Counting patients' time: days at home can be built, and a length-of-stay target would have misjudged the week's best frailty result

A quiet week for new clinical evidence, and a useful one for a hard question: how is patients’ time counted? Twelve items across the five domains cleared review, almost all single-centre or from outside the UK and none practice-changing, with Palliative Care and OPAT thin. Alongside them, a systematic review of measurement, an English programme evaluation and two cohorts from Ontario and the United States were read for that single question. The best frailty-specific result of the week improved survival and reduced delirium while lengthening hospital stay, which a length-of-stay target alone would have scored as a failure. No new national clinical development was identified in the five domains.

Time away from healthcare is the only countable stand-in for patients’ time found, and frailty and age lower it

The measurement question has a first usable vocabulary. A systematic review of 56 studies of how time burden is defined in cancer care found that most measures count direct healthcare time, that measures of time away from healthcare such as days alive and out of hospital or days at home appear in 20 of the 56 studies (35.7%), and that no study used a patient-reported definition Synthesis (Wee et al., PMID 42809115). It is oncology-specific, and the authors’ proposed three objective domains transfer to frailty and community pathways more readily than the evidence behind them does.

What a days-at-home measure looks like in practice comes from Ontario. A population-based retrospective cohort study of 693 adults with acute traumatic spinal cord injury admitted to trauma centres between 2009 and 2023 linked administrative health records to test days alive and at home at 180 days as a patient-centred outcome Primary Study (Lozano et al., PMID 42770972). The median was 63 days (interquartile range 0–118), and it was lower with older age, greater injury severity, complete injury and frailty. It correlated with motor function at rehabilitation discharge (ρ 0.69), with days at home from 180 days to one year (ρ 0.61) and from one to two years (ρ 0.47), and inversely with one-year healthcare expenditure (ρ −0.41); the minimum important difference was about 23 days. This validates an outcome measure; it does not evaluate a service, and the population is a trauma one, with no frailty-pathway focus.

A third study shows the same idea applied to frailty, in a very different setting. In a retrospective cohort of 260 patients aged 65 and over receiving commercial CAR-T therapy for myeloma or B-cell lymphoma at one US centre, an automated electronic frailty index separated robust or pre-frail, frail and very frail patients Primary Study (Enns et al., PMID 42826706). Frail and very frail patients had higher odds of grade 3 or worse toxicity (odds ratios 3.28 and 4.84) and fewer days alive and out of hospital in the first 30 days (β −2.37 and −3.79); only very frail patients had worse survival. The therapy is highly specialised and the cohort is from a single centre, so it shows that a score built from health records can track time out of hospital, not that it would do so across a general frailty caseload.

Taken together these three extend the published pattern on measurement fidelity from handover to outcome measurement: days at home can be built and validated from linked records, yet no study defined time in patient-reported terms. That is a loose extension rather than corroboration, since all three come from oncology or trauma and none from a UK frailty pathway. It also continues a line from last week’s synthesis, which found that frailty scores predict death but not readmission and that social disadvantage marks who is held up: identification and measurement are advancing faster than evidence that either changes what happens next.

The best frailty-specific result of the week would have been scored a failure on length of stay

A quality-improvement evaluation of a geriatric co-management service in oncologic surgery at one US health system, launched in 2021, analysed 4,359 cases between July 2018 and August 2024 and compared 309 propensity-matched pairs Primary Study (Wolfgang et al., PMID 42832427). Co-management was associated with lower odds of hospice or in-hospital mortality (odds ratio 0.27, 95% confidence interval 0.12–0.61) and of delirium on the Confusion Assessment Method (0.29, 0.11–0.73); ICU admission was lower but not significantly (18 against 25, p=0.069). Length of stay was longer (8.2 against 6.2 days, p=0.035). Patients were less likely to be discharged home (odds ratio 0.51) and more likely to go to a skilled nursing facility (2.25) or inpatient rehabilitation (8.52, with a very wide interval of 1.04–69.6).

The limits are real. It is a single-system, observational evaluation matched on measured covariates only, in a US post-acute system with no direct NHS equivalent, and the outcomes carried here are clinical and administrative, so how patients and families experienced the longer stay and the move away from home is not known. For the published pattern that earlier specialist input is associated with better outcomes this qualifies the claim without refuting it: it stands for mortality and delirium, and should not be read as a claim about time or destination. The open question is whether the longer stay and lower home-discharge rate reflect a more appropriate destination, or a gap in rehabilitation and community capacity that a model with strong intermediate care could avoid. Either way, a service judged on length of stay or home discharge alone would have been marked down for the very intervention that cut deaths and delirium.

