Rehabilitation should be carried out at a proper facility with a rehabilitation medicine department; the intensity must be set by therapists — more intensity does not necessarily mean better results.

The most common mistake is simply sending a patient home to lie down and hope for a natural recovery. Care delivered by a coordinated, multidisciplinary team helps many patients survive one year later and continue living at home. However, patients should not overdo it on their own. In a large multi-center clinical trial, early, frequent, and more intensive activity within 24 hours after a stroke actually led to poorer outcomes three months later. Thus, the amount and timing of exercise must be determined by the rehabilitation department and therapists.

Inpatient rehabilitation costs several hundred yuan per day; the out‑of‑pocket portion varies by region. Treatment is billed on a monthly basis. Daily exercises are required and can be quite monotonous.

A Cochrane systematic review included 29 trials with 5,902 participants, comparing organized inpatient stroke‑unit care with other forms of care. At the final follow‑up (median 1 year), the odds ratio for adverse outcomes — death, disability, or need for institutional care — was 0.77 (95% CI 0.69–0.87). The odds ratio for death alone was 0.76 (95% CI 0.66–0.88), and for death or dependence it was 0.75 (95% CI 0.66–0.85); evidence quality is moderate. In absolute terms, for every 100 patients receiving stroke‑unit care, an additional 2 survive, 6 continue living at home, and 6 regain the ability to care for themselves. Conversely, the AVERT trial randomly assigned 2,104 patients across 56 acute stroke units in five countries to either “standard care” or “standard care plus very early intensive activity.” After three months, 480 patients (46%) in the early‑activity group achieved a good outcome (modified Rankin Scale 0–2) versus 525 patients (50%) in the standard‑care group; the adjusted odds ratio was 0.73 (95% CI 0.59–0.90, P=0.004). Death rates were 88 versus 72 patients, with no statistically significant difference (odds ratio 1.34, 95% CI 0.93–1.93, P=0.113).

Caution: Both studies involve stroke patients; their findings cannot be directly applied to other disabling conditions such as spinal cord injury, amputation, or burns. The AVERT trial demonstrates that “the earlier and more intense the activity, the better” is not true; it does not mean that early activity is useless. In both trials, the majority of participants began walking within 24 hours; the difference lay in frequency and total volume of exercise. When selecting a facility, verify whether it has a rehabilitation medicine department and qualified physical and occupational therapists — do not rely solely on a “rehabilitation” sign. Information on referrals between primary and tertiary hospitals and related cost thresholds can be found in Section 24.1 (step‑by‑step referrals). Community‑based rehabilitation and assistive devices are discussed in Section 16.7 (contact the county disability‑affairs office).

Langhorne P, Ramachandra S; Stroke Unit Trialists' Collaboration (2020). Organised inpatient (stroke unit) care for stroke: network meta-analysis. Cochrane Database of Systematic Reviews, 4:CD000197. https://doi.org/10.1002/14651858.CD000197.pub4；AVERT Trial Collaboration group (2015). Efficacy and safety of very early mobilisation within 24 h of stroke onset (AVERT): a randomised controlled trial. Lancet, 386(9988):46-55. https://doi.org/10.1016/S0140-6736(15)60690-0