Patient voice — stroke

This directory is the patient-experience layer of the stroke knowledge base. It describes what stroke survivors and unpaid caregivers report about emergency care, rehabilitation, communication, identity, fatigue, work, medication and life after discharge. It complements clinical outcomes: an mRS score can summarize dependence but cannot capture aphasia-related exclusion, cognitive effort, changed family roles or the work required to navigate fragmented services (Reed 2012, PMID 21985186; Sarre 2014, PMID 23883420).

Files in this layer

File	Contents
organizations.md	Verified organizations and networks across regions; services, focus and research/advocacy roles; fetch failures quarantined
themes.md	Aggregate thematic synthesis; every established theme has at least two independent sources
sources.md	Annotated PubMed and public-web sources, search/fetch record and explicit coverage limits

The wiki-level synthesis is planned for wiki/patient-experience-and-advocacy.md; this literature layer preserves the evidentiary substrate.

Method — session of 2026-08-30

Published research

Live PubMed E-utilities searches combined stroke or stroke survivor with:

• lived experience, patient experience, qualitative, meta-synthesis, identity, adjustment;
• caregiver experience, caregiver unmet needs, hospital-to-home transition;
• aphasia, communication, healthcare, identity;
• post-stroke fatigue, sensory impairment, invisible impairment;
• return to work, vocational activity, young stroke;
• patient priorities, unmet needs, peer support, physical activity barriers;
• medication-taking, adherence, self-management.

Every PMID used in these files was returned by a live PubMed query and then resolved through esummary or efetch in this session. Qualitative syntheses were preferred over isolated stories; primary qualitative studies were retained where they expose an under-measured domain.

Organizations

Organizations were sought by region and by population need: general stroke, younger stroke, aphasia/communication, caregiver support and rehabilitation. Each main-table organization had its own site fetched on 2026-08-30 and returned HTTP 200. A fetched page verifies current web presence, not charity registration, service quality or national reach. Sites returning 403, DNS failure or 404 are listed separately and their content is not used.

Public story sources

Two public organizational story hubs were fetched successfully: Stroke Support Station Singapore’s “Stories & Impact” resource filter and Heart & Stroke Canada’s public stories page. They were used only to confirm the existence and genre of public testimony; no private individual is named, quoted or profiled here. Other guessed story URLs that returned 403/404 were excluded.

Ethics rules applied

1. Public sources only. No closed group, member forum, social-media comment thread, messaging channel or logged-in community was accessed.
2. Aggregate synthesis. A pattern becomes a theme only when supported by at least two independent published or public organizational sources.
3. Paraphrase. No patient quotation is reproduced in this build. This is stricter than the repository maximum of 15 attributed words.
4. No private names or identifying details. Public story pages are cited at hub level and summarized in aggregate.
5. No cross-source dossiers. Details from multiple sources are never combined into an individual narrative.
6. Voice is not efficacy evidence. A testimonial can identify priorities or harms; it cannot establish treatment benefit.
7. Communication access matters. Aphasia can exclude the people most affected from interviews and surveys. Studies that adapt interviewing are particularly important, while proxy reports are not assumed to equal survivor reports (Clancy 2020, PMID 30261762; Musser 2015, PMID 25685553).
8. Caregiver voice is distinct. Caregiver burden and survivor preference are reported separately even when care is interdependent (See Toh 2022, PMID 35986588; Denham 2022, PMID 32393074).

How to interpret the synthesis

• Qualitative findings establish recurring patterns and mechanisms, not prevalence.
• Survey prevalence is instrument- and denominator-dependent. The proportion with any unmet need ranges widely because questionnaires cover different domains (Lin 2021, PMID 33879490; Chen 2019, PMID 31110106).
• “Recovery” has several meanings: impairment change, independence, participation, identity and valued-role resumption may move differently.
• Stroke populations are heterogeneous by type, severity, age, communication ability, culture, income, rehabilitation access and time since stroke.
• Public organization stories are selected narratives, not representative samples.

Updating this layer

• Re-fetch every organization before adding or materially revising its entry; record HTTP outcome and access date.
• Add a new theme only after finding a second independent source.
• Prefer accessible research designs that include aphasia and cognitive/communication disability rather than excluding them for convenience.
• Record language, country, time since stroke and recruitment setting in sources.md when a source materially shapes a claim.
• Preserve failures and regional gaps; do not infer that a missing website means no organization or no patient community exists.
• Propagate any material synthesis change to the planned patient-experience wiki page and the condition log in the coordinating build session.

Coverage boundary

This build is strongest for English-language research and organizations with stable, indexable websites. It includes verified organizations in Europe, North America, Australasia, Southeast Asia, South Africa and Brazil, but it does not constitute a census. South Asia, most of Africa, the Middle East, East Asia outside Singapore, and Spanish/Portuguese-language qualitative research remain underrepresented. Full coverage accounting is in sources.md.