Acute Frailty

An acute admission is dangerous for a frail older person: about three in ten leave less able to manage daily life than before they fell ill, rising to six in ten over the age of 90; most able-bodied older inpatients spend most of the day in bed; and one in four has delirium. Part of that loss starts with the illness before admission, and part is caused by what happens in hospital, where lying in bed, imposed bed rest and delirium are the measurable harms. The content of good acute frailty care rests on trials: geriatric assessment, ward-based geriatric care, delirium prevention and early movement. The clocks the NHS sets around it (30 minutes, 1 hour, 6 hours, 70 hours a week) rest on professional consensus, and no study tests them. The national measures count speed and flow, not whether people end up alive, at home and able to do what matters to them; NHS England's own 2026 guide says its frailty metrics are not nationally mandated or collected. Abroad, the strongest signal is how deeply geriatricians are involved, not the badge a service carries.

What practitioners take as self-evident, tested part by part

1. “Earlier is better.”

   • Long waits for care in an acute episode harm older people.

     Harm from exposureLow confidenceConsistent, graded observational evidence across England and France; see what the setting itself does, above. · PMID 35042695 · PMID 37930696 · PMID 40229528

   • Giving the same specialist input sooner than it is already given improves outcomes.

     What worksVery low confidenceWhere timing has been randomised, results lean to no gain: HIP ATTACK, ENABLE III, AVERT. · PMID 32050090 · PMID 25800768 · PMID 25892679

   Intuition is right about delay and not shown about speed. The evidence fits a threshold: long waits harm, while bringing forward care that is already reasonably prompt has not helped in the trials that tested it. Avoiding long delays is the better-supported aim.

2. “Hospital is bad for frail older people, and ten days in bed ages their muscles by ten years.”

   • Many older people leave hospital less able than they were.

     How often it happensModerate confidenceAbout three in ten, rising steeply with age and frailty (Loyd et al., 2019; Covinsky et al., 2003). · PMID 31734122 · PMID 12657063

   • Bed rest causes loss of strength and fitness.

     Harm from exposureModerate confidenceExperiment in healthy volunteers (Kortebein et al., 2008). · PMID 18948558

   • The hospital stay, rather than the illness that caused it, causes most of the loss.

     Harm from exposureLow confidenceTwo thirds of the loss at discharge had begun before admission (Covinsky et al., 2003); immobility in hospital is strongly linked to decline but confounded by severity (Brown et al., 2004). · PMID 12657063 · PMID 15271112

   • Ten days in bed ages an older person's muscles by ten years.

     Harm from exposureVery low confidenceNo study read states a years-of-ageing equivalence; it appears to be an extrapolation from a 2007 research letter on bed rest and skeletal muscle in healthy older adults, whose figures could not be read at source because PubMed holds no abstract. The companion study found strength fell but everyday physical performance did not, and in acutely admitted patients grip strength and muscle mass did not fall significantly. · PMID 17456818 · PMID 18948558 · PMID 28286250

   The substance is right and well evidenced: many frail people leave hospital worse, and lying in bed weakens muscle. The slogan is not: it is an extrapolation, not a finding. The more exact and more useful point is that much decline starts before admission and much of what happens in hospital (immobility, delirium, bed rest without a reason) is modifiable.

3. “Staying in hospital longer than necessary is worse for frail older patients.”

   • Excess stay waiting to leave causes harm.

     Harm from exposureVery low confidenceMostly low-quality studies; those delayed are frailer to begin with (Rojas-García et al., 2017). · PMID 28898930

   • What happens during the days in hospital (lying in bed, delirium) is linked to decline.

     Harm from exposureLow confidenceMobility was the strongest process factor in decline and length of stay itself had a small direct effect (Zisberg et al., 2015; Brown et al., 2004; Witlox et al., 2010). · PMID 25597557 · PMID 15271112 · PMID 20664045

   • Getting people up and moving during an acute stay prevents the loss.

