Solving the deeply complex problem of excessive alarms will require a multipronged strategy

?Walk through the halls of just about any hospital, and a cacophony of sounds—from a steady monotone signal and high-pitched beep to an intermittent ping and constant whoosh—will most likely assault your ears. Clinical alarm alerts have reached astronomical proportions—a University of California San Francisco study reports that one hospital experienced approximately 2.5 million alarms in a single month. In spite of broad industry efforts to manage this overload of alarms, solutions to the problem remain elusive.

Alarm fatigue is not the only concern. A multipronged approach is required to address a complicated situation that entails a disparate array of causes such as unrecognized alarm conditions, often due to inappropriate configuration practices; failing to reset medical device limits when a new patient is connected; choosing inappropriate monitoring settings for different measurement categories; and selecting priority levels that don’t match the patient’s condition.

Addressing the Problem

In 2002, The Joint Commission (TJC) attempted to address the clinical alarm management issue and created the first National Patient Safety Goals (NPSG) program. Little progress was made, and 5 years later, the organizations took a different approach, requiring that alarm decibels be turned up. That move proved to be counterproductive to the care environment.

[sidebar float=”right”]

Lessons from Boston Medical Center

Read an exclusive online case history detailing how James Piepenbrink’s department of clinical engineering at BMC worked with a task force of nurse managers, nurse educators, residents, attending physicians, and IT staff to implement a successful alarm management program.


A 2011 Medical Device Alarms Summit sponsored by the Association for the Advancement of Medical Instrumentation (AAMI), the US Food and Drug Administration, the American College of Clinical Engineering, and TJC identified clarion themes and challenges and prioritized actions. Fast-forward some 4 years, and little has changed. While some health systems have witnessed a modicum of success, regulatory bodies are hoping to achieve more widespread positive outcomes.

In 2013, TJC redefined NPSG, making a more concerted effort to ameliorate the alarm dilemma. In Phase I, hospitals are required to establish alarm management as an organizational priority and to identify the most important alarms within the institution by January 1, 2014. Phase II takes effect on January 1, 2016, and mandates that hospitals implement policies and processes for managing alarms.

Robert Maliff, director, Applied Solutions Group at the ECRI Institute, notes that the issue of alarm management has been on its “shout” list for years, and his group has actively been involved in helping hospitals tackle the problem. He surmises that most hospitals have complied with Phase I requirements of NPSG, but that they are concerned about next steps. “Everyone has a base analysis, but how do you change practices and culture?” he says. Many hospitals are unsure of how to begin the process of determining the number and source of alarms, assessing clinician workflow, and finding viable solutions.

The first order of business, according to Maliff, is to engage leadership in overseeing alarm management policies, strategies, and implementation. “This is a critical point. Senior leaders have to set expectation and deadlines,” he says, adding that “2015 will be a year of practice change.”

In 2014, AAMI’s Healthcare Technology Safety Institute formed the National Coalition for Alarm Management, a group of healthcare thought leaders charged with sharing resources and developing alarm management guidelines that will enable all health systems across the country to improve patient safety and care.

Julie A. Reistetter, MS, RN, chief nursing officer for Banner Telehealth, is one of nearly 80 individuals from healthcare institutions, academia, industry, and professional organizations that belong to the coalition. “Our goal is to create a compendium of recommendations,” she says, explaining that cooperation among all entities will be necessary to achieve positive outcomes. “The ultimate goal is alarms equal action.”

Changing Clinical Practice

Although hospitals realize they have to address the issue of alarm management, many need guidance to ensure they embark on an appropriate and efficient course of action. Maria Cvach, DNP, RN, FAAN, assistant director of nursing, clinical standards, for The Johns Hopkins Hospital, points out that alarms in critical situations, such as those for ventricular tachycardia or ventricular fibrillation, are seldom overlooked.

“They make distinct noises so those alarms don’t get missed,” she says. “What gets missed are the subtle ones that occur over time; for instance, a minor arrhythmia, slowly falling blood pressure, or when leads come off when a patient moves.”

Cvach identifies five gaps that affect a hospital’s ability to implement a workable alarm management program:

  • Inability to analyze events to understand the root causes.
  • Lack of evidence related to how to set alarm settings.
  • Little understanding of what kind of signals (melodic, symphonic, vibrational, etc) alarms should elicit.
  • Dearth of knowledge about whether and how long to monitor.
  • Failure to understand the best type of technology to alert clinicians.

