Jinfeng Township Brings VLR into Primary Care: Five Cards for Community Health Governance

On October 9, 2026, the Taitung County Government reported that Jinfeng Township Health Center had received an honorable mention in the Indigenous and Offshore-Island category of the Health Promotion Administration's 16th Primary Health Center Awards. The county described Jinfeng as one of Taiwan's first Indigenous townships to introduce a Voluntary Local Review, or VLR, into local governance. The health center entered under the theme “From VLR to Vitality: Age-Friendly and Socially Co-Created Health Governance in Jinfeng Indigenous Township,” connecting rural health care, older-adult support, community culture, and social co-creation.

The most useful lesson for Taiwan's 55 Indigenous townships is not the award itself. It is that two activities often kept far apart—a local sustainability report and front-line primary care—were brought into one service chain. A VLR is often understood as a local government's report aligning its work with the United Nations Sustainable Development Goals. For residents, however, sustainability cannot remain in a long document. It must answer practical questions: how far an older person travels for care, whether chronic-care follow-up breaks, whether language support exists, whether health programs respect culture, how transport and digital exclusion are addressed, and who must respond after residents raise a concern.

Taiwan's National Council for Sustainable Development explains that voluntary local review helps local authorities take stock, disclose action, connect it to sustainability goals, and sustain improvement through bottom-up participation. Taitung County's VLR also identifies evidence, participation, meaningfulness, and actionability as guiding principles. Applied to health governance in an Indigenous township, the purpose is not to attach an SDG number to every health-center activity. It is to build a closed loop from problem to action, evidence, feedback, and correction.

The first card covers “baseline need and accessibility.” A health center must know who is receiving service, who remains outside it, and how distance, transport, weather, mobility, care arrangements, language, information, and digital capability create different barriers. Visit counts describe people who reached a facility; they do not prove that people who did not appear had no need. Dispersed mountain settlements, limited public transport, wet-season road risk, or the absence of a companion may all hide need from administrative statistics.

Minimum fields include service area, data date, the source of population and health-need estimates, people actually reached, reasons for non-contact, travel time, alternatives, language support, the responsible contact, and the next follow-up. Maps can help identify coverage gaps, but household registration or phone location must not be used to infer an individual's health condition. For small communities, public statistics must also prevent combinations of age, condition, location, and time from re-identifying a person.

The second card covers “cultural safety and shared decisions.” Taiwan's Indigenous Peoples Health Act connects cultural safety to Indigenous knowledge systems and requires health services to respect culture while improving fairness and appropriateness. Community health governance therefore cannot be completed by translating an existing class into an Indigenous language. Nor can a single consultation stand in for participation. Residents need a continuing role in defining the problem, selecting indicators, shaping service, authorizing data use, interpreting results, and revising action.

For each action, the record should identify who proposed it, who is affected, which language is used, who confirmed its cultural basis, whether it involves traditional medicine or ethnobotanical knowledge, what may be public, and what must remain within an authorized group. Consent for audio, images, interviews, and health data should be separate. Participation in an event must not be treated as permission for later research, publicity, or model training. AI may organize approved minutes or prepare multilingual drafts, but it must not decide who represents a community or send restricted knowledge into an external system.

The third card covers the “health-service journey.” A resident's experience is rarely one appointment. It can run through information, screening, scheduling, transport, consultation, tests, medication, referral, rehabilitation, home support, and follow-up. A break at any point can undo earlier work. If a VLR reports only the number of sessions and participants, it misses the real costs of repeated explanations, travel between providers, inability to take leave, unstable connectivity, and unanswered referrals.

The journey card records waiting time, transport, language needs, required documents, the referral destination, response deadline, follow-up result, and remedy contact at each point. Acute events must follow clinical and emergency procedures without waiting for AI-generated advice. For chronic conditions or older-adult care, reminders may flag overdue follow-up, but any risk classification needs professional review. Residents must retain the right to refuse, choose alternatives, and correct their records.

The fourth card covers “indicators, evidence, and data quality.” Sustainability indicators are not decorative reporting devices. They should let residents and implementers see whether a problem improved. Health governance can combine four types: inputs such as staffing, service hours, and available transport; processes such as referral completion, language support, and follow-up; outcomes such as fewer interruptions or improved health literacy and daily function; and equity measures comparing access across settlements, age, gender, and mobility.

Every indicator needs a definition, source, update frequency, denominator, missing-data note, limitation, owner, and change record. Two completion rates with different denominators are not directly comparable, and missing data must not be recorded as zero. Interviews and older people's accounts are not ornaments around quantitative results; they explain why numbers changed. If AI classifies feedback or drafts summaries, the original record, model version, and human review must be retained, with checks for systematic errors involving Indigenous languages, accents, and local usage.

The fifth card covers “public feedback and accountability.” A VLR should not end with periodic publication. Commitments, progress, limitations, and next steps need to be disclosed in a form residents can understand, question, and see answered. Disclosure can have three layers: plain-language service information and aggregate indicators for everyone; de-identified community or service analysis; and individual records available only to authorized professionals. These layers must not be mixed in one downloadable file.

The accountability card lists the commitment, responsible unit, start date, checkpoints, current status, obstacle, resident feedback, response, and next update. If a target is missed, publishing the cause and correction is more valuable than cosmetically improving an indicator. Residents also need phone, paper, in-person, and digital feedback channels. People with weak connectivity, difficulty reading, or limited form experience cannot be limited to an online portal. Complaints involving personal health or culturally sensitive content require a confidential human process with clear referral and response times.

The five cards operate as one chain. The baseline card identifies people and barriers that have been overlooked. The shared-decision card tests cultural fit and resident choice. The journey card traces where service breaks. The evidence card asks whether action produced improvement. The accountability card returns the result to residents and starts the next revision. Without any one of these, a VLR can become an upward-reporting exercise detached from daily life.

A local government could begin with a 90-day pilot. During the first 30 days, select one health issue prioritized by residents and establish the baseline, sources, and shared-decision rules. During the second 30 days, connect the service journey, referrals, and three to five indicators. During the final 30 days, publish an accessible result, gather resident feedback, and revise the process. The objective is not maximum data collection. It is whether gaps are found, responses become faster, cultural authority is clear, and residents can understand and correct the record.

Digital tools can reduce work when carefully bounded. Forms can save offline in low-connectivity areas. Dashboards can warn that evidence is stale or a referral is overdue. Speech tools can create text that remains pending until confirmed, and multilingual systems can prepare translation drafts. Every tool still needs a paper or human alternative and a clear statement of storage, access, retention, and error handling. In small communities especially, convenience must not come at the cost of anonymity, cultural rights, or trust.

Jinfeng Township Health Center's recognition does not provide a standard package that every township should copy. It provides a method worth testing: translate the global language of sustainability into locally actionable health questions, then connect residents' experience, front-line service, and public indicators. When a VLR can answer who remains unserved, why, who must improve the situation, and when a response is due, it becomes more than a report. It becomes a governance tool that communities can use every day.

Sources

• Central News Agency service: Jinfeng Township Health Center recognized after bringing VLR into Indigenous-township health governance
• Ministry of Health and Welfare: 2025 Healthy City and Age-Friendly City award list
• National Council for Sustainable Development: Voluntary sustainability reviews
• National Council for Sustainable Development: Taitung County Voluntary Local Review
• Laws and Regulations Database: Indigenous Peoples Health Act