Every company in this category would like the answer to be a plain yes, and Bedford sells these systems. The published research says something more useful than yes: it depends on what you measure, and above all on whether the device is actually worn and used. This page reviews that research in plain language: what the evidence supports, what it does not support, and where it is silent.
This is a plain-language review of the published research on personal emergency response systems (PERS): the alert pendants, wristbands, and monitored buttons marketed to older adults living at home. It was prepared by the research desk of Bedford Medical Alert®, a Canadian company that sells monitored PERS. We state that interest plainly so readers can weigh it. The review was written under a standing rule that null and negative findings are reported with the same prominence as positive ones, because a page like this is only useful if every line of it can be checked.
Prepared by the Bedford Medical Alert research desk, August 2026; every source is cited in full below. Every citation was resolved live against PubMed, ClinicalTrials.gov, or the named official source on July 6, 2026, and the literature is current to that date; no figure is quoted from memory. This page is part of our research section, and a companion review covers medications and fall risk.
This page is educational. It is not medical advice, and it makes no claim that any product, Bedford’s included, produces the outcomes described in the studies below. Anyone concerned about their own fall risk, or a family member’s, should speak with a doctor or care team.
About one in five Canadian seniors living in the community reports a fall in a given year, and falls are the leading cause of injury-related hospitalization among older Canadians [1]. In 2022, 7,189 Canadians aged 65 and older died following a fall [2]. The most dangerous scenario is well documented: falling, being unable to get up, and remaining on the floor for an hour or more, which researchers call the long lie. In a classic UK study, half of the people who lay on the floor for more than an hour were dead within six months [3]. In a US study of people found helpless in their homes, two thirds of those down for more than 72 hours died, compared with about one in eight of those helpless for under an hour [4].
Personal emergency response systems exist to shorten that interval, and the research on them splits cleanly in two. The case for fast access to help is strong. The case that owning a button, by itself, changes hard health outcomes is weak: the best randomized trials found no significant effect of standard PERS or telecare packages on hospital admission, mortality, care-home placement, anxiety, or fear of falling [5, 6], and an economic analysis of the largest trial found the service, as deployed, was not cost-effective [7].
The most consistent explanation for that gap sits in the adherence literature. At the moment of need, most owners do not use the device: across studies, roughly a quarter of subscribers never wear the button, under half wear it consistently, and among owners who fell while alone, more than 8 in 10 did not press it [8, 9, 10, 11].
Read together, the evidence supports a modest, checkable position. Delayed help after a fall is genuinely dangerous. An alert system can only shorten that delay if it is on the body and activated: worn and pressed, or worn and detected. And no sensor on the market catches every fall.
About 20% of community-dwelling Canadian seniors report a fall in the previous year, and falls are the leading cause of injury-related hospitalization among Canadian seniors [1]. National surveillance recorded 7,189 fall-related deaths among Canadians 65 and older in 2022, and 81,599 fall-related hospitalizations in fiscal 2023/24 (excluding Quebec) [2]. About a quarter of Canadians aged 65 and older live alone [12], and nearly all older Canadians surveyed say they plan to live independently in their own homes as they age [13]. The full set of verified Canadian figures, with their scope notes and source links, is kept on our seniors’ falls in Canada statistics page.
The harms of delayed help after a fall have been replicated across four decades and three countries. In Wild’s 1981 UK survey of 125 home fallers, 20 people lay on the floor for more than an hour; half of them died within six months, and fallers’ one-year mortality was five times that of matched controls [3]. In a prospective US cohort of 1,103 community-dwelling adults 72 and older, 47% of non-injured fallers could not get up without help after at least one fall, and inability to rise predicted lasting decline in daily activities [14]. In the San Francisco series of 367 people found helpless or dead at home, total mortality was 28%; among those helpless for more than 72 hours it was 67%, versus 12% for under an hour, and 62% of survivors could not return to independent living [4]. In a Cambridge cohort of 110 people aged 90 and over, 80% of fallers could not get up after at least one fall, 30% lay on the floor for an hour or more, and 82% of falls happened while the person was alone [11]. The four-study story is told in full, and calmly, in the long lie: what happens when you cannot get up.
One emergency-department cohort followed 413 older fallers, of whom 115 owned a PERS. Only 18 of the 115 used it to call for help. Those who did spent less time on the ground and had lower six-month mortality, but owners as a group showed no advantage over non-owners, and the small user subgroup differed in ways that could explain part of the benefit [10]. A 12-month Australian cohort found no differences between alarm purchasers and non-purchasers in time spent on the floor or in hospitalization after an emergency [15]. The mechanism is plausible; the population-level effect is undermined by non-use, which the adherence section below documents.
Wondering what this research means for someone in your family? Call 1-888-755-3055 for a complimentary, no-pressure phone consultation. We will tell you plainly what a device can and cannot do.
