Alarm fatigue: the human-system interface.

The Joint Commission’s 2014 National Patient Safety Goal .06.01.01 to reduce harm associated with clinical alarm systems highlights the need to manage the number of alarm signals and their noise (bells, dings, buzzers, etc). However, in the clinical area, staff rely on audible and visual alerts to prompt them to immediate interpretation, clinical assessment, and sometimes lifesaving action. To this effect, the advancement of technology has provided medical device manufacturers with the ability to provide a great number of physiological data point sensors at a low cost. This has resulted in an increased number of alarms being delivered to the care provider who is then expected to quickly synthesize and interpret a growing volume of data points into useful clinical information. One of the many problems is that frequent false or false-positive results give rise to alarm fatigue, whereby the healthcare providers becomedesensitized to the noise associatedwith the alarm. An important factor depicted in the work of Bliss et al is if an alarm is accurate 90% of the time, individuals will respond to the alarm 90% of the time. Conversely, if the alarm is accurate 10% of the time, individuals will respond only 10% of the time. This demonstrates that workers will modify their response rate to the alarm system’s reliability. Subsequently, it becomes a patient safety mandate for clinical alarms to be valid not only to improve the accurate response rate, but also to reduce alarm fatigue. Alarm fatigue is a human response tomachine interface. When alarms are heard incessantly, the individual can be overwhelmed. TheWorld Health Organization’s Guidelines for Community Noise defines noise as unwanted sound. It further describes the capacity of noise to induce annoyance. For example, the sound pressure level (dBA) of the indoor environment, when it equals 35 dBA, can account for moderate annoyance. In a 2005 Johns Hopkins study, the daytime hospital noise level for the year 1960 was determined to be 57 dBA and in 2005 it was 72 dBA. Konkani andOakley in their literature reviewequated sound sources that may occur on a daily basis with sound pressure levels. A noisy vacuum cleaner at a 10-m distance is equivalent to 50 dBA and a noisy lawnmower at a 10-mdistance is equivalent to 60 dBA. In addition to annoyance, the human response to sound at these levels can become one of fatigue and/or desensitization to the alarm(s) that are asking the healthcare provider to promptly intervene. When this occurs, staff may miss, ignore, or disable alarms. The Pennsylvania Patient Safety Authority (in 2011), an independent state agency, reported that over a 6-year period, 31 of 35 reported deaths related to physiologic monitoring were attributed to human error. Of the 35 events, 28 involved telemetry monitors. Human error involved in the 31 events included equipment not connected to either the patient or the equipment (n = 14), transport outside the unit for diagnostic testing (n = 6), inadequate response to alarms (n=6), andalarmsbeing silenced (n=4). To counteract human error, the Food and Drug Administration strongly advocates the application of human factors engineering into the safe design of alarm systems. By incorporating human factors engineering into the design, the user interface may Author Affiliations: Regional Clinical Practice Consultant (Ms GuardiaLaBar), Kaiser Permanente Northern California Patient Safety, Oakland; Clinical Practice Consultant (Dr Scruth), Critical Care/Sepsis, Clinical Effectiveness Team, Regional Quality and Regulatory Services, Oakland, California; Professor of Applied Psychology (Dr Edworthy), Cognition Institute, Plymouth University, Devon, England; Director of Adult Services and Caring Science Integration (Ms Foss-Durant), Kaiser Permanente Northern California, Patient Care Services, Oakland; and Director of Biomedical Technology Integration (Mr Burgoon), Kaiser Permanente, Enterprise Application Services and Emerging Technologies, Oakland, California. The authors report no conflicts of interest. Correspondence: Elizabeth Ann Scruth, PhD,MPH, CCNS, CCRN, FCCM, Quality and Regulatory Services, Kaiser Permanente, 1950 Franklin St, Oakland, CA 94612 (elizabeth.scruth@kp.org). DOI: 10.1097/NUR.0000000000000039