Evidence Explorer · MS-N277 · Supportive physical care · Appraised November 8, 2026
A 19-site Japanese randomized trial found that a structured three-week rehabilitation program preserved activities of daily living better than usual unstructured rehabilitation in selected inpatients with terminal cancer. Attrition was substantial: 130 were randomized but only 77 had complete primary-outcome data.
Evidence at a glance
- Population: 130 inpatients with terminal cancer, ECOG performance status 2–3, life expectancy at least three weeks and no severe symptoms.
- Intervention: structured rehabilitation for three weeks.
- Comparator: usual unstructured rehabilitation.
- Primary result: modified Barthel Index changed −1.31 versus −15.51; between-group difference 14.21 points (95% CI 1.77–26.64), exceeding the stated 9.25-point clinically important difference.
- Evidence label: positive multicenter evidence with major missing-data and selection limits.
What was studied
The intervention preserved patient-reported physical function and no serious harms were reported. It was delivered in inpatient hospice or palliative-care units, not as unsupervised exercise.
Only 77 of 130 randomized participants had complete day-22 data. In a population with short prognosis, missing outcomes are understandable but can bias estimates if decline or death differs by group.
What this evidence cannot establish
The eligibility criteria excluded severe symptoms and required an estimated life expectancy of at least three weeks, so findings may not generalize to people who are unstable or actively dying. The study addresses function and comfort, not survival or cancer control.
Safety and practical context
Goals should center on comfort, transfers, self-care and energy conservation. Stop or adapt activity for distress, falls, unstable fractures, severe breathlessness, uncontrolled pain, delirium, bleeding, acute infection or rapid decline. Rehabilitation should never feel compulsory near the end of life.
Questions for the care team
- Which personally meaningful activity is the program trying to preserve?
- How will intensity change if symptoms or prognosis change?
- What is the plan if participating becomes burdensome rather than helpful?
Sources and identifiers
- Multicenter randomized trial — PMID 42304645
Funding and conflicts: The study lists Japan Agency for Medical Research and Development grants. Several authors reported outside honoraria; the PubMed record states these were outside the submitted work.
Source access and dates: Original PubMed records and available full texts were checked November 8, 2026. Last literature search November 8, 2026.
Educational evidence summary, not individualized medical advice. Supportive-care outcomes do not demonstrate tumor response, recurrence prevention or longer survival.
