Alarm Fatigue in a Step-Down Unit
A 24-bed step-down unit averages 180 alarms per bed per day, of which staff estimate fewer than 5% require action. The unit uses continuous pulse oximetry, cardiac monitoring, and infusion pumps. Two months ago a patient on telemetry had a delayed response to a sustained desaturation; the review found the alarm was audible but the assigned nurse was in another room and the alarm had been silenced repeatedly. Default alarm limits were set at device installation five years ago and never reviewed. The unit has no alarm-management committee and no policy defining who may change default limits. Staff report they 'just turn things down' to cope.
1. How would you diagnose the alarm burden before changing any settings?
Read the worked answer
Measure before intervening; otherwise you cannot tell whether changes helped.
- Alarm audit: pull alarm logs by type, parameter, time of day, and bed. Classify as actionable, non-actionable, or technical (lead-off, artifact).
- Identify the top contributors - usually a small number of parameters/limits generate most alarms. Look for limits set far from the patient's baseline, low battery/technical alarms, and duplicate alarms across devices.
- Interview staff about which alarms they ignore and why; observe response behavior.
- Review the delayed-response event against the log: was it a true alarm, was it silenced, how long until response.
Reasoning: a 5% actionability rate suggests the problem is configuration and default settings, not staff diligence. Diagnose the source before prescribing a fix.
2. What governance and configuration changes would you recommend, and who should own them?
Read the worked answer
Governance first, then configuration.
- Establish a multidisciplinary alarm-management committee (nursing, physicians, BME, respiratory, pharmacy, IT, safety) with authority to set and change default limits.
- Policy: define who may change defaults, require approval and documentation, and prohibit ad hoc silencing as a coping strategy.
- Configuration: individualize limits to patient baseline where appropriate; extend delays for parameters prone to transient artifact; reduce duplicate alarming across devices.
- Ownership: committee owns defaults; bedside staff own patient-specific settings within policy; BME owns device configuration and change control.
Reasoning: the absence of a policy is itself a finding. Without governance, any setting change is arbitrary and won't stick.
3. How would you measure whether the changes worked, and what would trigger a rethink?
Read the worked answer
Use both process and outcome measures:
- Process: alarms per bed per day (target reduction with a floor - you don't want zero), actionability rate, silencing frequency, time-to-response for true alarms.
- Outcome: adverse events, delayed responses, unplanned transfers, and staff-reported burden.
- Balancing measure: ensure the reduction did not come from disabling alarms that catch real deterioration - track missed-event or near-miss reports.
Trigger for rethink: if actionability rate rises but true-alarm response time worsens, or if any deterioration event is linked to a suppressed alarm, revisit the settings. Re-audit after 30 and 90 days and report to the safety committee.