     What worksModerate confidenceRandomised trial: structured exercise from the first days of the stay, mean age 87, reversed the loss of function seen with usual care (Martínez-Velilla et al., 2019). · PMID 30419096

   Likely true, and unproven as a count of days. The evidence points less at the number of days than at how they are spent: immobility, imposed bed rest and delirium are where the harm is measurable, and getting people moving is the one part a trial has tested and found to work. A shorter stay spent in bed is not the same gain as a stay spent upright.

What an acute admission does to older people with frailty

1. About three in ten older people admitted to an acute hospital go home less able to manage daily life (washing, dressing, getting to the toilet) than they were before they became ill. The rate has not fallen as stays have shortened.

   How often it happensModerate confidenceDaily activitiesMeta-analysis of cohorts and trials, 7,375 people aged 65 or over in acute medical and surgical care: 30% (95% CI 24% to 33%). Down from high for large heterogeneity, partly because daily activities were measured in different ways (Loyd et al., 2019). · PMID 31734122

2. The risk climbs steeply with age and with frailty: from 23% of people aged 70 to 74 to 63% of those aged 90 or over, and, after an admission, new disability within a month in 35% of frail older people against 5% of those who are not frail.

   How often it happensModerate confidenceDaily activitiesCohort of 2,293 medical inpatients aged 70 or over (Covinsky et al., 2003); community cohort of 754 people aged 70 or over followed monthly for ten years (Gill et al., 2010). Both from the United States, the first from the 1990s. · PMID 12657063 · PMID 21045098

3. Much of the loss is already there when the person arrives. Of the 35% who left worse than two weeks before admission, two thirds had declined during the illness before admission and did not recover; a third lost function for the first time in hospital. A further 20% declined before admission and recovered during the stay.

   How often it happensModerate confidenceDaily activitiesOne two-hospital cohort of 2,293 people; function two weeks before admission was recalled, sometimes by a relative (Covinsky et al., 2003). The illness that leads to admission, not only the stay, drives the decline. · PMID 12657063

4. Older inpatients who can walk spend most of their stay lying in bed: 83% of the time in one study, a median of 17 hours a day in another, with a median of about 43 minutes a day on their feet.

   How often it happensModerate confidenceTime uprightTwo accelerometer cohorts, 45 men in a United States veterans hospital and 43 medical patients in Denmark; small, but closely consistent (Brown et al., 2009; Pedersen et al., 2012). · PMID 19682121 · PMID 22972940

5. About one in four older medical inpatients has delirium, and the rate has not changed in forty years.

   How often it happensModerate confidenceDeliriumMeta-analysis of 33 studies using standard diagnostic criteria: 23% (95% CI 19% to 26%), no change from 1980 to 2019; adult medical inpatients, not only older people, with signs of growing publication bias (Gibb et al., 2020). · PMID 32239173

6. After new disability in hospital, recovery within a year is the minority outcome: of people discharged with new or added disability, 41% had died at a year, 29% were alive but not recovered and 30% were back to where they had been. Some recover later.

   How often it happensLow confidenceRecovery and deathOne cohort of people aged 70 or over admitted to general medicine in 1993 to 1998 (Boyd et al., 2008). Graded low, the more cautious of two passes, because it is a single, older cohort, and recovery continues past a year: among disabled older women who declined after a hospital stay, half recovered fully within 30 months (Boyd et al., 2009). · PMID 19093915 · PMID 19694869

7. Frail older inpatients die more often, go home less often and come back sooner than those who are not frail.

   How often it happensModerate confidenceDeath, home, readmissionMeta-analysis of 57 cohorts: death RR 2.26, discharge home RR 0.79, readmission at 30 days RR 1.47 (Kamo et al., 2026). In over a million older people in English hospitals, the fifth with the highest Hospital Frailty Risk Score had 30-day mortality OR 1.71 and long stay OR 6.03, though the score predicts poorly for any one person (Gilbert et al., 2018). Same direction everywhere; the size varies widely. · PMID 42642794 · PMID 29706364

What the setting itself does, and how sure we are it is the cause

1. Bed rest itself takes strength and fitness from older people: ten days in bed cut leg strength by about 13% and aerobic capacity by 12% in healthy volunteers in their late sixties. Their scores on everyday physical tasks did not change.

