BRENNEN

Industries / Health / Healthcare / Senior living and assisted care

Falls and safety

Falls and safety is keeping residents from harm inside the building: falls and the fractures they cause, restraints and bed rails, fire, and infections that spread from resident to resident.

7

Milestones achieved, from 1990 to 2022.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1990

    Fewer residents tied down

    Physical restraints are belts, vests and other devices that stop a person moving freely, long used to keep residents from falling or wandering. OBRA 1987 limited them in US nursing homes from 1 October 1990. Their use fell from 37.4% to 28.1% of residents between 1990 and 1993 in one 254-home study, and later national data put it at about 19% in 1995 to 2000, 6.9% in 2005 and 1.4% in 2015.

    Lessona practice once seen as the safe choice almost disappeared under a federal rule backed by inspections.

    ebi.ac.ukpmc.ncbi.nlm.nih.gov

    Credit

    US Congress (OBRA 1987); C. Hawes and colleagues; S. A. Fashaw, K. S. Thomas, E. McCreedy and V. Mor

  2. 2001

    Japan aims for zero restraint

    When Japan's long-term care insurance began in 2000, the operating standards for insurance-funded care facilities banned physical restraint except when urgent and unavoidable. In March 2001 the health ministry's Zero Restraint Campaign council published a guide to ending restraint, which many providers still use. Kyoto Prefecture reported in March 2010 that the share of its facilities using restraint had nearly halved.

    pref.kyoto.jpjri.co.jp

    Credit

    Japan's Ministry of Health, Labour and Welfare

  3. 2006

    Bed rail gaps get limits

    People can be trapped, and die, in gaps between a bed rail, the mattress and the bed frame. The US Food and Drug Administration (FDA) received about 691 entrapment reports from 1985 to 2006, including 413 deaths; long-term care facilities reported most of the entrapments. A workgroup of the FDA, the US Department of Veterans Affairs, Health Canada, manufacturers and care providers named seven zones where people get trapped, and in March 2006 the FDA set gap limits based on the breadth of a head (120 mm) and of a neck (60 mm), checked with a cone-and-cylinder test tool. In July 2023 the US Consumer Product Safety Commission made similar tests mandatory for portable bed rails sold for adults, after counting 284 entrapment deaths from 2003 to 2021; the proposed rule's count linked 50 deaths to nursing homes and 40 to assisted living.

    Lessona hazard became a pass-or-fail check once it was defined by measurements anyone with the tool can repeat.

    fda.govgovinfo.gov

    Credit

    FDA and the Hospital Bed Safety Workgroup; US Consumer Product Safety Commission

  4. 2007

    Hip pads fail a trial

    Hip protectors are padded underwear meant to absorb the blow of a fall onto the hip. In the HIP PRO trial, 1,042 residents of 37 nursing homes around Boston, St Louis and Baltimore wore a protector on one hip only, so that each person's unprotected hip served as the comparison. The trial was stopped after 20 months: fractures occurred on 3.1% of padded hips and 2.5% of unpadded ones, with adherence at 73.8%.

    Lessona device that absorbs force in the lab still has to be tested on real falls, and this design gave no protection even among residents who wore it most.

    ebi.ac.uk

    Credit

    D. P. Kiel and colleagues

  5. 2008

    Sprinklers in every nursing home

    In 2003 fires at nursing homes in Hartford, Connecticut, and Nashville, Tennessee, killed 31 residents. Neither home had automatic sprinklers, and federal standards did not require them, the Government Accountability Office reported in July 2004. A Medicare and Medicaid rule of 13 August 2008 required every long-term care facility to be fully sprinklered by 13 August 2013. It counted 2,446 facilities that needed systems, at a cost of about USD 847 million, and cited fire association data showing an 82% lower chance of death in a sprinklered building.

    Lessonthe deaths came in buildings without sprinklers; CMS noted that no long-term care facility sprinklered throughout had ever had a fire that killed more than one person.

    gao.govfederalregister.gov

    Credit

    US Government Accountability Office; Centers for Medicare & Medicaid Services (CMS)

  6. 2022

    A falls program that worked

    A 2018 Cochrane review of 71 trials in care facilities found the effect of exercise and of combined fall-prevention programs uncertain; only vitamin D, in residents low in it, probably reduced the rate of falls, by about 28%. The UK FinCH trial then randomized 84 care homes to the Guide to Action Care Home program, which trains staff and gives them a checklist to find each resident's fall risks and act on them, or to usual care. The unadjusted falls rate in the program homes was 0.57 times the usual-care rate (95% confidence interval 0.45 to 0.71), without limiting residents' activity, and the program was cost-effective at the thresholds the NHS uses.

    Lessona follow-up study found the program worked in different ways in different homes, and in some, training and awareness alone cut falls.

    ebi.ac.ukebi.ac.ukebi.ac.uk

    Credit

    P. A. Logan and colleagues, funded by the National Institute for Health Research

  7. 2022

    A court faults COVID guidance

    In 2020 care home residents made up 47% of COVID-19 deaths in June and 46% in October across 21 countries, by figures from the London School of Economics' International Long-Term Care Policy Network summarized by the US health department; Canada reported 85% in June and South Korea 8% in October, though countries counted deaths differently. On 27 April 2022 England's High Court ruled that the government's March 2020 policy on discharging hospital patients and its April 2020 care home admissions guidance were irrational, because they did not advise that a new resident without symptoms, unless tested negative, be kept apart from other residents for 14 days.

