response or code events occurred in either unit. Nursing surveys (n=253) reported positive trends of more manageable alarm burden and decreased perceived number of alarms. Family surveys (n=160), however, reported decreased sleep quality and increased perceived alarms. 

Conclusions: Implemented changes to bedside monitor alarm conditions decreased total alarm frequency in both cardiac acute care and pediatric ICU without compromising safety.