   Harm from exposureModerate confidenceStrength and fitnessExperiment, 11 healthy volunteers aged about 67, no control group (Kortebein et al., 2008). Direct evidence that bed rest causes the loss; indirect for frail people who are also ill, and in acutely admitted patients a meta-analysis found no significant fall in grip strength or muscle mass over the stay (Van Ancum et al., 2017). Low as applied to frail inpatients. · PMID 18948558 · PMID 28286250

2. Older inpatients who barely move do far worse, in a graded way, and much of their immobility is imposed rather than needed: a third had bed rest ordered, and nearly 60% of bed-rest episodes in the least mobile had no recorded medical reason.

   Harm from exposureLow confidenceFunction, care home, deathCohort of 498 medical inpatients aged 70 or over: low mobility against high, loss of daily activities OR 5.6, new care-home admission OR 6.0, death OR 34.3 (Brown et al., 2004). In a path analysis of 684 patients, in-hospital mobility was the strongest process factor in decline, and length of stay itself had a small direct effect (Zisberg et al., 2015). Sicker people move less, so the size of the link overstates the cause; an odds ratio of 34 for death is itself a sign of confounding. · PMID 15271112 · PMID 25597557

3. Long waits for a bed are linked to more deaths and more delirium, rising with the length of the wait: from five hours onwards in England, a night on an emergency-department trolley in France, and over ten hours for delirium.

   Harm from exposureLow confidenceDeath, deliriumEngland, all ages, 5.2 million admissions: deaths within 30 days rose with every hour from five to twelve, one extra death per 82 patients delayed beyond six to eight hours (Jones et al., 2022). France, 97 departments, people aged 75 or over: in-hospital mortality 15.7% against 11.1% after a night in the department, aRR 1.39 (Roussel et al., 2023). Time to the geriatric ward: death or loss of function at six months rose from 42% to 71% across quarters (Bannelier et al., 2025). Against: a meta-analysis of 52 studies found harm only beyond 24 hours in patients going to intensive care (Lauque et al., 2022). Upgraded for dose-response, downgraded for inconsistency; sicker patients wait longer. · PMID 35042695 · PMID 37930696 · PMID 40229528 · PMID 27160482 · PMID 36614835

4. Delirium goes with later death and care-home admission even after allowing for age, illness and prior dementia, and probably with later dementia.

   Harm from exposureLow confidenceDeath, care home, dementiaMeta-analysis of high-quality studies adjusted for age, sex, comorbidity or severity and baseline dementia: death HR 1.95, institutionalisation OR 2.41; dementia OR 12.5 but from two studies with a confidence interval from 1.9 to 84 (Witlox et al., 2010). Death and care home sit at the upper end of low; dementia at low. · PMID 20664045

5. Staying longer than necessary while waiting to leave is linked to death, infection, low mood and lost mobility.

   Harm from exposureVery low confidenceDeath, infection, mobilityMixed-methods review of 35 studies in OECD countries; seven of its ten quantitative studies were of low or moderate quality and the authors urge caution (Rojas-García et al., 2017). People whose discharge is delayed are frailer and more dependent to begin with, which is often why they wait. · PMID 28898930

What the trials settle

1. Comprehensive geriatric assessment started on admission makes it more likely an older person is alive and in their own home at 3–12 months, and less likely they move into a care home. It does not change mortality or dependence, and may cost slightly more.

   What worksHigh confidenceAlive and at homeCochrane review, 29 trials, 13,766 people (Ellis et al., 2017) · PMID 28898390

2. Acute geriatric units that take patients in the acute phase reduce functional decline at discharge and make living at home afterwards more likely, without changing mortality. They also go with fewer falls and less delirium.

   What worksModerate confidenceFunction and homeTwo meta-analyses, 11 and 13 studies, mixed designs (Baztán et al., 2009; Fox et al., 2012) · PMID 19164393 · PMID 23176020

3. Multicomponent non-drug delirium prevention cuts new delirium by about two fifths, from 18% to 11%. No effect on mortality has been shown.

   What worksModerate confidenceDeliriumCochrane review, 22 trials (Burton et al., 2021) · PMID 34826144

4. Structured exercise from the first days of an acute stay, in people with a mean age of 87, reverses the loss of function that usual care produces, with no adverse effects.

