A Hong Kong trial found that an app linked to symptom reporting and nurse follow-up helped maintain quality of life for people with advanced cancer receiving palliative care at home. It tested a staffed service added to usual care, rather than a download that worked on its own.
That distinction matters for anyone considering a similar programme. The technology created a route for patients or caregivers to report problems; the intervention also included people who could respond.
A trial in community palliative care
The SUPPORT+ randomized trial, published in JAMA Network Open on 1 June 2026, enrolled 1,214 adults across six Hong Kong palliative care clinics. Participants had advanced solid cancers and were no longer receiving systemic anticancer treatment. The median age was 78.
Researchers assigned 590 participants to the additional digital service and 624 to usual care, with 18 weeks of follow-up. Usual care still included outpatient palliative care and community nursing when clinically indicated. The comparison was an added monitoring-and-response pathway, not care versus no care.
Caregivers were the app users for 821 participants across the trial. That makes family involvement central to the findings, rather than a minor workaround for people unfamiliar with phones.
What happened after a symptom report
Participants or caregivers submitted weekly assessments using the Integrated Palliative Outcome Scale. The intervention combined symptom-specific guidance with alerts to trained palliative care nurses. Participants could also request contact through a nurse-consult function.
The paper describes nurses reviewing alerts during office hours and contacting participants within one working day. Its linked Supplement 2 adds telephone contact when the digital platform could not be used, follow-up after a missed weekly diary and documentation of clinical decisions. This was not a continuously staffed emergency-response promise.
The University of Hong Kong’s October account explains the local service model. It describes the same trial, not a second independent evaluation. For the response standard and outcome definitions, the original paper and supplement provide the more precise account.
Quality of life and hospital use answer different questions
The primary outcome was change in health-related quality of life. The trial reported better maintenance of the EQ-5D-5L utility score with the added service. This measure combines aspects of mobility, self-care, everyday activities, pain and anxiety or depression; it is not a direct measure of tumour response.
Hospital use also favoured the intervention, but a widely repeated percentage needs care. In the paper’s week-18 change analysis, 63 of 367 intervention participants and 108 of 379 controls had worsening unplanned-hospitalization episode counts: 17.2% versus 28.5%. These are categories of change from baseline among participants in that analysis, not the proportion of all randomized patients who had any admission.
The supplement separately tabulates admission counts during follow-up. Those tables use different available-data denominators, reinforcing why the figures should not be collapsed into a single “admission rate”. Emergency-department visits did not differ significantly, and the study did not demonstrate a survival benefit.
The limits are also implementation questions
During follow-up, 343 participants died and 96 were lost to follow-up. Analyses used observed outcomes among participants alive at each assessment; the findings do not describe an unchanged cohort of all 1,214 people completing every measure. Patients and care nurses were not blinded, and the study could not separate the effects of reporting, guidance and nurse contact.
The trial received Hong Kong government research funding; the app was supported by the Bank of China (Hong Kong) Centenary Charity Programme. The paper identifies a SUPPORT+ patent holder and reports other author relationships in its disclosures. Those are source-study interests, separate from this publication’s funding.
For services elsewhere, the unresolved questions include staffing cost, caregiver workload and fit with existing escalation routes. The paper itself calls for further cost-effectiveness and caregiver-focused work. Unlike the Oxford hospital-at-home wearable study, this was an evaluation of palliative symptom support. Its useful lesson is that a reporting interface and the capacity to act on reports belong in the same service assessment.