Making Alarms Count

Culture change and teamwork – not capital investment – allow one medical center to create an alarm management system that keeps patient safer…and nurses happier

What contracting executive hasn’t heard the beeps, chirps and screams of telemetry-related alarms in their facility’s ICU or med/surg units? Designed to keep patients safe, too many alarms, sounding off too often, can actually compromise patient safety. That’s because nurses can, in essence, tune them out. It’s a phenomenon known as “alarm fatigue.” And it’s a problem.

In fact, ECRI Institute listed alarm hazards as one of its top 10 safety hazards for 2014.
“Excessive numbers of alarms – particularly alarms for conditions that aren’t clinically significant or that could be prevented from occurring in the first place – can lead to alarm fatigue, and ultimately patient harm,” according to ECRI. “Beyond alarm fatigue, patients could be put at risk if an alarm does not activate when it should, if the alarm signal is not successfully communicated to staff or does not include sufficient information about the alarm condition, or if the caregiver who receives the alarm signal is unable to respond or is unfamiliar with the proper response protocol.”

89 percent reduction
A couple of years ago, Boston Medical Center tackled the issue head-on, and reduced the average number of alarms from 90,000 a week to just under 10,000 a week for a 24-bed unit — an 89 percent reduction in alarms. They did it without buying new technology, but rather, through teamwork and planning. And today, patients are just as safe as they were before. In some ways, safer.

“The ability for nurses to hear alarms, spend more direct time with patients and have a more pleasant working environment all contribute to safety,” says James Piepenbrink, BSBME, director, department of clinical engineering. Piepenbrink presented Boston Medical Center’s alarm-reduction strategy at the 2013 UHC Annual Conference, along with Deborah Whalen, MSN, APRN, ANP-BC, clinical service manager and nurse practitioner of cardiology; and Patricia Covelle, RN, director, critical care nursing. He shared their story with The Journal of Healthcare Contracting.

“Boston Medical Center has been looking at alarm management consistently since 2008,” explains Piepenbrink. “At that time we were expanding telemetry coverage and investing in new technology, and we found wide variability in device and alarm defaults across the different telemetry care areas.” To address the issue, the medical center convened a multidisciplinary team comprising the chief medical officer, nursing director, nurse managers, clinical instructors, physicians (attendings and residents) and clinical engineering, to standardize alarm defaults across the hospital.

“The prime trigger was a couple of incidents that occurred in teaching hospitals that appeared in the press, and we wanted to be certain that we were not at risk for a similar event,” he says. “This caused us to continue to evaluate individual parameters and identify areas of opportunity where we were ensuring that critical – that is, clinically significant – alarms were identified and captured.”

The team found that “as a result of creating alarms for the multitude of individual parameters, there was an overwhelming amount of noise on the units, and with this, frustration,” explains Piepenbrink. “[W]e enhanced the sensitivity and specificity of alarms, but we didn’t address the impact of these alarms.

Less critical alarms
“In 2011, the team started looking at ways to create a more balanced alarm management program by evaluating alarm data from the telemetry system so that we could identify opportunities to better manage alarms and the accompanying noise. In looking at the default settings as well as the alarm data, we found that there were a large percentage of less critical alarms that were ringing off and overwhelming the staff.”

The team focused on audible cardiac monitor alarms on med/surg telemetry units.

“The medical/surgical areas, in our opinion, were more vulnerable, based on two principle reasons,” explains Piepenbrink. “First, the telemetry technology had limited visibility for the staff to view alarms compared to the ICUs or other technology-intensive areas. The other piece was the staffing ratios. ICUs typically are staffed at 1:1 or 1:2, whereas on the med/surg areas, that ratio is higher. Technology-intensive areas have devices in each room where staff can view and respond quicker. Additionally, managing individual patient alarms and settings is more common in these areas.”

The team’s goal was to enhance the use of the existing telemetry system without layering additional technology on top, he continues. “We are one of a few safety net hospitals in the state, and we choose our capital investments shrewdly. Our feeling was that if we were to evaluate the use of the telemetry system and the process in which staff respond to alarms, and use data to help drive change, we could leverage our current investment and create a safer and more satisfying environment for our patients and staff.”

Targeting least critical alarms
Starting in October 2011, the team began evaluating alarm data in one of the medical center’s busier telemetry units. Compiling alarm data, reviewing the definitions and assessing the risks led the task force to create some major changes to alarm defaults, says Piepenbrink. Additionally, an enhanced serial observation of how staff responded to alarms was another data point the team used to frame the pilot.

