A useful alarm communicates a relevant condition and leads to a clear response.
In brief: alarm fatigue is the progressive loss of attention or response capacity caused by repeated exposure to frequent, redundant or poorly actionable alerts. Managing it requires measuring the actual burden, prioritising according to clinical risk, adapting limits and delays, defining escalation rules and reviewing outcomes without compromising the detection of critical events.
IEC 60601-1-8 establishes priority categories, consistent signals and requirements for alarm systems in medical electrical equipment. Compliance at device level must be complemented by system-wide management: monitors, ventilators, pumps, medical gases, electrical supply, HVAC and integration systems all operate simultaneously in an ICU or operating room.
Why alarm fatigue in ICUs and operating rooms is a system problem
An observational study published in JMIR recorded an average of 152.5 alarms per bed per day in the ICU studied. This figure should not be treated as a universal reference. It shows how records can identify sources, time periods and concentration patterns before an intervention. Without a baseline, a change may reduce apparent noise while worsening the response.
Four phenomena that are often combined should also be separated: technically false alarms; true but clinically non-actionable alarms; duplication of the same condition; and relevant alarms that arrive late, without context or to the wrong person. Each problem requires a different measure.
| Observed signal | Cause to investigate | Evidence required before making a change |
|---|---|---|
| Many brief alarms that resolve themselves | Overly sensitive limits, artefacts or lack of an appropriate delay | Duration, parameter, patient status and proportion requiring intervention |
| The same condition appears on several devices | Duplication between the device, central station, repeater and secondary system | Propagation map, timings and recipients |
| Alerts are acknowledged, but the response is delayed | Unclear ownership, overload, poor audibility or lack of escalation | Time from activation to acknowledgement and action |
| Alarms are silenced during routine tasks | Configuration poorly adapted to the clinical workflow or insufficient training | Reason, silence duration, context and automatic recovery |
The clinical objective is to increase the proportion of signals that lead to the correct decision without missing important events.
A five-decision process: measure, classify, configure, verify and govern
1. Measure the actual burden, not perception
The inventory should include physiological, technical and infrastructure alarms. Frequency, duration, priority, acknowledgement, action and resolution should be recorded for each source. Staff perception and system records provide different information: a highly disruptive alarm may be infrequent, while another that attracts little attention may recur hundreds of times.
2. Classify by actionability and risk
Priority should relate the colour and tone assigned by the manufacturer to the detected condition, its clinical consequence, the time available to respond and the responsible professional. A technical alert may be critical if it compromises system availability; a physiological deviation may not require intervention if it is expected during a specific phase.
3. Configure with context and safe limits
Customising thresholds, introducing justified delays or removing duplication can reduce non-actionable alerts. Each change must have a defined scope, owner, clinical criterion and rollback option. A 2025 study in a neonatal ICU observed an 84% reduction in alarms through filtering and delays, with no significant differences in the critical events studied. The result applies to the population, devices and protocol examined in that study.
4. Verify the complete response
Testing must confirm alarm audibility, identification of its source, comprehension of its priority, delivery to the correct role, escalation and traceability. It should also cover network failures, loss of a screen, temporary silencing and return to the normal configuration.
5. Govern changes throughout the lifecycle
Alarm behaviour changes when a monitor is replaced, equipment is added, a protocol is modified or staffing is reorganised. Alarm management should therefore be integrated with operational continuity in operating rooms and ICUs and the governance of connected equipment.
How to apply alarm management to HERMES and Q Panel
Within the Tedisel ecosystem, HERMES centralises operating room controls and, depending on the configuration, includes functions related to medical gas, ventilation and electrical supply alarms. Q Panel can integrate alarm repeaters, medical gas controls, displays and other elements. This centralisation brings the information together; clinical priority must be defined within the project.
Scope limit: HERMES and Q Panel form part of the environment’s control and display infrastructure. The safety of the complete system depends on the connected equipment, available interfaces, agreed configuration, hospital protocols and clinical responsibility.
The digital architecture of the connected operating room provides the integration layer; alarm management adds rules for priority, context and response. Connecting signals without governing them can transfer the noise to a central screen.
Acceptance matrix: what the hospital should test before operation
| Test | Verifiable criterion | Participants |
|---|---|---|
| Source and priority | Each signal identifies the equipment, condition and urgency level without ambiguity. | Supplier, clinical engineering and clinical users |
| Audibility and visibility | The alarm can be perceived at the intended work positions under realistic operating noise. | Occupational safety, engineering and the clinical team |
| Escalation | If the first recipient does not respond, the alert changes level or reaches the defined role. | IT, nursing and department management |
| Silence and pause | Duration, visual indication, recovery and permissions match the protocol. | Users, clinical engineering and supplier |
| Integration failure | Loss of network, screen or interface does not remove the primary alarm or conceal degraded mode. | IT, security, supplier and users |
| Logging | Activation, acknowledgement, change and resolution are recorded with a synchronised timestamp. | IT, quality and clinical leads |
Centralisation is useful when it reduces the distance between signal and decision. If it merely accumulates alerts, an orderly interface becomes another point of saturation.
Infographic: from noise to a verifiable clinical response
The infographic summarises the project logic: measure the burden, prioritise by risk, contextualise the signal and convert it into a response with an owner, escalation path and evidence. Configuration is one part of the process; validation, review and governance keep the system current when equipment and protocols change.
Less noise only matters if the response improves
Managing alarm fatigue requires understanding the actual system, intervening according to clinical criteria and verifying the effect. The decisive metric measures how many relevant signals are acknowledged, understood and addressed within the intended timeframe, as well as how many alarms are removed.
HERMES, Q Panel and the connected infrastructure can centralise display and control within this strategy. The outcome depends on an architecture that coordinates equipment, people and protocols. In critical care interoperability, reducing cognitive load requires retaining the context needed to interpret each signal.
Technical sources: IEC 60601-1-8:2006+A1:2012+A2:2020; Poncette et al., JMIR 2021; and Kalden et al., Acta Paediatrica 2025.
Frequently asked questions about alarm fatigue
What is alarm fatigue in an ICU?
It is a reduction in attention or response capacity caused by repeated exposure to frequent, redundant or poorly actionable alarms. It can lead to delays, inappropriate silencing and difficulty distinguishing genuinely critical signals.
Does reducing alarms mean lowering safety limits?
Safe alarm reduction involves removing duplication, artefacts and non-actionable alerts through justified limits, delays, maintenance, context and escalation. Every change must be validated and reversible.
What data should be measured before changing the configuration?
At a minimum: source, type, priority, frequency, duration, time, acknowledgement, action, resolution, recipient and relationship to the clinical condition. Silencing, escalation and incidents should also be recorded.
Can HERMES eliminate alarm fatigue?
HERMES can centralise controls and certain alarms depending on the configuration. Complete management depends on the devices, integrations, protocols, responsible roles and hospital validation.
What is the difference between centralising and managing alarms?
Centralising brings signals together in one interface. Managing defines which signals are relevant, how they are prioritised, who responds, when they escalate, what happens during a failure and how performance is reviewed.
Who should govern clinical alarms?
A multidisciplinary team involving clinical management, nursing, clinical engineering, engineering, IT, security, quality and suppliers. Every decision should have an owner and be documented.
We can examine how to integrate display, control, equipment and clinical workflow from the project stage. Talk to Tedisel Medical.