Early mobilisation of ventilated patients is recommended in guidelines, but the recommendation rests on modest evidence and the effective dose has never been clear. EVER tested a structured six-step programme across five tertiary hospitals in South Korea, randomising 169 adults with sepsis or respiratory failure expected to need invasive ventilation for at least 48 hours to the programme or usual care, with the Functional Status Score for the ICU at ICU discharge as the primary outcome and follow-up to 12 months.
The intervention was genuinely delivered — this is not a failed implementation. Mobilisation began at a median 30.0 hours against 45.9, patients had a median 11 sessions against 4, and accumulated 330 minutes of mobilisation against 120. Nearly three times the dose, started 16 hours earlier.
And nothing changed. Functional status at ICU discharge was 23.6 (SD 11.8) against 22.2 (10.8), p=0.44. Every long-term outcome measured — quality of life on EQ-5D, physical and mental component scores on SF-36, post-traumatic symptoms, cognition on MoCA-BLIND, and post-intensive-care syndrome — improved in both arms over 12 months, similarly, with no significant differences.
One finding invites over-reading and should not be. Patients who actually achieved step 4 or higher — sitting to standing or beyond — had better functional status than matched usual-care patients, 30.6 against 23.3, p<0.01. That is not a treatment effect. Patients who can stand are patients who were well enough to stand, and the comparison is between people separated by their illness rather than by their allocation. It tells you who recovers, not what makes them recover.
The practical position: keep mobilising, because it is safe, humane, and prevents the harms of immobility that this trial was not designed to capture. But stop treating the intensity of the mobilisation protocol as the lever that determines functional recovery, and be sceptical of business cases built on that premise. With 169 patients this trial cannot exclude a modest benefit, and its usual-care arm was already mobilising by 46 hours — in a unit where mobilisation currently starts on day five, the comparison would be quite different.
- Continue early mobilisation for its safety and humanity, not on a promise of better functional outcomes.
- Do not build a business case for extra physiotherapy staffing on the expectation of improved 12-month function.
- Ignore the step-4 subgroup finding as evidence of effect; it reflects who was well enough to stand.
- Note the usual-care arm mobilised by 46 hours — the result may not apply where current practice is far slower.
- At 169 patients the trial cannot exclude a modest benefit; it excludes a large one.
The statistics, in plain English
The subgroup result is the trap here and worth naming precisely. Comparing patients who achieved a high mobilisation step with matched usual-care patients breaks randomisation, because reaching step 4 is itself an outcome determined by how sick the patient is. This is the same error as concluding that adherent patients do better because of the drug — adherence marks people who are already doing well. The main result is a clean null: a 1.4-point difference on a scale with standard deviations above 10, with p=0.44. What 169 patients cannot do is exclude a small real effect, so read this as 'a large benefit has been ruled out' rather than 'mobilisation does nothing'.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free