Fast take: A 6-week mixed-methods feasibility study in the Bidibidi Refugee Settlement (Uganda) found mPallCare (a VHT-facilitated mobile app and clinician dashboard for palliative symptom and outcome monitoring) to be feasible and acceptable but sensitive to translation, connectivity, and workflow constraints (JMIR Mhealth Uhealth, published 26 June 2026). Key metrics: n=32 patients; 84.9% (163/192) symptom-report completion; 59.4% (266/448) outcome-report completion; 67% combined (429/640). Read the original study at JMIR Mhealth Uhealth. Evidano is an AI-powered qualitative data analysis platform that accelerates thematic coding, cross-segment comparisons, and stakeholder-ready reports.
Key Takeaways
Evidano is an AI-powered qualitative data analysis platform that accelerates thematic coding, cross-segment comparisons, and stakeholder-ready reports.
The mPallCare pilot (n=32, 6 weeks) showed acceptable symptom-report completion (84.9%) and lower outcome-report completion (59.4%), with a combined completion of 67%, and feasibility that depended on translation, connectivity, and workflow supports.
- mPallCare paired a VHT-administered Android app and a clinician dashboard, collecting data in four languages and providing self-management guidance.
- Operational challenges included navigation and translation confusion, intermittent sync failures, device availability, and competing VHT duties that contributed to declining engagement.
- UX and product teams should prototype local phrasing, test rating scales with frontline VHTs, and consider adaptive reporting schedules to reduce data fatigue.
- Researchers should capture metadata (device, sync timestamps, VHT ID, zone) to analyze engagement drivers and validate dashboard actions with clinician notes.
Findings snapshot
| Metric | Value | Source / Note |
|---|---|---|
| Publication date | 26 June 2026 | JMIR Mhealth Uhealth |
| Study setting | Bidibidi Refugee Settlement, Uganda | Refugee + host integrated services |
| Sample | 32 people living with advanced cancer | Purposive convenience sample |
| Study length | 6 weeks (reporting Aug–Oct 2022) | Weekly symptom; outcome every 3 days |
| Symptom reports | 84.9% (163/192) | Target ≥65% for feasibility |
| Outcome reports | 59.4% (266/448) | Decline over time; connectivity & workload factors |
| Combined completion | 67% (429/640) | Meets conservative feasibility threshold |
| Top symptoms reported | Headache, muscle pain, dizziness | Table 2 in source |
What happened (plain English)
The pilot tested a VHT-administered Android app plus a clinician web dashboard to capture longitudinal symptom and outcome reports for people receiving palliative care.
The pilot combined a VHT-administered Android app (symptom checklist and APCA Palliative Outcome Scale) with a web dashboard for clinicians, VHTs visited patients to collect one symptom report per week and one outcome report every three days, data were captured in four languages, and self-management guidance was displayed after reports.
- Design: app plus dashboard, reports flagged (green, yellow, red) for clinician action.
- Operational inputs: 20 trained VHTs, local translations (English, French, Swahili, Juba Arabic), secure deidentified server.
- Challenges observed: initial navigation and translation confusion, intermittent sync errors, device availability and competing outreach duties, and modest decline in engagement across six weeks.
- Perceived benefits: improved communication, quicker clinical response, and increased patient empowerment and self-management use.
So what for researchers, UX & policy teams
For qualitative researchers
The study shows value in pairing structured longitudinal PROs with qualitative interviews to explain drop-offs and validate dashboard actions against clinician notes.
This study shows value in pairing structured longitudinal PROs with qualitative interviews (n=20 interviews, 10 patients, 10 clinical/VHT). Use mixed methods to explain drop-offs and validate dashboard actions against clinician notes.
Collect metadata (device, sync timestamps, VHT ID, zone) to enable cross-segment analysis of engagement and to model operational constraints.
For UX/product teams
Translation and local phrasing were critical and the study found numeric rating scales unfamiliar to some participants, requiring locally meaningful wording and prototypes with translators.
Translation and local phrasing are critical: numeric rating scales were unfamiliar to some participants and required locally meaningful wording. Prototype with translators and frontline VHTs and include pictorial or audio affordances.
Minimize friction by streamlining reporting frequency or allowing adaptive schedules when symptom burden is high to reduce data fatigue.
For policy & health system planners
Integration matters and clinicians in the study asked to map dashboard outputs into DHIS2 and referral workflows, which requires governance and budget planning.
Integration matters: clinicians asked to map dashboard outputs into DHIS2 and referral workflows. Plan for governance, data sharing agreements, and budget lines for devices, connectivity, and training.
Do more, faster with Evidano (mapped to this use case)
Problem: messy, multilingual inputs → Evidano
Evidano ingests transcripts, app logs, and CSV exports and normalizes local terms with translation memory and custom dictionaries.
Import interview transcripts, app logs, and CSV survey or export files into Evidano. Use custom translation dictionaries and translation memory to normalize local terms (for example, Bari, Ma’di, Juba Arabic) before coding.
Problem: inconsistent coding & slow synthesis → Evidano
Evidano accelerates reproducible coding with AI-assisted codebook import and automated thematic analysis combined with human-in-the-loop validation.