   What worksModerate confidenceFunctionOne randomised trial, single centre, 370 people (Martínez-Velilla et al., 2019) · PMID 30419096

5. After hip fracture, joint care by orthopaedic surgeons and geriatricians lowers in-hospital and one-year mortality and delirium.

   What worksModerate confidenceSurvivalTwo meta-analyses, 37 and 18 studies (Van Heghe et al., 2021; Grigoryan et al., 2014) · PMID 34591127 · PMID 23912859

6. For selected older people referred for admission, hospital at home is a safe alternative to a ward: no difference in mortality or readmission, about half the chance of living in residential care six months later, and probably lower cost.

   What worksModerate confidenceHomeCochrane review, 20 trials (Edgar et al., 2024) · PMID 38438116

7. An individualised discharge plan slightly shortens the stay and reduces readmission. It does not change health status or mortality.

   What worksModerate confidenceProcessCochrane review, 33 trials (Gonçalves-Bradley et al., 2022) · PMID 35199849

How patients and carers experience it, and what they say matters

1. Frail older people who wait in the emergency department for a bed describe physical and psychological distress, and feeling disregarded, defenceless and resigned. Older people attending describe needs for comfort, help, food and drink. Confidence in the nurses, and a trusting relationship with them, sustains hope.

   Lived experienceModerate confidenceWaitingTwo small qualitative studies of frail patients boarding in the department, a study of frail patients' needs there, a study of people aged 75 or over after a visit, and a review of pre-2010 studies (Iozzo et al., 2025; Bjurbo et al., 2025; Venema et al., 2023; Cetin-Sahin et al., 2019; Hoon et al., 2012). No source found gives patients' own account of corridor care. · PMID 40429551 · PMID 40898611 · PMID 36773515 · PMID 32821794 · PMID 27820308

2. On acute wards, older people subordinate their needs to those of staff and other patients, holding back information and requests for help. They take technical care for granted and judge care mainly on how staff relate to them, including whether they are seen as a person.

   Lived experienceModerate confidenceBeing seenTwo meta-ethnographies, of 63 and 43 sources, from an overlapping team, so not fully independent; older people in general rather than frail people specifically (Bridges et al., 2019; Bridges et al., 2009). One consequence for measurement: people who hold back requests will under-report unmet need in complaints and satisfaction surveys. · PMID 31862528 · PMID 19854441

3. Most older inpatients think walking in hospital is appropriate, but few expect to and few have discussed it with a doctor. Symptoms, fear of falling, lines and drains, their own unwillingness, and a ward environment that discourages activity keep them in bed.

   Lived experienceModerate confidenceMovingA survey of attitudes to exercise in hospital, a review of 48 studies of barriers, and two small qualitative studies, in the United States, the Netherlands and Singapore; none specific to frailty (So and Pierluissi, 2012; Dijkstra et al., 2022; van Dijk-Huisman et al., 2022; Lim et al., 2020). Patients' own unwillingness is a real barrier, so lying in bed cannot be put down only to the ward. · PMID 22429118 · PMID 35908056 · PMID 35399054 · PMID 32268947

4. At discharge, what matters to older people is getting their life back at home. They experience leaving as negotiating for independence while professionals orient to their dependence, and poor preparation and information recur.

   Lived experienceModerate confidenceGoing homeA meta-synthesis of patients with multiple conditions, a study of transitional care, a qualitative study after emergency-department visits, a study of delayed transfers and a small Norwegian study (Shamim et al., 2025; Allen et al., 2016; Condon et al., 2024; Smith et al., 2022; Nilsen et al., 2021). · PMID 41017050 · PMID 27469975 · PMID 38448831 · PMID 36193616 · PMID 34117673

5. Older people and their carers are less likely than clinicians to see goal-setting in geriatric assessment as shared, and decisions about escalating treatment appear led by clinicians. In one acute medical unit, patients felt monitored and observed rather than actively treated.