“The data was extracted from the telemetry system by clinical engineering,” he says. “Log files containing each alarm type and parameter violation were extracted, and a grid was developed to show the specific alarm parameter violations, the frequency, and also the alarm acuity. We used these data to isolate the alarms that carried the least clinical significance, and these were the primary targeted alarms that we proposed be changed.”

Says Piepenbrink “We took the approach that all audible alarms were to be considered actionable. By downgrading some less significant clinical alarms and elevating others to crisis, we created an environment where any audible alert was deemed critical in nature and required attention. The data captured prior to the pilot showed that the largest percentage of alarms captured was categorized as warning alarms. By reviewing those and creating changes to ensure that we did not miss any critical alarms, we eliminated the warning alarms, which cut out the largest quantity of alarms.”

The medical center also slightly expanded the high and low limits for Heart Rate. “They were a bit too tight and a common violator,” he says.

The task force and unit leadership were presented these data and agreed that slight changes would not negatively impact the patients on this unit. They went on to propose to senior leadership that a pilot of the changes be done on the telemetry unit.

The medical center also instituted two key elements to help support the alarm default changes. The first was enhanced order sets for the staff to use. “Additional information was embedded in the order sets to help better define specific requirements for using telemetry as well as ensuring improved communication between the physician teams and the nursing staff,” explains Piepenbrink. “Having greater clarity on orders as well as what latitude there was to create safe, patient-specific changes led to greater autonomy on the unit and better management of the patient by the entire care team.”

The second key element was giving two nurses the ability to discuss a parameter or alarm change, make the change, and then notify the physician about a change in the orders. The task force felt that staff understood the system and their patients, and that it was very unlikely that two nurses would miss something that would negatively impact care.

Culture change
The project was as much about a culture change as it was a technology change, says Piepenbrink. “We were asking the care team to change the way they manage patients. We were making some significant changes to the alarms and this was a leap of faith for them. We felt that the process was sound, that we covered this in an extensive manner (internal marketing, local education, meetings, etc.), and that we would be counting on the staff to trust the changes and the process.

“[Giving] staff an opportunity to fully understand the rationale and the process was invaluable. We shared data, listened to their concerns, educated them on the anticipated results, and provided 24/7 support to them during the pilot. The success of the pilot was a direct result of the areas taking ownership of this,” and seeing the positive impact on their work, the environment and patients.

Given the success of the project, BMC rolled out the pilot to all 330 medical/surgical telemetry beds in the winter of 2012, with similar cardiac-rhythm and heart-rate alarm reductions.

“We have been on this journey for nearly six years, with no end in sight,” he says. “Alarm management is a term that requires constant review and tweaking to ensure that we are managing and using this technology in the best way possible. We are evaluating three additional parameter changes to further reduce the alarms, and we will roll this out in the same exact way we did the first time. We will extract the data, discuss the issues, detail the process, evaluate the staff impact, measure and monitor the results.”

Contracting executives: Take note
Contracting executives can take away a lesson or two from Boston Medical Center’s alarm management project, says James Piepenbrink, BSBME, director, department of clinical engineering.

One of the most labor-intensive pieces was extracting the alarm logs from the system, he says. “Many vendors make this difficult to extract, so purchasing departments can initiate the discussion with vendors at the time of purchase…about how organizations can best utilize the technology they are purchasing.”

Contracting executives should also negotiate periodic customer education with their telemetry equipment vendors, he continues. “Manufacturers provide education at the time of installation, but little follow-up after that. Having a partnership where periodic support is available would greatly enhance the user experience and embolden the use of this expensive technology.”

Alarm primer
Not all alarms are the same, explains James Piepenbrink, BSBME, director, department of clinical engineering, Boston Medical Center.

Crisis alarms are intended to signal conditions that are life-threatening (Asystole, etc.). “There is a distinct tone to this alarm, and the alarm is stored in the alarm history,” he says.

Warning alarms are those that can carry a higher acuity, such as heart rate violations (rate too high or too low), and that reset if the patient’s condition returns to normal. The alarm has a specific sound and will be stored in the alarm history. Example: A post-op patient gets out of bed or goes for a walk, and there is a transient increase in heart rate because of the activity. The heart rate often stabilizes after a brief period, so the alarm violation – while stored in the alarm system history – stops.

Advisory alarms are the lowest acuity audible alarm. They signal parameters that do not require immediate response but that do need to be captured in the alarm history for review.