The alarm fires at 2:14 AM.
Two CNAs leave their assigned residents and walk the wing. The resident is disoriented, standing near the door, but not attempting to elope. They redirect her, reset the alarm, and return to their rounds.
This happens nineteen more times before morning.
No one calls it a crisis. It is filed under "normal operations."
But it is not normal. And the cost it generates does not appear on any line item.
The Math Nobody Does
Each false alarm response pulls two staff members off their assigned tasks for an average of eight minutes.¹
Twenty false alarms per night. Two staff. Eight minutes each.
That is 320 staff-minutes — more than five hours of nursing attention — redirected every night away from residents who were not at the door.
Annualized: nearly 2,000 hours. Almost a full FTE.
Most memory care facilities do not count this. The labor appears in the payroll — it is simply invisible as a category. It reads as "care delivered." What it actually represents is care diverted.
Alert Fatigue Is Not a Behavioral Problem
The Joint Commission identified alarm fatigue as a top patient safety concern after analyzing hundreds of adverse events tied directly to alarm desensitization.¹
The mechanism is well-understood: when alert systems produce continuous low-signal output, clinical staff develop systematic desensitization. Response times increase. Alarms are silenced before they are investigated. Staff learn — unconsciously and rationally — that most alarms mean nothing.
This is not a failure of professionalism. It is a failure of system design.
An alarm that fires without discrimination is not providing safety coverage. It is manufacturing noise. And a staff that has learned to treat alarms as noise is a staff that will, at some point, be slow to respond to the one alarm that mattered.
That is the liability. Not the false alarms themselves. What the false alarms have trained the team to do.
The Turnover Cascade
Alarm-dense environments accelerate burnout — and burnout accelerates turnover.
Annual CNA turnover in long-term care exceeds 50 percent.² The cost to replace a single CNA ranges from $3,000 to $5,000 in recruiting, onboarding, and orientation time. For licensed nurses, the replacement cost climbs to $12,000–$17,000 per departure.²
A 40-bed memory care unit with four staff departures per year attributable to burnout and working-condition fatigue represents $12,000 to $68,000 in replacement costs — generated, in part, by an alarm system meant to reduce risk.
The exit alarm is not a line item under "operations cost." It should be.
The Survey Exposure
CMS surveyors reviewing elopement incidents do not evaluate only whether an alarm was present. They evaluate whether the facility's response to that alarm was adequate.³
A documented pattern of delayed response, silenced alerts, or staff unable to account for resident location during a shift creates a finding that interventions were insufficient — regardless of whether hardware was in place.
Alert fatigue is not a defense. In a surveyor's review, it is evidence that the facility's monitoring system had degraded to the point where it could not reliably perform its function.
The alarm that produced 20 false responses before the one real event is not a system that was working. It is a system that had already failed.
The Exit Alarm Paradox
Exit alarms were designed to protect residents. They do — in environments where alarms are rare enough to command full attention.
In memory care, that condition does not hold.
An exit alarm answers a single binary question: is a resident at this door right now? It cannot distinguish a resident walking past from a resident in active pre-departure agitation. It cannot assign urgency. It cannot tell staff whether this event follows three hours of elevated restlessness or is a random positional drift.
Every event is equal severity. Which is how every event becomes background noise.
Signal, Not Noise
The solution is not fewer alarms. It is alarms that carry meaning.
Urgency scoring changes what the alert communicates. A resident whose behavioral baseline shows elevated agitation over the past two hours, increasing movement toward exit zones, and deviation from their normal nighttime pattern — that resident's door event is a signal. It warrants immediate response.
A resident whose pattern shows no deviation, who walked past the same door an hour ago without incident — that event is noise. It can be deprioritized without risk.
A system that distinguishes between the two stops burning staff attention on noise and reclaims it for the events that matter. The alarm no longer trains the team to stop listening. It trains the team to respond — because when it fires, it means something.
The staff responding to false alarms at 2:14 AM are not failing at their jobs.
They are compensating — with attention, presence, and diminishing reserves — for what the system should already know.
The cost of that compensation does not disappear when the shift ends. It accumulates in exhaustion, turnover, and the slow erosion of the response discipline that makes the next real event survivable.
An operations problem this predictable should not require an incident to measure.
Sources
¹ The Joint Commission. Sentinel Event Alert Issue 50: Medical device alarm safety in hospitals. jointcommission.org, 2013.
² PHI National. Workforce Data Center: Direct Care Workers in the U.S. phinational.org; NSI Nursing Solutions. 2024 NSI National Health Care Retention & RN Staffing Report. nsinursingsolutions.com.
³ Centers for Medicare & Medicaid Services. State Operations Manual, Appendix PP — F689 Accident Hazards. cms.gov.