Thinking back to my time as a nurse on a busy telemetry unit, I can tell you how most days began. Clock in by 6:55 a.m. and throw my lunch in the fridge. Then, I check the assignment board, see who is in charge and how many patients I have. Next, I had to find the nurses who will report on my four to six patients, which usually includes a quick synopsis of any pertinent issues and the plan for the day.

This sounds relatively easy. However, I am trying to accomplish these modest but critical goals amid near-constant interruptions. The nurses’ station phone is ringing; the telemetry alarms are going off; the phlebotomist is asking questions; new orders are being entered; medications are being modified or discontinued; discharge orders are being written; and transport is zipping patients off the unit for various scheduled procedures.

While all of this is going on, a patient’s pulse oximeter keeps alarming. Each time I get the alarm on my phone, I stop what I’m doing and hurry into the room to check on him. I usually find him sitting in the bed, fidgeting with the TV remote, and knocking off the pulse ox finger probe.

By the 10th time this has occurred, I am ready to stop rushing to his room because there is no intervention required.

Enough is Enough

For most of my career as a nurse, I wished there was a way to help determine if an alarm was real or not. Unfortunately, one of the biggest problems facing nurses today is that the onslaught of medical technology designed to help us care for our patients is drowning us in false alerts.

The increase in the number of medical devices with alarming capabilities, including bedside physiologic monitors, pulse oximetry machines, bedside telemetry, infusion pumps and ventilators, has only exacerbated the problem, as has the lack of standards on the proper configuration of alarm parameters.

But the quantity of alarms is just one problem among many. Many of the alarms sound the same. Additionally, individual alarms may not have any clinical relevance, but because they did “sound” an alert, a nurse is required to respond—over and over again.

This drives me crazy.

Clinical alarm signals are supposed to alert caregivers that an intervention with a patient is required. But 85 percent to 99 percent of alarms are little more than anxiety-inducing wild goose chases that send nurses scrambling and disrupt the recovery of our patients.

The problem is so widespread that healthcare associations and governing agencies, like the ECRI Institute and The Joint Commission, have documented the patient safety dangers inherent in alarm proliferation. Alarm fatigue has caused direct-care staff to arbitrarily adjust threshold settings on devices or shut them off completely, dramatically increasing the chances of a sentinel event.

Taking Control

Achieving measurable progress in clinical alarm management requires that hospitals identify and support internal champions in all relevant departments, including nurses, respiratory therapists, biomedical engineers, and information technology staff.

Two of the biggest challenges for clinical alarm management are separating clinically relevant alarms from non-actionable alarms—things like a patient sensor momentarily detaching or a temporary Wi-Fi disruption—and managing the ever-growing number of alarm-enabled medical devices.

These interdisciplinary experts must come together to assess the current state of the clinical alarm environment—by unit and facility—including establishing baseline alarm quantities by department, patient cohort, and time of day or night; reviewing current alarm settings; identifying and developing targets for reduction; implementing an agreed upon plan using a staged or phased approach; and evaluating appropriate interventions, policies, and standards.

Critically, it’s important to understand that clinical alarm management is not just about reduction of alarms. It’s also about providing the right caregiver a holistic, real-time view of a patient’s condition at the right time (i.e., before intervention is required).

Smart alarms provide this real-time perspective and can include trending alarms, which expand or contract patient alarm limits on individual devices; consecutive alarms, in which patterns of a consistent alarm detected, occurring over a clinician-defined period of time; and combination alarms, in which multiple parameters from different devices occurring simultaneously may together indicate a degraded patient condition.

The Bernoulli One platform has the ability to create smart alarms which operate on live real-time data allowing the platform to annunciate alarms independent of the patient care device alarming. One of their smart alarms is the combination alarm which is triggered when two patient values violate their limits simultaneously. The triggering conditions may be simple limit thresholds or the individual parameters may meet sustained criteria where individual parameters meet the limit criterion for a specified duration. This would certainly help reduce the number of nuisance alarms I receive on a daily basis.

Alarm management is a classic example of interdisciplinary leadership, involving clinical, IT, biomedical engineering, and other departments. Alarm management is also much more than simply reducing non-actionable alarms; it’s a gateway for more seamless care and a way for hospitals to leverage hard data to make continuous improvements to its care and response processes.