   Lived experienceLow confidenceShared decisionsA review of 91 patients' and carers' accounts of inpatient assessment, a small qualitative study in one unit, and a review graded by its authors with GRADE-CERQual that notes patients' views are seldom studied (O'Shaughnessy et al., 2023; Darby et al., 2016; Warner et al., 2023). Thin patient data. · PMID 38066435 · PMID 26801585 · PMID 37588625

6. Family carers of older inpatients, mostly of people with dementia or delirium, want to be recognised, listened to, informed and supported. Poor communication feeds their worry, and some compensate or find information for themselves.

   Lived experienceModerate confidenceCarersA review graded with GRADE-CERQual (29 studies), a meta-synthesis on dementia, a rapid review its authors call weak, a scoping review and a qualitative study (Lee-Steere et al., 2024; Burgstaller et al., 2018; Mockford, 2015; Liebzeit et al., 2023; Mahmoud et al., 2024). Some samples include intensive care and younger adults. · PMID 38284503 · PMID 28639361 · PMID 29355178 · PMID 37598425 · PMID 39578718

7. Delirium is remembered as distressing, and anxiety, flashbacks and shame about behaviour can persist after discharge. Relatives want information, communication and a part in care. Clinical measures treat delirium as over once it resolves.

   Lived experienceModerate confidenceAfter deliriumReviews of patients' and relatives' experience (Kuusisto-Gussmann et al., 2021; Steffen et al., 2025; Bélanger and Ducharme, 2011; Meyer et al., 2023; Shrestha and Fick, 2020). Downgraded for relevance: many samples are from intensive care or mixed ages, not frail older people. · PMID 33991125 · PMID 41122003 · PMID 21999421 · PMID 37722790 · PMID 32374518

8. Asked what matters in acute care, older people with frailty in one study described autonomy (information, control, feeling secure) and function. Research measuring outcomes after emergency care mostly measures physical function; security and self-determination are rarely assessed.

   Lived experienceLow confidenceWhat mattersOne interview study of 22 older people with frailty, and its follow-on outcome measure (van Oppen et al., 2022; 2024); other studies of what older emergency patients want rank symptom relief and a diagnosis first, or abilities and emotions (Chera et al., 2024; Griese et al., 2025); a review of outcome measures after emergency care, in which 84% of studies measured physical function (Gettel et al., 2024). No core outcome set for acute frailty care was found. · PMID 35216550 · PMID 39414637 · PMID 39514495 · PMID 41251602 · PMID 38366698

What the NHS standards ask, and whether anyone has tested them

1. FRAIL strategy NHS England · 2024

   Standing

   Strategy with a self-assessment tool

   What it asks

   Clinical Frailty Scale, NEWS2 and the 4AT within 30 minutes of arrival for people aged 65 and over; a multidisciplinary acute frailty team at least 10 hours a day (ideally 12), seven days a week; comprehensive geriatric assessment for everyone scoring 6 or more on the frailty scale, a threshold agreed with the British Geriatrics Society; advance care planning support at 7 or more; no more than 8 hours in the frailty service; a discharge summary within 24 hours.

   What is measured

   Self-assessment only. A 2023/24 incentive payment (CQUIN) rewarded frailty identification in emergency departments.

   Does meeting it improve outcomes?

   Its core, geriatric assessment, is backed by trials. Its time limits are consensus; no study tests them.

2. The Model Acute Pathway NHS England with the Royal College of Physicians, Society for Acute Medicine and British Geriatrics Society · 2026

   Standing

   Standards; trusts self-assess and report to their boards

   What it asks

   Frailty Scale and 4AT within 30 minutes of arrival; for people scoring 5 or more, a senior decision-maker skilled in frailty within 1 hour of referral (before 10am if overnight) and a consultant-level frailty expert within 6 hours by day or 14 hours overnight.

   What is measured

   Nationally, only the share of patients discharged within 72 hours. Everything else is local.

   Does meeting it improve outcomes?

   Too recent to have been evaluated.

3. Best practice guide for NHS frailty pathways NHS England · 2026

   Standing

   Guidance; its metrics are suggested, and in its own words 'not nationally mandated or collected'

   What it asks

   At least 70 hours a week of acute frailty service; geriatric assessment started within 1 hour of admission for people scoring 5 or more; a further assessment at home within 24 to 48 hours of discharge; a full assessment bundle (4AT, medication review, falls assessment, personalised care plan).