In addition, clinical workflow needs to be taken into consideration. In the past 20 to 30 years, more private rooms, as well as more high-acuity units, have been created, and caregivers are responsible for a higher number of patients, according to Tim Gee, “connectologist” and principal at Medical Connectivity Consulting. He adds that units have also become geographically larger, which may affect the auditory impact of medical device alarms. In order to manage alarm safety, hospitals must improve clinical practice as it relates to alarm generation. “They have to make sure [alerts] are reliably transmitted to the caregiver, which is a notification problem,” says Gee, adding that data and analytics are crucial to understanding the underlying issues. “You can’t manage what you can’t measure.”

Additionally, within many hospitals, each department is responsible for its own equipment. Forcing the institution out of a comfortable silo mind-set to work in cross-functional fashion poses another great challenge, according to Gee. To further confuse matters, hospitals often use various vendors for different pieces of equipment, which increases the complexity of buying, installing, and updating devices. Also, while alarm information is available in older devices, it may be more difficult to extract.

Weighing the Options

So how and where should a hospital start? “It can’t be the hospital alone, and it can’t be the scientist alone,” Cvach says. “It takes an interdisciplinary approach to formulate ideas and come up with solutions. You need to merge scientists, health care, and vendors.” For an effective solution, she explains, “You need a discussion between people who work with devices, purchasers, and designers about how to make a better system. The situation involves not only alarm fatigue, but also all alarm hazards. The goal is to drive the false alarm rate down by incorporating delays and multiparameter settings. There are multiple components.”

Allen Enebo, senior product manager for Extension Healthcare, suggests that hospitals take a step-by-step approach. “Hospitals have to plan out the project, looking at workflow and the overall concept of interruptions. It’s not just a matter of alarms,” he says. “To have a full solution, you need to manage all the alerts and messages that go to the nurse and cause interruptions.”

Companies like Extension gather data to determine needs, which facilitates reporting and analytics to measure improvement. “It varies between individual department, patients, and rooms. You have to have the data to understand how to improve the care,” Enebo says. He points out that measurements should also be taken at different times of the year, such as during flu season, when inpatient census may rise. The goal is to obtain metrics that encompass all alarm-related data, including baseline, progression, and outcomes.

But once the data has been collected and analyzed, the work is not done. Enebo emphasizes the importance of conducting follow-up studies to chart ongoing progress. “The hospital needs to do analysis with a fine-toothed comb,” he says. To be successful or show improvement, you have to continue to run reports, and the data needs to be easy to understand and meaningful.

Paul McGurgan, national program director, alarm management, for Excel Medical Electronics, asserts that the quality assurance process is iterative and comprises many layers. When Excel’s software captures waveforms, it enables the user to go back to see if the alarm is “true,” an artifact, or nonactionable. “The value of our Navigator is that it drills down to the patient or time or specific alarm,” he says. This extracted data is put into a database from which custom reports are created. The initial report offers a high-level, system-wide alarm profile.

The next report provides measures at the departmental level, and the final piece drills down to individual beds. “Initially, you need to see the global perspective, then peel back the layers a little at a time. The data doesn’t always give you an answer. It might give you more questions, but it points you in the right direction,” McGurgan says.

While equipment is an important part of the equation, human interface is another key component, according to McGurgan. Drawing attention to small measures, such as proper skin preparation and lead placement, can contribute to reducing nonactionable alarms.

Extracting data from devices can be time and manpower intensive. Gee explains that hospitals have two options when it comes to alarm notification systems: an in-house solution or a purchased one. “Everyone wants to be automated, but no one wants to change things. Ideally, you should no longer be doing something manually if it’s automated. This applies to all aspects of alarm safety,” he says.

Clinical engineers have the expertise and training to take on this task. But given the time-intensive aspect of the task, hospitals may opt to purchase a middleware system that automatically aggregates data from devices and applies rules to automate workflow related to some task. The vast amount of captured information ranges from orders, diagnostic results, medication lists, alarms, and other patient-specific data. The resulting rules, which can be customized based on device, patient, and environment, may include delays, an escalation notification feature, and/or multiple parameters.

Gee notes that when considering middleware, a hospital should make sure the manufacturer interfaces with its medical devices so the systems are compatible. Hospitals should also make sure the middleware supports mobile devices. “You want a system that understands the event from beginning, middle, and end, since an alarm initiates action during every stage. Remember that you’re not just buying a widget, but creating a system of systems to do specific tasks,” Gee says.

