Therapies and AHPs

The standing claim

Value in therapies is functional independence gained per episode. The best-evidenced lever is delivery by a coordinated multidisciplinary team; rehabilitation intensity matched to potential is the second. That it is decided early, before deconditioning sets a ceiling, is the working hypothesis this domain shares with acute frailty (D01): plausible, and consistent with shorter hospital stays doing no harm, but not tested as timing. The evidence is unusually clear about what works — and unusually honest about what is asserted but unproven.

Where practice stands

The firmest, most quantified standards in this domain are flow standards, not outcome standards: the Urgent Community Response two-hour standard (respond to ≥70% of appropriate two-hour referrals within two hours), and NICE NG74’s expectations — reablement offered as the first option, crisis response within two hours, bed-based intermediate care started within two days. NHS England’s Intermediate Care Framework organises this as therapy-led, “Home First” step-down. For stroke specifically, NICE NG236 (2023) raised the therapy-intensity expectation sharply — to at least three hours a day of multidisciplinary therapy, five days a week (up from the old 45-minute target). There are no comparable national standards for community rehabilitation outcomes or intensity.

What good looks like

The data-led exemplar is a model, not a country: coordinated, AHP-led stroke Early Supported Discharge (ESD). According to PubMed, the Cochrane review (Langhorne & Baylan, 2017 — 17 RCTs, n=2,422) found ESD reduced death or long-term dependency (OR 0.80, 95% CI 0.67–0.95), reduced death or institutional care (OR 0.75, 0.59–0.96) and shortened hospital stay by about six days, on moderate-grade evidence, with costs ranging from 23% lower to 15% higher across trials. Crucially, the benefit was greatest where a coordinated multidisciplinary team delivered it, and in patients with mild-to-moderate disability; without a coordinated team the results were inconclusive. The evidence specifies what “good” requires, which is exactly why it is the exemplar. It maps directly onto the Intermediate Care Framework’s therapy-led step-down. Goal-directed rehabilitation also reaches dementia: in the GREAT trial (475 people with early-stage dementia), home-based cognitive rehabilitation delivered mostly by occupational therapists produced large gains in the goals people chose, sustained at nine months and confirmed by carers, though with no difference in quality of life, mood or cognition and no cost-effectiveness on QALYs (Clare et al., 2019). The person’s own goal moved; the standard measures did not, which is itself a finding about what we measure.

The honest counter-case is reablement. It is the policy default (NICE NG74 makes it the first option), and it is internationally fashionable. But the evidence does not yet support the enthusiasm: a systematic review (Legg et al., 2015) found no studies meeting its inclusion criteria, which admitted non-randomised designs, and concluded reablement is “an ill-defined intervention… with no evidence to suggest it is effective” at either increasing independence or reducing care use. Some service-level evidence is encouraging — an Australian randomised trial of restorative home care (Lewin et al., 2014; 750 people) found fewer home-care hours, fewer emergency presentations and unplanned admissions, and lower combined costs over two years. One trial against one null review, ten years old: the evidence is thin rather than settled, and newer evidence is owed a look. So reablement is sound in design and policy, unproven in sustained effect — and treating its international reputation as proof is precisely the exchange-value error this publication guards against.

Through the lenses

The measure is use value — what the person can actually do afterwards — not the count of therapy contacts (the activity unit). McGilchrist’s whole-person attention is the clinical correlate: function is a property of the whole person in their context, not of any single discipline’s silo, which is why the coordinated team, not the sum of separate inputs, is what the ESD data rewards. The VAT lever is functional independence per episode; cross-domain, this is the active defence against the deconditioning that decides acute-frailty outcomes (D01) and the rehabilitation that lets palliative patients spend their time where they choose (D03).

Confidence and limits

The stroke-ESD evidence is moderate-grade and meta-analysed — but condition-specific (stroke), MDT-dependent, and concentrated in mild-to-moderate disability; its benefits weaken at one and five years. The reablement evidence is genuinely weak, and we state that plainly rather than dress policy enthusiasm as proof. The firmest national standards are operational-flow targets, not therapy-outcome standards — a structural gap. (NICE NG236’s “three hours” refers to total multidisciplinary therapy, not physiotherapy alone, and itself exceeds the directly-evidenced 1–2 hour physiotherapy band — a committee consensus, honestly labelled.)

What we are watching

Whether community rehabilitation acquires outcome standards rather than only flow targets; whether reablement finally generates RCT-grade evidence to match its policy status; and whether the NG236 therapy-intensity expectation is matched by the workforce to deliver it.