Integration changed one measured flow but not admissions, and counted cost rather than time at home

The English reference point is older and sobering. RAND Europe’s 2019 evaluation of the Southwark and Lambeth Integrated Care Older People’s Programme reported evidence of reduced A&E attendance in the programme’s fourth year against what would have been expected without it, but no evidence of reduced emergency admissions Audit/Data (RAND Europe, Research Report RR2231). The programme cost £149 per resident aged 65 and over against hospital-cost savings of £86. It is carried here from the evaluation’s published summary and not its full text, it was published in 2019, years before the rest of this week’s evidence, and it describes a London programme with its own population profile. It matters because it shows whole-system integration moving one flow and not another, and because it measures cost and attendance, not how much of their time older residents spent at home.

Home-delivered rehabilitation looks feasible, but gain in function for frail older people is not shown

Two meta-analyses published this week bear on home and remote delivery, and both come with heavy qualifications. Nine randomised trials in 632 adults with cancer found home-based exercise rehabilitation improved fitness (standardised mean difference 0.42) and quality of life (0.63) with adherence above 75%, though heterogeneity was high (I² 62% and 84%) and the population was people with cancer, not older frail adults Meta-analysis (Winnige et al., PMID 42811812). A network meta-analysis of 92 randomised trials in chronic musculoskeletal pain found small to moderate average gains from digital interventions, certainty predominantly low or very low, and no basis for calling exercise-based virtual reality or telerehabilitation superior Meta-analysis (Wang et al., PMID 42842940). Different conditions from the sarcopenia network meta-analysis behind the published pattern on remote exercise delivery, so this bounds how far “remote delivery is optimal” can be generalised beyond sarcopenia, without refuting it.

The patient’s own account points the same way. A qualitative analysis of free-text responses from 101 homebound older adults (mean age 74) in the 12-week Strength on Wheels trial of home-delivered meals plus a home-based exercise programme, a trial which had found no difference in physical measures but improved loneliness in the exercise group, reported that participants enjoyed the meals and the exercise materials Primary Study (Cruz et al., PMID 42847733). They also reported ongoing pain and fear of falling, had mixed experiences with a commercial activity tracker, and named social engagement, physical therapy, transport and home modification as the supports they needed. It is a qualitative study of one trial’s participants in a different system, but it is consistent with the HERO trial recorded in the published pattern on earlier specialist input: a multicentre randomised trial of 740 older people with frailty in which 24 weeks of extended home-based exercise gave no 12-month benefit in physical health-related quality of life (adjusted difference −0.22, 95% confidence interval −1.47 to 1.03) Primary Study (Prescott et al., PMID 41557460; earlier signal). That is weak confirmation, since the populations and designs differ and Cruz is qualitative. Home delivery widens access; it has not been shown to improve function for frail older people, and the people in the Strength on Wheels trial asked for therapy input and environmental change, not a printed programme.

Care-home and community psychosocial provision: novelty and prescribing are not shown to add value

Immersive virtual reality was no better than a low-technology group activity for dementia symptoms in a multicentre French randomised trial of 56 nursing-home residents, which compared group virtual reality with non-digital cognitive stimulation, twice a week for six weeks at 45 minutes each time Primary Study (Giroux et al., PMID 42837535). There was no significant between-group difference in behavioural and psychological symptoms (Neuropsychiatric Inventory) or in overall well-being, immediate well-being rose in both arms, social interaction was more frequent in the active control, and adverse events were few. It is small and formative-stage, so it is a cautionary negative and not a definitive one.

In the community, the UK-led systematic review of green health prescribing of 21 studies found that many reported improvement in at least one outcome but that significant between-group effects were uncommon and certainty was low or very low Synthesis (Butt et al., PMID 42823701). Neither study supports a new psychological-therapy offer. Both counsel against assuming that novelty, or a policy-favoured prescribing route, adds value before it has been tested.