The UK Whole Systems Demonstrator, a cluster randomized trial of 2,600 people with social care needs whose equipment included pendant alarms, found hospital admission rates of 46.8% with telecare versus 49.2% with usual care (not significant), and no significant effects on mortality or care-home admission over 12 months [5]. Its economic evaluation estimated roughly GBP 297,000 per quality-adjusted life year, with a 16% probability of cost-effectiveness at the UK threshold [7]. A Toronto randomized trial (86 patients given a 30-day PERS loan after an emergency visit for a fall) found identical emergency-department return rates in both arms and no significant difference in hospitalization [6].
The frequently cited positive studies are older and methodologically weak: cost-benefit modeling, demonstration projects, pre-post comparisons without control groups, and narrative reviews reporting program coordinators’ beliefs, from the 1980s and 1990s [16, 17, 18, 19]. One modern randomized trial is often miscited as proof that PERS works: in that 370-person study, both arms already had a PERS; what was tested was predictive analytics plus nurse outreach layered on top. Its primary outcome was null, with significant reductions on secondary readmission measures, and it was co-authored and funded by the device industry [20, 21].
Surveys and qualitative studies in several countries, including Canada, consistently find that users report faster help, more security, more confidence, and a feeling of being able to stay home longer [22, 24, 25, 26]. In an Ontario qualitative study of 30 subscribers aged 80 and over, participants described strong psychological value in being able to get help, alongside genuinely complex decision-making about when pressing the button is warranted [26]. Controlled studies using validated scales tell a more muted story: the Toronto trial found no significant change in anxiety or fear of falling [6]; a UK controlled study of fall detectors found no group difference in falls efficacy, quality of life, or morale, and warned that some non-wearers may lose confidence [27]; and the Whole Systems Demonstrator questionnaire study found a small benefit on one mental well-being scale and little else [28]. The reported sense of security is real and worth respecting as experience. It has not reliably converted into measured psychological outcomes.
This is the most consistent finding in the PERS literature and the least discussed in marketing. At the moment of need, non-use is the norm. In a German interview study of 52 community subscribers, 24% never wore the button and only 14% wore it around the clock; of those who fell while alone and lay on the floor for more than five minutes, 83% did not activate the alarm. Satisfaction with the service was high and did not predict use [8]. In the emergency-department cohort above, 18 of 115 owners used the device after a fall [10]. In the Cambridge over-90s cohort, call alarms were widely available but went unused in most falls that led to a long lie [11]. In a New York survey, under half of subscribers were fully compliant (wearing the button when alone); compliance was higher with self-purchase, a fall history, and proper instruction, and lower when family members made the purchase [9].
The documented reasons repeat across countries: the device comes off in the bath, the shower, and bed, which is where risk concentrates [22]; people do not want to bother anyone or do not define the event as an emergency; they wait to see whether they can manage alone [23]; and cognitive impairment, which in the Cambridge cohort was the only characteristic that predicted long lies, also makes activation less likely [11]. Among at-risk older adults who do not own a device, the main reported barriers are not seeing the need (57%), cost (37%), and not knowing the option exists (23.7%) [24].
Everything in this section describes a device that was available and not used. That specific failure, the unpressed button, is the problem automatic fall detection was built to address.
Automatic detection is the industry’s answer to the adherence problem above: a sensor designed to call for help when no button gets pressed. It removes the activation step, which is exactly where the evidence says summoning help most often fails. The published real-world evidence says the help is real and the coverage is not complete. When 13 published fall-detection algorithms were applied to 29 real-world falls of older adults, mean sensitivity was 57.0%; the best algorithm reached 82.8%, and false-alarm rates in continuous monitoring ranged from 3 to 85 per day [29]. A systematic review of 92 fall-detection projects found that only 7.1% of wearable-detector projects had been tested with older adults in real-world settings at the time [30]. Real-world validation has improved since these reviews, but near-perfect accuracy figures still come mostly from simulated falls.
One question the published literature has not answered is whether automatic detection measurably shortens time-to-help compared with button-only systems. No trial of that comparison has been published, and this page will not pretend otherwise. The case for automatic detection rests on the adherence evidence above, not on outcome data: it removes the step at which the research shows help most often fails to be summoned. It does not remove the wearing step. Automatic detection can summon help for a person who cannot or does not press; it works only while the device is worn, and on the current evidence, no sensor on the market catches every fall. Our earlier note on fall-detection research looks at the same evidence from the technology side.
A 10-year US study of 18,660 emergency responses triggered through PERS providers found that emergency crews made patient contact in 36% of responses and provided transport in 25%; lights-and-siren transports were 1.4% [31]. Many activations are cancelled before arrival or resolve as low-urgency checks. An activation starts a triage conversation. It does not automatically summon an ambulance.
