Alarm Fatigue: Experts share solutions that prevent sentinel events
by Don Sadler
Everyone is familiar with the fable of “the boy who cried wolf.” A shepherd boy shouted to the villagers twice that a wolf was attacking their sheep just to amuse himself. When a wolf really did attack, nobody from the town responded, and the sheep were scattered.
These are alarms that sound from the many different monitors present in the OR that do not drive an actionable response by OR personnel. The result is what is now commonly referred to as “alarm fatigue.”
The ‘Cry Wolf’ Effect
“There’s no question that clinically non-significant alarms have led to a ‘cry wolf’ effect,” says Dr. Maria Cvach DNP, RN, Assistant Director of Nursing, Clinical Standards, at the Johns Hopkins Hospital in Baltimore, Md. “When an alarm is viewed as insignificant most of the time, staff will eventually begin to ignore the warning. The problem is that staff could potentially miss a clinically actionable alarm.”
Alarm fatigue may occur “when the sheer number of monitor alarms overwhelms clinicians, possibly leading to alarms being disabled, silenced or ignored,” Cvach adds.
The ECRI Institute has listed alarm fatigue as the number one health technology hazard for the past three years.
Every type of medical device and monitor has some kind of alarm: EKG, pulse oximeter, end tidal CO2, cardiac output, respiration, and so forth.
“So the OR can just become a cacophony of alarms,” says Dr. Purna Prasad, BE., M.S., Ph.D., C.C.E., Director, Clinical Technology and Biomedical Engineering, Information Technology/Information Services, Stanford University Medical Center.
“At first, nurses want to respond to every alarm,” Prasad adds. “But as time goes on, they build up a tolerance and become desensitized to them. Their minds are becoming conditioned to unresponsiveness, which is very dangerous in the OR.”
“We have seen hospital staff completely overwhelmed by alarms, unable to distinguish between alarms that are clinically non-significant and those that require action on their part,” adds Rikin Shah, Senior Associate, Applied Solution Group, at the ECRI Institute.
According to Cvach, clinicians are exposed to as many as 700 physiologic monitor alarms per patient day. And research indicates that between 85 percent and 99 percent of all alarms are either false or clinically non-significant, not requiring clinical intervention.
These clinically non-significant alarms disrupt patient care 71 percent of the time, and clinicians disable alarms 78 percent of the time according to a 2011 survey conducted by the Healthcare Technology Foundation.
Cvach gives several examples of clinically non-significant alarms that often sound:
» An alarm that occurs when a threshold setting is breached momentarily. For example, the pulse oximeter drops momentarily when the blood pressure cuff inflates.
» Alarms that sound due to incorrect monitor programming or inappropriate patient individualization. For example, the pulse oximeter alarm is programmed too high for a patient who has a chronically low saturation concentration.
» Alarms that sound due to system or technical issues. For example, the pulse oximeter alarm sounds because the probe cannot read the patient’s pulse correctly.
There are many potentially negative consequences of alarm fatigue, both for patients and OR personnel. In a worst-case scenario, it can result in serious patient complications and even death.
“Non-detection of changes in a patient’s condition can result in a failure to rescue, the results of which can be fatal,” says Prasad.
Meanwhile, many OR nurses are suffering from impaired hearing and concentration, elevated blood pressure and stress levels, disorientation and distraction, and sleep disruption and loss due to the cacophony of alarms sounding in the OR.
Alarm volumes inside hospital ORs can reach 120 decibels, which is as loud as a chain saw. The World Health Organization (WHO) recommends that OR noise levels not exceed 30 to 40 decibels.
Sentinel Event
Clinically non-significant alarms have resulted in sentinel events. The Joint Commission’s Sentinel Event database includes reports of 98 alarm-related events between January 2009 and June 2012, with 80 of these resulting in death, 13 in permanent loss of function, and five in unexpected additional care or extended stay.
In an effort to help deal with the problem of clinically non-significant alarms and alarm fatigue, the Joint Commission has included this in its 2014 Critical Access Hospital National Patient Safety Goals.
In 2014, hospital leaders should be identifying the most important alarm signals to manage, based on such factors as whether specific alarm signals are needed or unnecessarily contribute to alarm
noise and alarm fatigue. By 2016, hospital leaders should establish policies and procedures for managing alarms that address the following:
» Clinically appropriate settings for alarm signals
» When alarm signals can be disabled
» When alarm parameters can be changed
» Who in the organization has the authority to set alarm parameters
» Who in the organization has the authority to change alarm parameters
» Who in the organization has the authority to set alarm parameters to “off”
» Monitoring and responding to alarm signals
» Checking individual alarm signals for accurate settings, proper operation and detectability
By this date, hospital leaders should also educate staff and licensed independent practitioners about the purpose and proper operation of alarm systems for which they are responsible, notes the Joint Commission.
Meeting the Challenge
Prasad sums up the challenge of reducing clinically non-significant alarms and alarm fatigue as follows: “Selecting the right monitoring parameters with the right setting for the right patient, at the right time and by the right clinicians, to obtain actionable clinical data to promote patient safety.”
He breaks solutions into three broad categories:
1. Proper alarm configuration and threshold settings
This includes eliminating redundant alarms, clinically irrelevant alarms, and multiple devices monitoring the same physiological signals, as well as optimizing the alarm to the individual patient’s condition and range of values.
2. Improved alarm algorithms
This involves balancing sensitivity and specificity. Algorithms should be developed that accommodate technical and artifact detection and decrease false alarms and fatigue. For example, research indicates that a 19-second alarm delay would reduce false-positive alarms by 67 percent, according to Prasad.
3. Improved medical device alarm features
These might include non-auditory means for presenting alarms, like color-coded displays or vibrations on smartphones and tablets that are ported to alarms.
In addition, Shah encourages routine replacement of electrodes, proper patient skin prep and electrode interface/placement, and replacing device batteries at a set time every day or every shift.
“Steps like these have been shown to reduce alarm burden and create an environment where alarms that do sound are alarms that require action,” says Shah.
In an effort to raise awareness of clinically non-significant alarms and alarm fatigue, ECRI Institute has created an Alarm Safety Resource website at https://www.ecri.org/forms/pages/Alarm_Safety_Resource.aspx.
The site includes a wide range of free resources to help nurses and clinicians better understand the problem and minimize alarm fatigue. This includes some free sample content from ECRI Institute’s new publication, The Alarm Safety Handbook: Strategies, Tools and Guidance.
“Determining and implementing strategies to improve alarm safety and reduce alarm fatigue will allow alarm management policies and procedures to be improved across the organization,” says Shah. “This will increase the quality of care delivered and improve patient safety across all units of the hospital.”
