Staff at a burn center ICU worked to reduce alarm fatigue. They developed four modules of best practices and educational material.

  • Many alarms were related to ECG leads becoming loose or disconnecting from the skin.
  • The team consulted physicians and nurses to develop new best practices.
  • Some improvement slipped when staff and leadership changed.

When a team set out to address alarm fatigue at a North Carolina burn center ICU, they found success with implementing new best practices that addressed some of the most common reasons for nuisance alarms. But they also found those wins can slip when staff changes bring new people who were not trained in the updated ways and new leadership that was not there for the initial effort.

The North Carolina Jaycee Burn Center at the University of North Carolina Medical Center reduced nonactionable and false alarms from the baseline of more than 100 alarms per bed per day. The staff developed new skin preparation processes to improve the adherence of ECG leads for burn patients.

The effort was rooted in The Joint Commission’s National Patient Safety Goal (NPSG.06.01.01) on clinical alarm safety in 2014, says Rayna Gorisek, MSN, RN, CCRN, CNL, lead author of the report on the hospital’s success. Gorisek was a clinical nurse IV in the Jaycee Burn Center at the time and now is the clinical nurse leader in the surgical ICU at the Durham VA Medical Center.

The NPSG spurred hospital leaders to look at their own alarm rates. The burn center identified specific issues that needed attention.

“We identified particular things that involved our patients in the burn ICU that were not necessarily common to the rest of the hospital,” Gorisek says. “We wanted to make an improvement, so we started with seeing what exactly was going on in our ICU with alarms, which we hoped would lead us to some education pathways and show us how to make the biggest impact.”

Beginning in January 2016, the Jaycee Burn Center team collected a four-week baseline alarm sample that revealed a mean of 110 alarms per bed per day — lower than the 187 alarms per bed they had seen from other facilities but still high enough to warrant action.

The alarms were differentiated as critical, alert, or inoperative. The critical alarms required an immediate response and might be triggered by life-threatening conditions such as apnea or asystole.

Alarms were triggered when predefined parameters approached thresholds that might trip a critical alarm. Inoperative alarms simply meant the monitor could not detect a signal. A common cause of an inoperative alarm was an ECG coming loose or disconnecting from skin. This was particularly important, Gorisek explains, because loose or detached ECG leads were responsible for 96% of inoperative alarms in the baseline.

“With burn patients, it can be a constant challenge to keep the leads on because the skin is damaged and there are substances that have to be applied to the burns that, unfortunately, can make it difficult for the leads to adhere,” she explains. “We identified that as a problem unique to our type of care that could be contributing to the number of alarms.”

Four Intervention Modules

Using the baseline data, the team created four educational modules for the burn center staff to raise awareness, apply evidence-based tactics for improvement, and to promote interprofessional collaboration, Gorisek says. One of the resources they consulted was a practice alert, “Managing Alarms in Acute Care Across the Life Span: Electrocardiography and Pulse Oximetry” from the American Association of Critical-Care Nurses.

Gorisek, assisted by a critical care nurse, presented one module each month to all burn center ICU nurses, nursing assistants, and respiratory therapists at monthly staff meetings and unit in-service training sessions. The first module explained the baseline data, along with education about alarm fatigue and how it can affect patient safety. The module also explained The Joint Commission’s NPSG on alarm safety.

“We tried to impress on them how much this can affect patient safety. We used an example that was almost like a sentinel event, trying to make it more impactful,” she says. “We wanted them motivated to bring it to their everyday practice, to think that this is something that actually could happen to them and their patients.”

The example involved a patient with a tracheostomy whose pulse oximeter fell off and did not alert the nurses to a lack of oxygen — with a fatal outcome. In the example, alarm fatigue may have contributed to the nurses’ failure to recognize the problem until it was too late.

“I read that to them to start the education portion, to make an impact about how bad this can be and show that it was something that could happen to any of us,” Gorisek says.

In addition, Gorisek and her team posted the NPSG on unit bulletin boards and reviewed it during daily unit safety huddles on the unit.