At the care-home end of palliative care, a Norwegian questionnaire study of 77 nursing-home physicians, with a response rate of 43.5% between November 2022 and December 2023, found that they rated the importance of palliative care higher than the quality delivered Primary Study (Sandsdalen et al., PMID 42834000). The widest gaps were in symptom relief (pain, depression, anxiety), existential care, individualised care, cooperation and continuity. The authors caution that the sample is small. It extends the published pattern that community settings are structurally underequipped into the care-home setting from the clinician side, but adds no weight, being non-UK with a low response rate. At the acute-ward end, a single US academic centre’s quality-improvement project in which geriatric and palliative clinical nurse specialists built a comfort-care initiative reported 80–100% compliance on individualised care plans, pain assessment and timely end-of-life orders on the pilot geriatrics unit Audit/Data (Meyer and Ladak, PMID 42842231). It reports no patient-level outcomes and cannot be read as evidence of effect on place of death.

Smaller items upstream and at the edges

Two frailty items sit upstream of the acute pathway. A narrative review by Italian geriatric and orthogeriatric authors argues that comprehensive geriatric assessment on admission is pivotal for detecting frailty after a fall or fragility fracture, and that Fracture Liaison Services could detect frailty and reinforce intrinsic capacity from the acute setting to home Synthesis (Ruggiero et al., DOI 10.1016/j.arr.2026.103394). It is a framing paper with no pooled data or new outcome evidence. A prospective US cohort of 2,554 men who were robust or pre-frail at baseline (mean age 71.6), reassessed about seven years later, found that after full adjustment only activity-limiting back pain was associated with progression from pre-frail to frail (odds ratio 1.40, 1.03–1.90) Primary Study (McNaughton et al., PMID 42844230). It is an association in men only, using phenotype-based frailty, so it suggests activity-limiting pain as a target for prevention without showing that treating it changes the outcome.

In OPAT, nothing new on outcomes, selection or stewardship emerged. Hospital-pharmacy protocol work in Madrid standardised the preparation of 16 antimicrobials for paediatric hospital-at-home, with fixed diluent volumes, defined in-use times and a deliberately conservative microbiological approach because home refrigeration cannot be controlled Synthesis (Merino-Pardo et al., PMID 42823275); it has no patient outcome data and is paediatric, but it states plainly that home storage conditions set the limits on in-use time, which matters for adult delivery to dispersed addresses. A first-in-human, industry-sponsored study of once-daily intravenous ertapenem with zidebactam in 54 healthy volunteers, dosed for seven days, found it well tolerated with predictable pharmacokinetics Primary Study (Lodise et al., PMID 42809435); with no efficacy data, it is horizon scanning for a possible OPAT agent and nothing to act on.

What the week says about whether patients’ time is being well spent

The honest answer is that the week’s evidence mostly cannot say, which is itself the finding. The one place a patient’s time was counted directly is days at home, and there it was lower for people who were older or frail. The best frailty-specific intervention traded days in hospital for lower mortality and delirium, and shifted patients from home towards skilled nursing and rehabilitation, so whether that trade was worth it depends on what patients and families would choose, which was not measured. Whole-system integration in London changed A&E attendance and not admissions. Home-delivered rehabilitation widened access, but the homebound older adults asked for therapy, transport and home modification, and the trials do not show their function improving. The home-delivery evidence summarised here reports nothing on the time of unpaid carers, who absorb the work when care moves home. No comparator found this week, as last week, counts patient time in value terms rather than bed-days or days at home.

National direction is waiting on its own timetable, and the standing record is narrower than its wording

The Department of Health and Social Care committed in May 2026 to publishing an interim product for the dementia and frailty framework in September and the full framework by the end of the calendar year Guidance (Department of Health and Social Care, 22 May 2026). A search on 9 October found no source reporting that the interim product has appeared; an absence in a search is not confirmation that it has not been published, but it is now past its own marker. The consultation on the allied health professions framework for 2027 to 2032, which the September edition noted was due to close on 18 September, has published no outcome that was found, and a professional body’s notice of 11 August says the final framework is due to be published in 2027 Guidance (Society of Radiographers). Nothing this week suggests that a national commissioning specification requiring validated frailty or outcome measurement for virtual wards has appeared.

Nothing this week contradicts a standing published pattern. That is a statement about a thin, largely non-UK week and not evidence that no pattern is overstated: the claim that remote delivery of exercise is optimal and the breadth of the claim for earlier specialist input are now the two published entries resting on narrower evidence than their wording suggests.

Open questions

• Can a days-at-home measure be built from readily available acute, community and social-care activity data, and could it be stratified by deprivation?
• Do the longer stay and lower home-discharge rate after geriatric co-management reflect a more appropriate discharge destination, or a rehabilitation and community capacity gap that a model with strong intermediate care could avoid?
• Will the dementia and frailty framework attach measurement or equity requirements to home-based care models? The interim product is now past its September marker.