Medical Alert Systems: Most Who Needed One Did Not Press It

If you are reading this because someone you love is unsteady on the stairs, the honest starting point is not a product comparison. It is a number.

What actually happens when an older person falls alone

The BMJ study followed 110 people aged over 90 — 90 women and 20 men — for a year, using fall calendars, phone calls and visits rather than relying on memory. Across 265 fall reports:

The authors are direct about what the long wait costs. Lying on the floor for a long time was “strongly associated with serious injuries, admission to hospital, and subsequent moves into long term care.”

That last item is the one families underestimate. The fall is an event. The hours afterwards are what often ends independent living.

The finding that inverts the sales pitch

Here is the sentence, verbatim:

“Call alarms were widely available but were not used in most cases of falls that led to lying on the floor for a long time.”

Read it twice. These were not people who lacked a device. The alarms were there. In most of the cases that ended in a long wait on the floor, nobody pressed them.

The study names the reasons, from the older people and their carers themselves:

• Perceptions of irrelevance — the device is for someone frailer, someone older, not me
• Concerns about independence — pressing it is an admission
• Practical difficulties — it was in the other room, on the charger, taken off for a bath

None of those is a hardware problem. Buying a more expensive system does not solve any of them.

The authors’ own conclusion goes to design and training rather than to purchase: people “need training in strategies to get up from the floor,” and “work is needed on access and activation issues for design of call alarms and information for their effective use.”

Does automatic fall detection fix it?

It is the obvious answer to an unpressed button, and it is genuinely useful. It is also oversold.

A 2017 validation study in PLoS One makes the gap explicit. Research groups, it notes, “have reported up to 100% accuracy in detecting falls in experimental settings.” Then the researchers tested against real life: 19 older adults and five younger adults wore tri-axial accelerometers through roughly 400 hours of ordinary daily activity, during which the older adults experienced 10 unanticipated real falls.

The result:

• 8 of the 10 real falls were detected from a single waist or sternum sensor
• No false alarms across about 28.5 hours from younger adults
• Among older adults, false positives ranged from 0 to 0.3 per hour

The authors describe this as higher detection and substantially lower false-positive rates than existing systems — and still call for bigger real-world validation.

Eight out of ten is meaningfully better than a button nobody presses. It is not the “100%” of a laboratory or a brochure. Buy it as a second layer, not as a guarantee.

The same paper notes the background reality plainly: “about half of all older adults who fall are unable to get up without assistance,” and the extended time on the floor “often leads to medical complications, including muscle damage, dehydration, anxiety and fear of falling.”

What this means for choosing one

The evidence reframes the decision. The question is not which system has the longest battery or the most features. It is which one this particular person will wear and use.

Wearability beats specification. A pendant that comes off for the bath is absent during one of the highest-risk activities in the house. Ask what happens in the shower before you ask about range.

Fall detection covers the unpressed-button problem, partially. On the best real-world data available, roughly eight falls in ten. Treat the other two as the reason the button still matters.

The independence objection is the real obstacle. The study found it in the participants’ own words. A conversation that frames the device as protecting independence — the alternative to a long floor wait is often the move into care — does more than any feature list.

False alarms have a cost. Up to 0.3 an hour in the older-adult data. A system that cries wolf gets switched off, which returns you to square one.

And practise getting up. The BMJ authors put training in strategies to get up from the floor alongside the technology, not after it. It is free.

Who is most at risk of the long wait

The study looked at what predicted difficulty and what predicted the long floor wait, and they were not the same thing.

Difficulty getting up was consistently associated with age, reported mobility, and severe cognitive impairment.

But for lying on the floor a long time, cognition was the only characteristic that predicted it.

That matters for families. A physically frail parent with intact cognition may struggle to rise but will call for help. Where memory or judgment is affected, the device is least likely to be used at the moment it is needed most — which is precisely where automatic detection earns its cost.

What this page is not saying

Not that medical alert systems do not work. The evidence here is about how they are used, not whether summoning help quickly matters. Getting help sooner after a fall plainly matters, and the study’s own concern is that people were not getting it.