Placing ECG Leads

The second module addressed how to place ECG leads correctly to reduce inoperative alarms as well as how changing the leads during each patient’s daily bath could improve skin adherence.

As part of this module, the alarm fatigue team searched for best practices on how to place ECG leads over burned tissue in an ICU to reduce instances of the lead coming loose, but they found none in the literature. In response, they asked for advice from burn center nurses and physicians.

That spurred the development of three best practice methods for securing ECG leads on patients in the burn center ICU. The leads for patients with chest burns could be applied directly into silver sulfadiazine cream. For patients with diaphoresis, nurses could use a liquid adhesive on the leads.

The third best practice addressed patients with Stevens-Johnson syndrome/toxic epidermal necrolysis, in which the skin peels off. Staff were instructed to wash chest wounds with soap and water, pat the area dry, and place the leads before applying silver-impregnated dressings.

Those best practices were presented with the same methods as the first module, but with the addition of one-on-one instruction. The third module addressed the correct use of pulse oximetry probes, with nurses taught to change the probe and move it to a different site each shift.

During this process, the alarms were turned off to avoid inoperative alarms. When burns on the fingers and toes hampered the typical use of oximetry probes, nurses were told to use an ear clip pulse if possible.

For each patient on a ventilator, Gorisek and her team worked with respiratory therapists and nurses to establish acceptable end-tidal carbon dioxide monitoring parameters, and importantly, to pause the alarms while performing tracheostomy care or suctioning. Those working with the patient at that time would know the alarms were false, Gorisek explains, but they nonetheless contributed to the cacophony of alarms in the unit.

Customizing Alarms and Settings

In the fourth module, nurses learned how to customize continuous patient monitoring system alarm parameters for each patient as well as how to review monitor settings at each patient handoff. They also learned how proper use of the monitors could improve patient safety.

“Tailoring alarms to be specific to the patient was very important. If we knew someone’s blood pressure was high, or if their baseline rhythm was an [atrial fibrillation] rhythm and we didn’t need to be alerted to that, tailoring the alarm to the individual could make a big difference,” Gorisek says. “We really made a lot of effort to educate nurses about that.”

The alarm fatigue team worked with each nurse individually to ensure they were all comfortable with customizing the alarm settings, then they mounted laminated 10 × 5 cm signs at the central monitoring stations to remind them. The sign included a rhythm strip graphic and asked, “Have you customized your alarms?”

After the initial education period, Gorisek’s team updated the staff at six-month intervals with the number and type of alarms on the unit. They also implemented an annual refresher on the four modules. New graduates coming to the ICU also were educated on the four modules.

The number of alarms decreased more than 50%. But six months after the project began, the alarms increased and were approaching the baseline numbers. Gorisek and her colleagues studied the data and determined significant changes in unit leadership and staffing contributed to the uptick in alarms.

Six months later, staffing had stabilized and education efforts resumed. The number of alarms fell to the those seen immediately after the first intervention period.

“The increase in alarms in January 2017 and the decrease in alarms in June 2017 were likely related to the stressors the nursing staff were experiencing,” the team wrote in their report. “These observations emphasize the importance of understanding the possible impact of unit-level culture and sources of work stress while undertaking improvement initiatives.”

The unit also implemented new default alarm values for the adult ICU unit in January 2017. The team concluded the decreases in alarm numbers in June 2017 and January 2018 probably were partially attributable to the new default alarm values.

Gorisek notes the literature indicates changing the default alarm values can reliably reduce overall alarm numbers.

“We tried to involve any stakeholders who were involved in this process when we were developing the modules and implementing our strategies for sustaining the improvements,” Gorisek says. “You have to keep monitoring and reviewing the data, and weeding the garden as you move forward. Perseverance is key.”


  • Rayna Gorisek, MSN, RN, CCRN, CNL, Clinical Nurse Leader, Surgical ICU, Durham VA Medical Center, Durham, NC. Email: rayna.gorisek@va.gov.