    Lessonguidance built around symptoms failed, in the court's words, to take into account the risk of spread by people without them.

    aspe.hhs.govcaselaw.nationalarchives.gov.uk

    Credit

    Lord Justice Bean and Mr Justice Garnham, in a case brought by Cathy Gardner and a second claimant on behalf of their fathers

Watch for

  1. Lamps that spot falls

    Nobi, a Belgian company, sells smart lamps with sensors to nursing homes. It says the lamp detects a person on the floor and flags a possible fall, lights the room when someone sits up in bed, and greatly shortens the time before staff arrive; no figures were given. An outside 2026 study of an unnamed camera-based fall detector in a 102-bed care facility and a 50-suite assisted living building found frequent false fall alerts, especially for residents in reclined wheelchairs.

    Achieved whenan outside evaluator reports how many falls the lamps catch and how many false alerts they raise per resident per day in care homes.

    healthcare-brew.comjournal.gerontechnology.org

    Claimed by

    Stijn Verrept, Nobi's co-founder, April 2026

  2. Early warning from sewage

    A team in Ontario analysed 5,095 weekly floor swabs from 10 long-term care homes, taken from August 2021 to January 2023, together with city wastewater data and weather. It reports that adding wastewater and weather improved how well the swabs predicted the 25 COVID-19 outbreaks in those homes, with wastewater among the two strongest signals. The paper is a preprint, not yet reviewed by other scientists. A US company, Aquora Research, says its NIH-funded models anticipated outbreaks in long-term care up to seven days ahead, with no publications listed.

    Achieved whena peer-reviewed study tests a model in homes not used to build it and reports lead time and false alarms.

    medrxiv.orgpandemicpact.org

    Claimed by

    Michael Fralick (Sinai Health) and colleagues, October 2025; Aquora Research on its grant record

Open

  1. 01

    Phone check for bed gaps

    An app that measures the gaps in entrapment zones 1 to 4 of a bed, rail and mattress from a phone's depth scan, flags any over the FDA limits, and is checked against the FDA's cone-and-cylinder tool. Tests use empty beds only.

    The brief
    Credit · unclaimed
  2. 02

    National outbreak warning backtest

    A public, weekly-updated test of whether wastewater levels of COVID-19, influenza A and RSV rise before cases among nursing home residents, state by state, using only published aggregate data, with lead time and false alarms for each state and virus.

    The brief
    Credit · unclaimed

Open 01 · the brief

Phone check for bed gaps

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An app that measures the gaps in entrapment zones 1 to 4 of a bed, rail and mattress from a phone's depth scan, flags any over the FDA limits, and is checked against the FDA's cone-and-cylinder tool. Tests use empty beds only.

Why it is still open

The FDA's 2006 method, which the 2023 bed rail rule also uses, places a physical cone-and-cylinder tool in each gap by hand. A 2019 patent held by Huntleigh Technology uses a depth camera of the Kinect type around the rails, but it watches for a body part or object crossing a keep-out zone while the bed moves; it does not measure the gaps. No published phone or depth-camera method checked against the FDA tool was found.

fda.govpatents.google.com

Built with what exists

An iPhone Pro or iPad Pro with a LiDAR depth sensor; Apple's ARKit; the Open3D point-cloud library; a commercial copy of the FDA test tool; second-hand hospital beds, rails and mattresses.

  1. First step

    Borrow one bed and one test tool, scan ten rail and mattress combinations, and compare the readings by hand.

  2. Done looks like

    50 bed, rail and mattress set-ups measured with both the app and the tool; app readings within 5 mm of the tool in at least 95% of zone checks, and no set-up that fails with the tool passed by the app.

Open 02 · the brief

National outbreak warning backtest

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public, weekly-updated test of whether wastewater levels of COVID-19, influenza A and RSV rise before cases among nursing home residents, state by state, using only published aggregate data, with lead time and false alarms for each state and virus.

Why it is still open

Studies so far sample the sewers or floors of a few homes themselves: five homes in Valencia, Spain, where virus in a home's sewage came 5 to 19 days before cases were found, with one home's false signal traced to a neighbouring building's sewer (Davó and colleagues, 2021), and 10 Ontario homes in the October 2025 preprint. The NIH-funded Aquora project lists no publications. No public comparison of city wastewater with resident cases across a whole country was found.

pmc.ncbi.nlm.nih.govmedrxiv.orgpandemicpact.org

Built with what exists

CDC's weekly NHSN data on COVID-19, influenza and RSV among nursing home residents by state (data.cdc.gov dataset tscn-ryh9, reported since 30 September 2024 and required for all three viruses from 1 January 2025); CDC's National Wastewater Surveillance System data on data.cdc.gov; Python and pandas.

  1. First step

    Download both data sets and plot one state's three viruses, wastewater against resident cases, on one chart.

  2. Done looks like

    For every state and each of the three viruses since late 2024, the median weeks of warning, the share of wastewater rises not followed by a rise in resident cases, and code and results published and rerun each week.