Data-Driven Implementation of Alarm Reduction Interventions in a Cardiovascular Surgical ICU.

BACKGROUND

Alarm fatigue in the ICU setting has been well documented in the literature. The ICU's high-intensity environment requires staff's vigilant attention, and distraction from false and non-actionable alarms pulls staff away from important tasks, creates dissatisfaction, and is a potential patient safety risk if alarms are missed or ignored. This project was intended to improve patient safety by optimizing alarm systems in a cardiovascular surgical intensive care unit (CVSICU). Specific aims were to examine nurses' attitudes toward clinical alarm signals, assess nurses' ability to discriminate audible alarm signals, and implement a bundled set of best practices for monitor alarm reduction without undermining patient safety.

METHODS

CVSICU nurses completed an alarm perception survey and participated in alarm discriminability testing. Nurse survey data and baseline monitor alarm data were used to select targeted alarm reduction interventions, which were progressively phased in. Monitor alarm data and cardiorespiratory event data were trended over one year.

RESULTS

Five of the most frequent CVSICU monitor alarm types-pulse oximetry, heart rate, systolic and diastolic blood pressure, pulse oximetry sensor, and ventricular tachycardia > 2-were targeted. After implementation, there was a 61% reduction in average alarms per monitored bed and a downward trend in cardiorespiratory events.

CONCLUSION

To reduce alarm fatigue it is important to decrease alarm burden through targeted interventions. Methods to reduce non-actionable alarms include adding short delays to allow alarm self-correction, adjusting default alarm threshold limits, providing alarm notification through a secondary device, and teaching staff to optimize alarm settings for individual patients.