Reducing alarm fatigue starts with cutting the number of alarms that don’t require action. On most hospital units, the vast majority of monitor alarms are either false, self-correcting, or clinically irrelevant. Between 2009 and 2012 alone, 98 alarm-related sentinel events were reported to The Joint Commission’s database, and 80 of those resulted in patient death. The problem isn’t that nurses stop caring. It’s that hundreds of alarms per shift train the brain to tune them out.
Fixing this requires changes at every level: how electrodes are placed, how thresholds are set, how teams communicate about alarms, and how technology routes notifications. Here’s what actually works.
Why Default Settings Cause Most of the Problem
Most hospitals run their monitors on factory-set defaults designed for a “normal healthy adult.” That means every patient whose baseline falls outside that narrow window, whether from COPD, heart failure, or chronic atrial fibrillation, generates a constant stream of alarms that no one needs to act on. The Agency for Healthcare Research and Quality describes these defaults as useful when a patient first arrives on a unit, acting as a safety net for detecting major deviations. But once you have enough data on a patient’s own normal, keeping those generic thresholds active just adds noise.
A patient with COPD whose resting oxygen saturation sits around 88% will trigger low-SpO2 alarms all day on a default threshold of 90%. A clinician who knows that patient’s baseline might reasonably lower the alarm to 80%, the point where actual intervention would be needed. That single change eliminates dozens of meaningless alerts per shift for that one patient.
Customizing Alarms to Individual Patients
The most effective intervention across the research is shifting from population-level defaults to patient-specific alarm parameters. The AACN’s practice alert on alarm management recommends that nurses customize alarm settings for individual patients and check those settings at the start of every shift, with any change in patient condition, and with any change in caregiver. This isn’t a one-time task. It’s an ongoing process.
In practice, customization means reviewing each patient’s trending data and adjusting heart rate ranges, SpO2 low limits, and arrhythmia alerts to reflect what’s actually abnormal for that person. Many hospitals require a physician order or dual-nurse sign-off before alarm limits can be changed, so implementing this takes coordination. If your unit has that kind of policy barrier, it’s worth raising with your alarm management committee, because the friction discourages nurses from making changes that would genuinely improve safety.
One medical ICU intervention lowered the default SpO2 alarm from 90% to 88% and extended the delay on the audible alarm from 10 seconds to 15 seconds. That extra five seconds filters out the brief, self-correcting dips that account for a large share of nuisance alarms, without compromising the detection of real desaturation events.
Electrode Prep and Probe Placement
A surprising number of alarms aren’t clinical at all. They’re technical artifacts from poor electrode contact. Standard preparation for cardiac monitoring includes shaving the skin at the electrode site, cleaning it, and drying it thoroughly before applying the electrode. Some protocols add light abrasion with fine sandpaper to remove the outermost layer of dead skin, which improves electrical conductivity.
The AACN specifically calls out proper skin preparation and electrode placement as a frontline strategy. Replacing electrodes that have been on too long (most manufacturers recommend every 24 to 48 hours), checking that leads aren’t pulling or kinked, and making sure pulse oximetry probes are on the correct finger with good perfusion can eliminate a significant portion of the “leads off” and artifact alarms that pile up throughout a shift.
Using Safety Huddles to Target High-Alarm Patients
One of the more practical approaches tested in hospitals involves adding alarm discussion to the daily safety huddles that many units already hold. In a cluster randomized trial, researchers generated paper dashboards each weekday morning showing the four patients on each unit with the highest number of high-acuity alarms over the previous four hours. Any patient with fewer than 20 alarms in that window wasn’t included. The dashboards listed the most frequent alarm types, current settings, and a script with checkboxes for changes agreed upon by the team.
Charge nurses led these discussions during the existing huddle, and the results were striking in terms of follow-through. Of the huddle discussions where changes were agreed upon, 95.6% of those changes were actually enacted at the bedside. The most common actions were:
- Discontinuing monitoring entirely (32.0% of changes), for patients who no longer needed it
- Monitoring only when asleep or unsupervised (23.8%), for patients who were stable but still on telemetry
- Widening heart rate parameters (12.7%)
- Changing ECG leads or wires (8.6%)
- Changing the pulse oximetry probe (8.0%)
- Increasing the SpO2 alarm delay (4.7%)
The biggest takeaway here is that nearly a third of changes involved taking patients off monitoring altogether. Unnecessary monitoring is one of the largest drivers of alarm fatigue, and it often persists simply because no one formally reviews whether a patient still meets criteria for it.
Secondary Notification Systems
Technology can help filter what reaches you and how. Secondary notification systems, sometimes called middleware, sit between the bedside monitor and the nurse’s mobile device. Instead of every alarm sounding overhead or at the central station, the system applies logic: it can delay, escalate, or suppress notifications based on the alarm type, duration, and priority level. Only alarms that meet certain criteria get forwarded to the assigned nurse’s smartphone.
One pediatric hospital implemented this alongside standardized alarm defaults and a nursing education module across all non-ICU areas. The key advantage is that nurses receive notifications specific to their assigned patients rather than hearing every alarm on the unit. This preserves the ability to respond quickly to genuine emergencies while reducing the background noise that drives desensitization.
Building a Unit-Level Alarm Management Program
Individual nurses can improve their own practice by prepping skin properly, checking settings every shift, and advocating for patient-specific thresholds. But lasting change requires a unit-level or hospital-level program. The AACN recommends that nursing leaders establish an interprofessional team to gather alarm data, identify the highest-volume alarm types, and develop unit-specific default parameters.
The Joint Commission’s National Patient Safety Goal on alarm safety (NPSG.06.01.01) requires hospitals to make improvements ensuring that alarms on medical equipment are heard and responded to on time. This gives nurse leaders institutional backing to push for changes that might otherwise get deprioritized.
A practical starting point: pull your unit’s alarm data for one week. Identify which monitors and alarm types generate the highest volume. In most cases, you’ll find that a small number of patients and a handful of alarm categories account for a disproportionate share of the total. Interventions targeted at those specific sources, whether it’s adjusting defaults for your patient population, improving electrode protocols, or establishing criteria for discontinuing monitoring, will have the largest impact with the least disruption to workflow.
One mixed-methods study found that a comprehensive alarm management program reduced high-priority technical alarms by 61.1%. The total alarm count in that study actually increased slightly due to other factors, which underscores an important point: success isn’t always about reducing the raw number. It’s about reducing the alarms that don’t matter so nurses can trust and respond to the ones that do.