   What is measured

   Suggested only. The one outcome-like measure is the share of people whose frailty score is the same or lower six months later. It asks for patient-reported outcome and experience measures, and says national assurance metrics will come with the modern service framework for frailty and dementia.

   Does meeting it improve outcomes?

   No evaluation. Its case studies report large gains, none of them independently evaluated.

4. Same-day acute frailty services NHS Improvement and NHS England · 2019

   Standing

   National ambition: every hospital with a major emergency department to run an acute frailty service at least 70 hours a week by December 2019

   What it asks

   Geriatric assessment within 30 minutes of an ambulance arrival; screening for geriatric syndromes within 1 hour for people scoring 7 or more; a multidisciplinary team 10 hours a day, seven days a week; out-of-hours admissions seen by noon the next day.

   What is measured

   Local.

   Does meeting it improve outcomes?

   It cites a medication-review trial for cutting adverse drug reactions, which the trial supports (21% to 12%), and for shortening stays by three days, which that trial did not find: median stay was 8 days in both arms (O'Connor et al., 2016).

   PMID 27365262

5. Co-ordinated frailty care for better outcomes, and the SAMIT 75+ dashboard Getting It Right First Time (GIRFT) · 2025

   Standing

   Guidance and a national benchmarking dashboard

   What it asks

   Frailty Scale as early as possible in any setting; a locally agreed threshold for geriatric assessment (commonly a score above 5 or 6); a single point of access seven days a week; the Frail-tED review of alternatives to the emergency department.

   What is measured

   SAMIT 75+, refreshed monthly, ranks hospitals on admission demand, early discharge, beds used by long stays, and readmission or repeat admission.

   Does meeting it improve outcomes?

   These are measures of demand and flow. None measures whether people are alive, at home or able to do what matters to them.

6. Silver Book II British Geriatrics Society · 2021

   Standing

   Good practice guide for the first 72 hours of urgent care

   What it asks

   Assessment as a continuing process, not a single event; unit-based geriatric services in preference to visiting liaison teams; a frailty champion on the acute medical unit.

   What is measured

   Not a measurement framework.

   Does meeting it improve outcomes?

   Its preference for units matches the trials of acute geriatric units. It says plainly that the evidence behind geriatric emergency care recommendations remains a challenge.

7. Society for Acute Medicine Benchmarking Audit (SAMBA) Society for Acute Medicine · 2022

   Standing

   Annual one-day national audit

   What it asks

   Benchmarks time to initial assessment (4 hours) and to consultant review.

   What is measured

   In 2022, 90% of 152 hospitals said they used a validated frailty tool, but only 45% of 3,604 patients aged 70 and over were assessed with one. Frail patients were less likely to be seen within 4 hours or reviewed by a consultant in time. Assessment by an acute frailty service went with more same-day discharge (odds ratio 1.55).

   Does meeting it improve outcomes?

   A process audit. The gap it measures, between policy on paper and assessment at the bedside, is the finding.

   PMID 39014306 · PMID 38192597

8. National Hip Fracture Database and Best Practice Tariff Royal College of Physicians · 2007

   Standing

   National audit since 2007; tariff paid for meeting care standards since 2010

   What it asks

   Prompt surgery, orthogeriatric review, delirium and bone-health assessment, mobilisation by the day after surgery.

   What is measured

   Every hip fracture in England, Wales and Northern Ireland, with 30-day mortality.

   Does meeting it improve outcomes?

   The one English frailty-related indicator set tested against outcomes, and the tests disagree. Against Scotland, 30-day mortality fell by 1.7 points more after the tariff; comparing English hospitals paid the tariff with those not, surgery within 48 hours rose 5.9 points but the 0.6-point fall in mortality was not significant. More orthogeriatrician time went with lower mortality (3.4% relative, per 2.5 extra hours per patient).