The research suggests families start with two questions, and neither is about a feature list. If your parent falls when no one is there, how long would it be until someone knows? And will the device actually be on their body, and pressed, when it happens?
Involve the older person in the decision. Compliance is measurably higher when people choose the device themselves and are properly instructed, and lower when family buys it for them [9]. Ask how the device will be worn in the shower and overnight, because the research is specific about where wearing stops: the bath, the shower, and bed [22].
Agree in advance on what counts as an emergency. People hesitate to press because they do not want to make a fuss, or they wait to see whether they can manage alone [23]. A plain rule set together, for example “if you are on the floor and cannot get up, press it, every time,” addresses the documented failure mode directly.
If the device offers automatic fall detection, treat it as a second layer behind wearing and pressing. It helps exactly where the adherence evidence says people fail, the moment no button gets pressed, and it carries the limits described above: it works only while worn, and it does not catch every fall.
If wearing or pressing is unrealistic, say so early. For people with cognitive impairment the evidence is specific: they are at the highest risk of a long lie and the least likely to activate a button [11]. Non-wearable approaches to fall detection exist as a category. Whether any particular approach fits a particular person is a conversation for the family and the person’s care team, with the real-world limits above in mind.
Treat outcome claims skeptically, from any vendor, Bedford included. If marketing says a device is “proven to reduce hospitalizations” or quotes a near-perfect detection percentage, the null-results section and the real-world sensitivity figures above are the check. The strongest honest claim this category can make is about access to help when it is needed, not about guaranteed health outcomes.
And none of this replaces medical judgment. Fall risk has medical causes that a doctor or care team can assess and sometimes reduce. A device, worn or not, is not a substitute for that conversation.
Bedford sells monitored medical alert systems, which is the commercial interest declared at the top of this page. Two things Bedford offers exist because of the adherence evidence above: wearable devices with automatic fall detection, such as the Go Anywhere watch, built to call for help when no button gets pressed, and Silver Shield, a non-wearable radar sensor for homes where wearing anything is unrealistic; when it detects a fall it alerts family or caregivers automatically, under protocols the family sets. The research frame on this page applies to Bedford’s products exactly as it applies to anyone else’s: a wearable helps only while it is worn, no sensor on the market catches every fall, and the strongest honest claim in this category is about access to help, not guaranteed outcomes.
The null trials tested specific services in specific systems: UK telecare as deployed around 2008 to 2010, and a 30-day device loan in Toronto. Those results establish that the tested deployments did not change the measured outcomes; they do not prove that no alert service could ever help anyone. But the burden of proof sits with anyone claiming outcomes, and no trial has yet met it.
No modern systematic review or meta-analysis of PERS effectiveness on hard outcomes exists. A targeted search of 2015 to 2026 reviews returned one device-accuracy review and one experience-focused review; neither addresses mortality, hospitalization, or institutionalization.
Caregiver outcomes are an evidence hole. Reviews and qualitative studies describe reassurance for families, but no controlled study quantifying caregiver burden reduction from PERS surfaced in any search. There are no numbers to cite, so this review cites none.
Cost-effectiveness in a Canadian public-payer context is unknown. The only trial-grade economic evaluation is British and was negative [7]; an Australian estimate of ambulance savings relied on a simulated comparator rather than an observed one [32].
Observational comparisons in this field are confounded. Device users differ from non-users; in the emergency-department cohort, users were younger, less likely to have dementia, and more likely to live alone [10]. Findings that depend on comparing them should be read accordingly.
This review is targeted, not a registered systematic review, and it was prepared by a seller of the product category it examines. The disclosure appears in the opening section, and the rule of reporting null findings at equal prominence exists precisely because of that interest.
Searches and citation resolution were performed on July 6, 2026 using PubMed (search, metadata, citation lookup, and identifier conversion) and ClinicalTrials.gov, plus web search for grey literature from named official sources (the Public Health Agency of Canada, Statistics Canada, and the National Institute on Ageing). Every PMID and DOI was resolved live during the research session, and each figure quoted is taken from the resolved record’s abstract or from the named official source. Preprints are not cited: the preprint interface used supports only date and category filters, with no keyword search and no geriatrics category, so a systematic preprint sweep was not possible. No PERS-specific review from Canada’s Drug Agency (CADTH/CDA-AMC) surfaced in searches, so none is cited.
If fast access to help is the piece you want to solve, start with a complimentary, no-pressure phone consultation at 1-888-755-3055. No long-term contracts, and the equipment is included with the subscription.
Prepared by the Bedford Medical Alert research desk. Literature search July 6, 2026; page published August 2026. This public edition is derived from an internal evidence brief; the internal brief is a source document and is not published.
