Trial finds structured rehab preserves function in terminal cancer

Maintaining activities of daily living is important for terminally ill cancer patients. In a randomized trial across 19 Japanese inpatient hospice and palliative care units, an Osaka Metropolitan University-led team found that a structured rehabilitation program preserved these everyday abilities better than usual unstructured care.

The study enrolled 130 terminally ill cancer patients between July 8, 2019, and Feb. 20, 2024. Participants had an Eastern Cooperative Oncology Group performance status of 2–3, a life expectancy of at least three weeks and no severe symptoms. They were assigned to a three-week structured program or to the rehabilitation normally provided at their unit.

The researchers measured activities of daily living with the modified Barthel Index, or mBI, which tracks practical abilities such as personal care and mobility. The average score changed by -1.31 in the structured-rehabilitation group, compared with -15.51 in the control group. That produced a between-group difference of 14.21, exceeding the minimally clinically important difference of 9.25. Patient-reported physical functioning was also significantly higher in the structured group.

The program’s value lies in making rehabilitation more consistent without treating every patient the same. It systematically organized recommended behaviors for therapists, while allowing care to reflect each person’s condition and goals. The researchers reported no serious harms, and first author Nanako Nishiyama said the approach could reduce variations in practice between institutions.

So what changes in practice? For patients in the final stage of cancer, rehabilitation need not be framed only as recovery or long-term strength building. The researchers say a practical guide could help therapists tailor rehabilitation to each patient's condition and goals, while making practices more consistent across institutions. The evidence is encouraging but bounded: the primary analysis included 77 participants with complete data, and the trial lasted three weeks.