Not that this cohort represents everyone. These were 110 people aged over 90 in and around Cambridge, followed for one year. The oldest old fall differently from a robust 68-year-old, and the figures above should be read as describing that group.

And not a substitute for a falls assessment. Recurrent falls are a medical event with causes — medication, blood pressure, vision, neuropathy, muscle loss — that are often treatable. We go through the causes in sudden leg weakness in seniors: 12 distinct causes. A device shortens the wait after a fall. It does nothing about why the falls are happening.

The short version

82% of falls in this cohort happened when the person was alone. 80% of fallers could not get up at least once. 30% lay there an hour or more, and that long wait tracked with serious injury, hospital admission and moves into long-term care.

The alarms were already in the houses. In most of the cases that ended badly, they were not used.

So buy for activation, not for features: something worn in the shower, something that detects a fall when the button is not pressed, and a conversation that makes using it feel like keeping independence rather than losing it.

Frequently Asked Questions

Do medical alert systems actually get used in a real fall?

Often not, and this is the most important thing to know before buying one. The BMJ cohort study of people aged over 90 found, verbatim, that “call alarms were widely available but were not used in most cases of falls that led to lying on the floor for a long time.” The devices existed in the homes; the button was not pressed. The researchers recorded the reasons from participants and carers: perceptions that the alarm was irrelevant to them, concerns about independence, and practical difficulties such as the device not being to hand. That is why the purchase decision should centre on whether the specific person will wear and activate it, rather than on the specification sheet.

How long do older people actually lie on the floor after a fall?

Longer than most families expect. In that one-year cohort of 110 people over 90, of those who fell, 30% (20 of 66) had lain on the floor for an hour or more, and 80% (53 of 66) were unable to get up without help after at least one fall. Across all 265 fall reports, 54% described the person being found on the floor. The consequences are the reason the number matters: the authors report that lying on the floor for a long time was strongly associated with serious injuries, admission to hospital, and subsequent moves into long-term care.

How accurate is automatic fall detection?

Better than a button that goes unpressed, and well short of the marketing. A 2017 PLoS One validation study noted that research groups “have reported up to 100% accuracy in detecting falls in experimental settings,” then tested performance against real life: across roughly 400 hours of ordinary daily activity, in which 19 older adults experienced 10 unanticipated real falls, the system detected 8 of the 10 from a single waist or sternum sensor. False positives ranged from 0 to 0.3 per hour among older adults, and there were none in about 28.5 hours of younger-adult data. Plan for roughly eight in ten, and keep the manual button as the backup rather than the other way round.

Does my parent really need one, or am I overreacting?

The figures suggest the risk is usually underestimated rather than overstated, particularly for anyone living alone. In this cohort 82% of falls happened when the person was alone, and the alone-ness is the whole problem: a fall with someone in the house is an incident, while a fall with nobody there can become hours on the floor. That said, this study followed people aged over 90, and a healthy 68-year-old is in a different risk category. The more useful question than age is whether they live alone, whether they have already had a fall, and whether they could get up unaided if they had one tonight.

Which matters more, fall detection or the button?

They cover different failures, which is why the better systems have both. The button fails when the person chooses not to press it, and the BMJ data shows that happened in most of the cases that ended in a long floor wait. Automatic detection covers exactly that gap, but misses roughly two falls in ten on real-world data and can generate false alarms. So: detection for the falls where the person will not or cannot call, the button for the two in ten detection misses and for emergencies that are not falls at all — chest pain, a stroke, a fire.

What predicts who will end up waiting on the floor?

Two different things predict two different problems in this study, and the distinction is practical. Difficulty getting up after a fall was consistently associated with age, reported mobility and severe cognitive impairment. But for the long wait on the floor specifically, cognition was the only characteristic that predicted it. In plain terms: physical frailty makes it hard to rise, but cognitive impairment is what predicts nobody being called. If memory or judgment is affected, automatic detection stops being a nice extra and becomes the part you are actually buying.

Is there anything free that helps?

Yes, and the study’s authors put it alongside the technology rather than beneath it: people “need training in strategies to get up from the floor.” Practising a safe method of getting from the floor to a chair, while someone is present to supervise, addresses the same problem the device does and costs nothing. Alongside that, the ordinary falls-prevention work — a medication review, checking blood pressure and vision, removing trip hazards, strength and balance exercise — reduces how often the situation arises in the first place. A device shortens the wait; none of it addresses why the falls are happening.

