Making Alarms Meaningful
By Sarah Richardson, CIO, NCH Healthcare
Alarm hazards are on the radar of safety institutions like the Association for the Advancement of Medical Instrumentation, Emergency Care Research Institute and The Joint Commission.
In this context, at NCH Healthcare System (NCH) of Naples, Fla., we are working hard to mitigate alarm fatigue.
This issue came to the forefront in September 2013, when we started sending secondary alarms to Apple iPhones. For a variety of reasons, our nurses received too many nuisance alarms.
In response, we dramatically limited the types of alarms that would fire on a nurse’s smart phone. We also formed an interdisciplinary Alarm Safety Committee, which implemented a number of interventions—such as proper skin preparation and changing the telemetry leads daily—to limit alarm misfires.
We also worked to perfect a new Alarm Management Shift Report, which displays the frequency, number and type of alarms for each bed by the hour. This data helps us mitigate non-actionable alarms and focus on alerts, with the potential impact to patient safety.
Reducing non-actionable alarms
In 2015, we incorporated this shift report into the workflow of a de-centralized, 45-bed telemetry unit. The report helped our clinicians adjust telemetry monitors and other interventions, which lowered the total number of alarms from 255,912 in January 2015 to 79,486 in April 2015—a 69 percent reduction in non-actionable alarms — without a negative impact on patient safety.
Today, we use this data and these reports to spot and correct a number of issues and take action to reduce unnecessary alarm noise. Through this data-driven process, we are minimizing alarm fatigue and making the organization a quieter, better care environment.
“Our alarm management team has come a long way,” said Kevin Smith, RN, a nurse manager who played a key role in developing the shift reports. “Really, there have been dramatic improvements. We’ve changed the whole culture of alarm safety.”
Challenges and barriers
As we worked through these alarm management issues, we faced a number of challenges, including:
• The lack of concurrent data
• Little data in the literature to guide the timing and escalation path of secondary alerting
• The realization that there were many alarms that we needed to evaluate
To mitigate these challenges, we created a standard set of retrospective reports and on-demand concurrent data. The data and staff feedback, continue to help us refine secondary alerting best practices and standards. While creating these policies, procedures and education rollouts are time-consuming; they have helped us safely implement our alarm management program.
With access to this data, our Alarm Safety Committee began to trust the analysis, which is now part of our basic process improvement initiative. The data helped us identify important alarms, implement an evidence-based approach, and measure the impact of the change.
Another challenge was getting staff members to “buy into” the proposed changes. As mentioned previously, our initial smart phone rollout did not go smoothly. The nurses got sick of the phones and tuned out the alarms.
Initially, nurses were skeptical when we proposed changes that would make a difference in the number of alarms. We took an important step when we let nurses change some of the parameters for these alarms, versus a “one size fits all” approach.
Allowing nurses to tailor the alarms for each individual patient (in line with the patient’s care plan and with permission from a physician) played a key role in getting them to incorporate smart phones back into their daily routines.
“Through this data-driven process, we are minimizing alarm fatigue and making the organization a quieter, better care environment”
As much as possible, we share the following lessons learned with other organizations.
Follow Data-driven Processes — In the battle against alarm fatigue, use data-driven processes, as well as retrospective and concurrent data, to build a successful alarm management system.
We also support the recommendations of the Association for the Advancement of Medical Instrumentation, Emergency Care Research Institute and The Joint Commission. Refer to these organizations to identify best practices and the most current evidence on alarm management.
Establish Sound Governance — Another key to reducing redundant, non-actionable alarms is understanding your alarm ecosystem. Form an interdisciplinary alarm committee with a sound governance structure. This committee can help you collect actionable data and additional metrics, and create a strong process improvement methodology.
Train your Clinicians — You may need to train your nurses to use smart phones. These phones may be ubiquitous outside the four walls of most hospitals, but within healthcare, they’re unusual. Typing on a small keyboard can be challenging, and many people don’t even own a smart phone.
Remember: Clinicians respond better to instructors with clinical backgrounds.
Do a Full Data Review — Finally, be sure to do a full data review on what alerts are firing and how often before you decide to send them to a nurse's phone. If you are able to collect data beforehand and understand it, you will be much smarter in how you deploy your smart phones.
We are rolling out our alarm management improvement processes and shift reports to all of our ICUs and other parts of the organization. Other improvements are also in the works.
Today, for example, the Alarm Management Shift Reports go to charge nurses every eight hours. In the future, we plan to give nurses access to real-time Alarm Hot Spot dashboards, which will show all of the individual alarms on their unit. These dashboards will help nurses figure out why the alarms are firing.
Importantly, we also are studying and quantifying the link between alarm management and patient outcomes. In the ICU, we plan to use a critical care acuity solution to measure possible links between severity-adjusted outcomes and alarm management.
Ultimately, we want to develop ways to proactively alert clinicians to potential adverse events before they occur—it’s the least we can do for our patients.