Use AI-assisted codebook import and automated thematic and frequency analysis to generate reproducible themes, co-occurrence networks, and hierarchical code to subcode maps, then validate and refine with human-in-the-loop coding.
Problem: comparing engagement across segments → Evidano
Evidano runs cross-segment analyses and time-series visualizations to identify where completion rates drop and correlate those drops with sync failures or device downtime.
Run cross-segment analyses (by VHT, zone, language, time) and time-series visualizations to spot where completion rates drop and correlate with sync failures or device downtime.
Problem: stakeholder buy-in & reporting → Evidano
Evidano exports visual reports and decision memos and supports encryption and on-prem options while not using customer data to train third-party models.
Export ready-to-share visual reports (word clouds, fitted trajectories, quote banks) and generate a concise decision memo for Ministry or donors. Evidano encrypts data, offers on-prem options, and does NOT use your data to train third-party models.
Problem: need rapid follow-up data → Evidano
Evidano can deploy AI avatar interviews to collect follow-up qualitative or structured data autonomously, with PII redaction and consent workflows.
Deploy AI avatar interviews to collect follow-up qualitative or structured data autonomously where VHT capacity is constrained, with PII redaction and consent workflow built in.
Actionable 7-step workflow to reproduce and extend the study with AI enablement
This seven-step workflow lists the concrete file exports and analysis steps to reproduce and extend mPallCare with AI-enabled qualitative analysis.
Step 1: Export app logs (per-report timestamps, flags, VHT ID, language) and interview transcripts; collect CSV baseline demographics.
- Step 2: Ingest all files into Evidano (documents and spreadsheets).
- Step 3: Run automated preprocessing: language normalization, custom dictionary translation, PII redaction.
- Step 4: Auto-generate themes, frequency tables, and time-series of completion by segment; review and refine codebook interactively.
- Step 5: Produce co-occurrence networks and hierarchical code maps to identify symptom clusters and trigger patterns.
- Step 6: Create stakeholder dashboards and downloadable briefings mapped to DHIS2 indicators for policy integration.
- Step 7: Pilot AI avatar follow-ups where VHT capacity or device access is limited; feed new transcripts back into the corpus for continuous learning.
Ethics & data note
The mPallCare study used deidentification and local ethics approval and teams should interpret trends cautiously in palliative populations.
The mPallCare study used deidentification and local ethics approval; similarly, Evidano supports encrypted storage and explicit policies: your data is not used to train third-party models. When working with palliative populations, interpret trends cautiously and monitor for unmet needs that cannot be safely met by program capacity.
FAQ: mPallCare
What was the mPallCare pilot study design and who participated?
The mPallCare pilot was a 6-week mixed-methods feasibility study in Bidibidi with VHT-facilitated mobile reporting and clinician dashboard monitoring, involving 32 people living with advanced cancer.
The pilot combined a VHT-administered Android app (symptom checklist and APCA Palliative Outcome Scale) with a web dashboard for clinicians, collected data in four languages, and included 20 trained VHTs.
What were the completion rates and the study’s feasibility conclusion?
Completion rates were 84.9% for symptom reports, 59.4% for outcome reports, and 67% combined, and the study judged mPallCare feasible and acceptable but sensitive to translation, connectivity, and workflow constraints.
The study reported 84.9% (163/192) symptom-report completion, 59.4% (266/448) outcome-report completion, and 67% (429/640) combined completion, meeting a conservative feasibility threshold for the pilot.
What operational challenges reduced engagement during the pilot?
Operational challenges included navigation and translation confusion, intermittent sync errors, limited device availability, competing VHT outreach duties, and a modest decline in engagement across six weeks.
The study observed initial navigation and translation confusion, intermittent synchronization failures, device availability constraints, and workload competition for VHTs, which contributed to declining outcome-report completion over time.
What UX changes should product teams prioritize for similar pilots?
Product teams should prioritize localized phrasing and rated scales, prototyping with translators and frontline workers, pictorial or audio affordances, and adaptive reporting schedules to reduce data fatigue.
The study recommends prototyping rating scales with translators and VHTs, including pictorial or audio affordances, and streamlining or adapting reporting frequency when symptom burden is high.
How can teams reproduce or extend the study using AI-enabled qualitative tools?
Teams can reproduce the study by exporting app logs and transcripts, ingesting them into an AI-enabled analysis platform, normalizing languages, auto-generating themes and time-series, and iterating with human coding.
Follow the seven-step workflow: export per-report logs and transcripts, ingest to Evidano for preprocessing (language normalization, PII redaction), auto-generate themes and time-series, refine the codebook, and produce stakeholder dashboards.
Wrapping up: next moves
If you are evaluating a pilot like mPallCare, prioritize richer metadata capture, localized UX fixes for rating scales and language, and dashboards that close the feedback loop to VHTs and patients.
- Convert transcripts and exports into evidence faster with Evidano: Evidano.
- Read the full study at JMIR Mhealth Uhealth to match methods and metrics precisely.
Ready to turn qualitative and dashboard logs into reproducible insights and stakeholder-ready reports? Try Evidano for free.