Alarm Fatigue and False Alarms: Why Fixed Thresholds Fail Babies

A monitor that alerts constantly is a monitor parents learn to ignore. That is not a minor annoyance, it is one of the most underappreciated failure modes in infant monitoring, and it has a name in clinical research: alarm fatigue. Here is what the evidence actually shows, and why the direction that research points to is personalized, not generic, thresholds.

What science says: the scale of the false-alarm problem

Alarm fatigue is a clinically documented phenomenon: repeated exposure to alerts, most of which are false, causes caregivers to become desensitized and slower to respond, including to the alerts that matter. This is not a fringe concern, it has been flagged by patient-safety bodies as a serious risk in monitored care settings.

The scale of the problem is striking. A comprehensive observational study across consecutive intensive care unit patients recorded 2,558,760 unique alarms over a 31-day period, an audible alarm burden of 187 alarms per bed per day. Of 12,671 arrhythmia alarms annotated by nurse scientists with 95% inter-rater reliability, 88.8% were false positives (Drew et al., 2014). In neonatal intensive care specifically, the alarm burden has been found to be similarly high, with the large majority of critical arrhythmia alarms turning out to be artifactual rather than real events (Li et al., 2018).

Why fixed thresholds generate so many false positives

A fixed threshold applies one limit, drawn from a population average, to every infant monitored. Normal heart rate, breathing rate, and movement patterns vary substantially from one healthy baby to the next, and even for the same baby across sleep states, feeding, and age. A threshold set for the "average" baby will mechanically fire for any baby whose normal physiology sits outside that average, even though nothing abnormal is happening.

This mismatch between a single cutoff and wide biological variability is the mechanical root of the false-alarm rates documented above. It is not a flaw in any one device; it is a structural property of any system that uses one number for every baby.

Alarm fatigue: a documented patient-safety issue, not just an inconvenience

Alarm fatigue has real consequences beyond irritation. Research on exposure to nonactionable physiologic monitor alarms found a measurable association between how many false alerts staff had already been exposed to and how quickly they responded to a subsequent alarm in a children's hospital setting (Bonafide et al., 2015). The pattern is intuitive and troubling at the same time: the more often an alert turns out to be nothing, the more a person's brain reasonably (but dangerously) starts predicting that the next one will be nothing too.

This dynamic is why hospital safety programs now specifically target alarm reduction and smarter alarm design, rather than simply adding more alerts, as a way to improve response quality. Fewer, more meaningful alerts protect attention; more, less reliable alerts erode it.

What the research direction looks like: individualized baselines

If the core problem is a mismatch between a single population threshold and wide individual variability, the logical fix is to stop comparing every baby to the same number. An individualized baseline approach learns what is typical for one specific baby, their own resting heart rate, their own breathing pattern, their own usual movement at night, over an initial period, and only flags a genuine deviation from that baby's own pattern rather than from a generic average.

This is precisely the direction Mothair's health reference profile takes: rather than applying one fixed number to every baby, it builds a personalized baseline over time so that alerts reflect a real change for that child, not simply a normal baby sitting outside an average. For more detail on how that specific feature works, see our article on Baby's Health Reference Profile. This article's purpose is different: to lay out the clinical evidence for why the false-alarm problem exists in the first place, and why personalization is the research-informed answer to it, not to describe the feature mechanics themselves.

Important: Mothair is a wellness tracking device for infants. It is not a medical device under EU Regulation 2017/745 (MDR) and never replaces professional medical care. The information in this article comes from public scientific sources and does not constitute medical advice, consult your pediatrician for any question about your baby's health or monitoring needs.

Does this apply to home baby monitors too?

The studies cited above were conducted in hospital intensive care and neonatal care settings, where alarm rates and false-positive rates are best documented and most rigorously quantified. Home consumer monitors have not been studied at the same scale or with the same rigor, so the specific false-positive percentages above should not be assumed to transfer directly to a nursery.

What does transfer is the underlying physiological principle: babies vary widely in their normal heart rate, breathing, and movement, and a single fixed cutoff cannot account for that variability without generating unnecessary alerts for some healthy babies while potentially under-alerting for others. That structural mismatch is why the research direction, in both clinical and consumer contexts, points toward individualized thresholds rather than one-size-fits-all ones.

FAQ

What is alarm fatigue?

Alarm fatigue is the well-documented desensitization that happens when clinicians or parents are exposed to frequent alerts, the vast majority of which are false. One landmark ICU study found 88.8% of annotated arrhythmia alarms were false positives, which pushes people to ignore or silence alerts, including the rare real ones (Drew et al., 2014).

Why do fixed thresholds generate so many false alarms?

A fixed threshold applies the same limit to every infant, even though normal heart rate, breathing, and movement vary widely from baby to baby. A threshold calibrated on a population average will mechanically trigger for any baby whose normal values sit outside that average, even when nothing is actually wrong.

What is an individualized baseline?

It is an approach that learns a specific baby's own normal values, their own resting heart rate, usual breathing rhythm, typical movement patterns, so it only flags a genuine deviation from that baby's own habits, rather than a deviation from a generic population average. This is the logic behind Mothair's health reference profile.

Does this problem also affect consumer baby monitors?

The cited studies focus on hospital monitors in intensive care and neonatal units, where the phenomenon is best documented and quantified, but the underlying physiological principle, wide individual variability not accounted for by a single threshold, applies just as much to home sleep-tracking devices.

Key Takeaways

• Alarm fatigue is a clinically documented patient-safety issue, not a minor inconvenience: 88.8% of annotated arrhythmia alarms in one large ICU study were false positives (Drew et al., 2014).
• Fixed thresholds fail because they apply one number to a population with wide individual physiological variability.
• Exposure to prior false alarms measurably slows response time to later alarms, including real ones (Bonafide et al., 2015).
• Individualized baselines, learning what is normal for one specific baby, are the research-informed direction for reducing false positives.
• This principle, documented in hospital settings, applies structurally to home monitoring even though the exact false-positive rates have not been studied at the same scale outside hospitals.