What should I ask a provider before buying?

Ask the questions the evidence points at rather than the ones the brochure answers. Is it waterproof and worn in the shower, given how much falling happens there? What is the real-world fall detection performance, and what is the false alarm rate? What happens if the wearer does not respond when the monitoring centre calls? Is there a monthly fee, and what does cancellation involve? Does it work outside the house? And the one families forget: what does the wearer think of it, because the study’s central finding is that an unworn or unpressed device protects nobody.

Sources

• Fleming J, Brayne C; Cambridge City over-75s Cohort (CC75C) study collaboration. Inability to get up after falling, subsequent time on floor, and summoning help: prospective cohort study in people over 90. BMJ 2008 Nov 17;337:a2227. PMID 19015185. PMCID PMC2590903. DOI 10.1136/bmj.a2227. One-year follow-up of 110 participants (90 women, 20 men) aged over 90, using fall calendars, phone calls and visits. Verbatim results: “54% (144/265) of fall reports described the participant as being found on the floor and 82% (217/265) of falls occurred when the person was alone. Of the 60% who fell, 80% (53/66) were unable to get up after at least one fall and 30% (20/66) had lain on the floor for an hour or more. Difficulty in getting up was consistently associated with age, reported mobility, and severe cognitive impairment. Cognition was the only characteristic that predicted lying on the floor for a long time. Lying on the floor for a long time was strongly associated with serious injuries, admission to hospital, and subsequent moves into long term care. Call alarms were widely available but were not used in most cases of falls that led to lying on the floor for a long time. Comments from older people and carers showed the complexity of issues around the use of call alarms, including perceptions of irrelevance, concerns about independence, and practical difficulties.” Authors’ conclusions include that “people need training in strategies to get up from the floor” and that “work is needed on access and activation issues for design of call alarms and information for their effective use.” Competing interests: none declared. Checked 25 August 2026. https://pubmed.ncbi.nlm.nih.gov/19015185/
• Aziz O, Klenk J, Schwickert L, Chiari L, Becker C, Park EJ, Mori G, Robinovitch SN. Validation of accuracy of SVM-based fall detection system using real-world fall and non-fall datasets. PLoS One 2017 Jul 5;12(7):e0180318. PMID 28678808. PMCID PMC5498034. DOI 10.1371/journal.pone.0180318. Five young adults and 19 older adults wore tri-axial accelerometers through approximately 400 hours of activities of daily living; the older adults experienced 10 unanticipated falls. Verbatim: “Research groups have reported up to 100% accuracy in detecting falls in experimental settings. However, there is a lack of studies examining accuracy in the real-world setting”; “our system was able to detect 8 out of the 10 falls in older adults using signals from a single accelerometer (waist or sternum)”; “our system did not report any false alarm during approximately 28.5 hours of recorded data from young adults. However, with older adults, the false positive rate among individuals ranged from 0 to 0.3 false alarms per hour.” Also verbatim on background: “about half of all older adults who fall are unable to get up without assistance” and the extended period on the floor “often leads to medical complications, including muscle damage, dehydration, anxiety and fear of falling.” Competing interests: none declared. Checked 25 August 2026. https://pubmed.ncbi.nlm.nih.gov/28678808/
• Scope note, stated as a limitation rather than a finding. The percentages above describe a specific population: 110 people aged over 90, mostly in and around Cambridge, England, followed for one year. They are not incidence rates for older adults generally, and they should not be read as applying to a healthy person in their sixties or seventies. The fall-detection figures come from a single validation study whose own authors call for larger real-world datasets. No product was tested, compared or endorsed in either paper, and nothing on this page reports the performance of any commercial system. Checked 25 August 2026. https://pubmed.ncbi.nlm.nih.gov/19015185/

Before deciding on a device. The strongest predictor of who actually falls is not a fall history or a physical test. In a prospective analysis, balance confidence explained 38.7% of the variance and the tests predicted nothing — what to ask instead.