   PMID 31362544 · PMID 32052455 · PMID 27915229 · PMID 26172938

What other systems show, judged on outcomes

1. Germany. In claims data from 121 certified centres, hospitals where the geriatrician worked on the surgical ward had a 28% lower 30-day death hazard than those with a low-frequency consult service, at similar cost. The signal is depth of involvement, not certification. ISAR predicts poorly (AUC 0.62–0.66) and, at the usual cut-off, flags 82% of attenders aged 75 and over. No controlled evaluation of early geriatric rehabilitation for medical patients was found.

   What worksLow confidenceGeriatricians integrated into hip-fracture care in certified orthogeriatric centres; a tariff for early geriatric rehabilitation in hospital; ISAR screening in emergency departments · PMID 42477569 · PMID 24969966

2. Japan. The Kihon Checklist predicts dependency well (frail HR 4.77), which is prediction rather than benefit. People who joined community salons or going-out programmes had about half the rate of new disability (HR about 0.49), in observational studies where people chose to take part. No verified study shows the insurance system itself improves function, hospital use or place of death.

   What worksLow confidenceUniversal long-term care insurance (2000); municipal community-based integrated care; the Kihon Checklist frailty screen; resident-run community salons · PMID 28479274 · PMID 25888596 · PMID 38569560

3. Denmark and the Nordic countries. Reablement is not shown to improve function or quality of life: the Cochrane review is very low certainty, and a 2025 Swedish trial and a 2026 meta-analysis of 27 trials found no clear effect. Preventive home visits work only in their stronger forms: more than nine visits cut nursing-home admission (RR 0.66), and a clinical examination improved function, while pooled across all trials there was no significant effect. Under Senior Alert, fewer than half of planned falls and nutrition interventions were carried out.

   What worksVery low confidenceReablement required by law before home help (Denmark); preventive home visits for everyone at home aged 75 and over (Denmark); Senior Alert prevention register (Sweden) · PMID 27726122 · PMID 40211212 · PMID 42380867 · PMID 11866651 · PMID 18375879 · PMID 34286078

4. Norway. Municipal acute beds reduced hospital admissions in a nationwide natural experiment, most among people aged 80 and over. After the reform, fewer older people died in hospital (adjusted rate ratio 0.87), but the deaths moved to nursing homes: deaths at home did not change.

   What worksLow confidenceThe 2012 Coordination Reform moved care to municipalities, with municipal acute beds and charges for delayed discharge · PMID 37151780 · PMID 33147248

Not settled, and held apart on purpose

1. Does assessment at the emergency-department front door add to what assessment on admission already does?

   What worksLow confidenceTrials disagree. A Finnish trial of geriatric assessment in the emergency department found no difference in hospital days over a year; an Irish trial of an emergency-department therapy and social work team cut admissions and improved function at six months; a review of 15 studies found a small effect on function only (Alakare et al., 2021; Cassarino et al., 2021; Hughes et al., 2019). · PMID 34215193 · PMID 34319971 · PMID 30875098

2. Does assessment before the crisis, at home or in the community, change what happens?

   What worksLow confidenceLittle or no effect on death or care-home admission; possibly fewer unplanned admissions; small gains in function in home-based programmes (Briggs et al., 2022; Hayes et al., 2025). · PMID 35521829 · PMID 39964054

3. Is giving the same specialist input sooner better than giving it a little later?

   What worksVery low confidenceNo frailty trial compares the timing of the same input. The hip-fracture evidence for 'early' assessment is mostly before-and-after studies with 'early' undefined, and a meta-analysis of five small trials of orthogeriatric care started before surgery found no significant effect (Mazarello Paes et al., 2025; Buecking et al., 2013). The one trial that does randomise timing is in advanced cancer: palliative care started early rather than three months later made no difference to quality of life, symptoms or mood, though one-year survival was better (ENABLE III; Bakitas et al., 2015). Where timing has been randomised outside frailty, the results lean to no gain: hip surgery at a median of six hours rather than 24 did not reduce deaths or complications (HIP ATTACK, 2020), and very early, higher-dose mobilisation after stroke made a good recovery less likely (AVERT, 2015). Very low means not shown, not wrong: avoiding long delays is better supported than racing to be first. · PMID 40449538 · PMID 23667392 · PMID 25800768 · PMID 32050090 · PMID 25892679