Know Your Default Settings

While understanding the types and number of alarms within a facility, as well as key related data, puts the hospital on the right track, users all need to have a working knowledge of the manufacturer’s default settings, and of when to change them and why. According to Bob Steurer, global product marketing manager for Spacelabs HealthCare, a manufacturer’s default settings are determined through clinical analysis. The settings are reviewed, based on feedback from clinicians and clinical educators, every 12 months, but they rarely change, he notes.

“We look at clinical feedback in a clinical setting. The manufacturer’s defaults are just the starting points. You need to customize them for each patient,” Steurer says.

But many hospitals often don’t understand that changing default alarm settings to reflect patient condition, location, and other critical related data can enhance patient care. So offering a thorough education and training program should be an integral part of any device purchase.

“It’s the manufacturer’s job to make sure the customer/partner understands this. It’s a standard part of the package,” he says. “We give an in-service when we sell equipment that goes beyond the basics. We teach the alarm limits, the settings and default settings, and inform them that they can modify [settings] per unit.”

He adds that changes don’t have to be big. For instance, for SPO2, a slight downward adjustment from 92 to 90 will reduce the number of false alarms significantly, according to Steurer. Or users could increase the delay time, which would also decrease nonactionable alarms, he adds.

Policies on who is authorized to change alarm settings also differ from one institution to another. Some hospitals require nurses to obtain a doctor’s order before modifying a setting, while others might use a team approach in which two caregivers evaluate the patient and agree to adjust the setting. Selecting an option that expedites appropriate changes to alarm settings tends to improve patient care, Steurer points out.

Monitor watchers have been suggested as another option as part of an alarm management solution. Although some facilities have decided to add this component to their alarm management plans, Steurer sees little value in the practice. “Like any other caregiver, what’s to say they won’t get alarm fatigue and miss an alarm? Monitor watchers won’t resolve the issue. They are just another set of eyes,” he says.

There is also the question of who should be monitored, according to Cvach. Some hospitals monitor every patient on a continuous or intermittent basis. “But there is no evidence to say that one way is better than another. At Johns Hopkins, the medical ICU nurses get
60 to 70 alarms per bed per day on just the patient monitor,” she notes. “You have to use other cues; otherwise, you would be chasing alarms all day. I want to highlight that the alarm in and of itself doesn’t drive response.” She recommends creating a hierarchy of alarms.

Working with Vendors

A successful alarm management program depends on the confluence of technology, patient acuity, workflow, and process, as well as collaboration with vendors. More important, a hospital’s policy and procedure plan should include a robust training program. Brian Gross, MSc, BSEE, RRT, SMIEEE, fellow scientist, clinical systems architect, Patient Care and Monitoring Solutions, at Philips Healthcare, points out that when a hospital buys an electronic medical record, it usually offers 8 to 15 hours of training.

But when the institution purchases a medical device with an alarm, training often consists of only 10 to 30 minutes. “Clearly, there is an opportunity for on-site training staff to work with the vendor team. Beyond that, as the facility creates an alarm committee, it should include the vendor in the process,” he says. “Vendors can help you determine how to quantify what you’re running and has insight on the technology. Once the alarm committee establishes a charter baseline, vendors can make recommendations, develop capabilities, and tell the hospital what others are doing.”

Banner Health in Arizona is a prime example of vendor-institution collaboration. Reistetter explains that Philips Healthcare collaborated with the health system to create a multiparameter alarm that considers individual settings which, in combination, provide value added to the device and to patient care. “The medical record number of every patient admitted to the ICU is entered at the central station. All parameters come into the software, eICU, which runs an algorithm and identifies adverse trends,” she says. “We can identify when a patient is trending negatively and intervene before something happens. Eyes and ears are constantly looking at data. If a patient alarms at a certain threshold, but other cues don’t indicate a problem, the program tells us the setting could be lowered.”

Can vendors do a better job when it comes to alarms? Gross thinks so. “We need to conduct more research to find newer ways to detect more appropriate settings. We also need to work with customers to better understand the challenges, including alarm recording and what is the current technology,” he says. “We also need to work with the product team to enhance algorithms and reduce risk. Advocacy with AAMI and other regulatory agencies will help provide guidance on what [these agencies] want to see that might not comply with standards. We need to create an environment in which we can learn from each other.”

As national healthcare systems await recommendations from the National Alarm Coalition, they continue to attempt to unravel the complications associated with alarm management through in-house task forces, sharing information with other institutions and referring to available online resources. Awareness of the problem has spawned a flurry of action, which hospitals hope result in positive outcomes this time around. 24×7

Phyllis Hanlon is a contributing writer for 24×7. For more information, contact editorial director John Bethune at [email protected].