Person-Centred Care in Taiwan: Choice, Autonomy and the Changing Relationship Between Services and Citizens

An older woman qualifies for long-term care support but wants to continue preparing part of her own breakfast. A disabled man needs assistance with daily activities but wants support organised around employment rather than around the most convenient service timetable. A family asks for more supervision because it worries about an older relative falling, while the relative values being able to leave home independently. None of these situations can be resolved simply by asking whether a service is available.

They raise a more demanding question: who gets to define what good support is for?

That question is increasingly important within the Taiwan Ageing, Long-Term Care & Community Support Knowledge Hub as Taiwan moves from the expansion achieved through Long-Term Care 2.0 into Long-Term Care 3.0. The new phase, implemented from January 2026, is intended to strengthen continuity across family support, home and community services, residential care, health care and social welfare while advancing healthy ageing, ageing in place and more integrated support.

Expansion matters. Yet a system can offer more services without necessarily giving people greater influence over how those services shape their lives. Person-centred care therefore needs to reach beyond respectful communication or individual care plans. It concerns how eligibility is assessed, how options are explained, how services respond to changing goals, how families participate, how risk is negotiated, how technology is introduced and whether the person remains an active participant when needs become complex.

The central policy challenge for Taiwan is increasingly to ensure that a larger long-term care system does not become a more standardised one.

Person-centred care begins with a different definition of need

Long-term care systems necessarily classify need. Taiwan uses assessment processes to determine eligibility and levels of long-term care need, enabling people who meet the relevant criteria to access publicly supported services within the applicable payment and benefit arrangements. Administrative consistency matters because public resources need transparent rules.

But an assessed level of dependency cannot describe a life.

Two people with similar mobility limitations may want very different outcomes. One may prioritise being able to bathe safely without relying on a daughter. Another may care more about reaching a neighbourhood association several times a week. A person with cognitive impairment may value remaining in a familiar apartment even if this requires a more complex combination of home support, family involvement and technology.

The distinction matters because service-led systems tend to translate assessment directly into tasks: bathing, meals, transfers, transport, supervision. Person-centred systems add another layer by asking what those tasks enable.

This is the difference between providing assistance and supporting a life.

Taiwan’s increasing emphasis on ageing in place creates a strong policy environment for this shift because remaining in the community is meaningful only if support preserves the relationships, routines and identities that make the community worth remaining in. The broader principle of tailoring support to the individual therefore has operational consequences for assessment, care management and service delivery.

A person-centred assessment needs to understand functional limitation while also identifying strengths, preferences, informal support, environmental barriers and personally meaningful outcomes. It should distinguish what somebody cannot currently do from what they could do with appropriate rehabilitation, equipment, environmental adaptation or carefully calibrated assistance.

LTC 3.0 creates more choice only if options are genuinely usable

Long-Term Care 3.0 builds on the service infrastructure developed under LTC 2.0 and aims to provide more continuous support across home, community, institutional, medical and social-welfare settings. The direction supports greater flexibility because people potentially have access to a broader combination of interventions rather than one dominant model of care.

Formal availability, however, is not the same as practical choice.

A person may theoretically be able to use day care but have no realistic transport. Home care may exist locally but not at the times that matter most. A family may know that respite is available but find the application or scheduling arrangements difficult to navigate. Rural and remote communities may have fewer alternatives than major urban areas. People employing migrant family caregivers have historically occupied a distinctive position within the system, although reforms have progressively widened access to specified community-based services.

Choice is therefore produced by the interaction of entitlement, local capacity and usable information.

Municipalities and county or city governments have an important implementation role because the practical service network is local. National policy can define eligibility and benefit structures, but the range of realistic options depends on provider availability, workforce, transport, opening hours and the density of community resources.

This creates an important governance test. A system should not measure choice only by counting how many service categories appear in policy. It also needs to understand whether eligible people can actually access different models within a reasonable distance and timeframe.

Organisations examining similar service-design questions can use the Digital Twin Scenario Modeller to explore how changes in demand, workforce and capacity may affect service options. It is not a Taiwan-specific planning instrument, but its scenario approach illustrates an important principle: individual choice depends partly on system capacity being designed before demand becomes impossible to accommodate.

Scenario: the service package is technically correct but the outcome is wrong

A 78-year-old man develops mobility difficulties after a period of illness. Assessment confirms that he requires assistance with several daily activities. His daughter initially asks for home-care visits concentrated around personal care because she assumes this is the main support available.

During care planning, however, the man explains that his priority is returning to a neighbourhood cultural group that he attended for many years. He can wash and dress slowly with some assistance, but he has stopped going out because navigating the apartment entrance and travelling independently have become difficult.

A task-based response could deliver regular personal care efficiently while leaving him increasingly isolated. A person-centred response looks differently at the same assessed need. Rehabilitation potential is considered. The home environment and appropriate assistive equipment are reviewed. Support is arranged to preserve the activities he can still perform rather than replacing them automatically. Community and transport options are explored alongside conventional home care.

The result may still include assistance with personal care, but its purpose changes. The care package is organised around maintaining function and enabling participation rather than completing the maximum number of tasks for him.

Review also becomes more meaningful. Instead of asking only whether scheduled visits occurred, the care team can examine whether mobility has improved, whether support remains proportionate and whether he has resumed activities that matter to him.

This is the operational significance of outcomes-focused support: service activity remains important, but it is no longer mistaken for the outcome itself.

Choice requires information people can actually understand

People cannot exercise meaningful choice if they do not understand their options.

Taiwan’s long-term care system involves assessment, service categories, payment arrangements, local providers and interfaces with health care, disability support and privately arranged care. Families may encounter it for the first time during a hospital admission, sudden deterioration or caregiver crisis. At that point, administrative complexity can make apparently simple decisions difficult.

Person-centred practice therefore depends on communication as much as on the number of options available.

Information needs to explain what support is available, what the person may need to contribute, how services differ, what can be combined and what happens if circumstances change. Communication also needs to accommodate sensory impairment, cognitive difficulties, language needs, health literacy and different levels of digital confidence.

This becomes particularly important where families undertake much of the navigation. Relatives can provide invaluable support, but information should not automatically bypass the person receiving care merely because a family member is easier for professionals to communicate with.

The wider principle of accessible information is therefore part of autonomy. A choice presented in language or a format the person cannot understand is not meaningfully theirs.

Family involvement should strengthen the person’s voice, not replace it

Family caregiving remains deeply embedded in Taiwan’s long-term care landscape. Relatives frequently organise care, provide daily assistance, employ migrant family caregivers, accompany people to health appointments and make practical arrangements across several services.

Person-centred care should not treat this contribution as peripheral. Families often know the person’s history, routines and communication better than professionals do. They may also carry substantial physical, emotional and financial responsibility.

Yet family involvement and personal autonomy are not identical.

A daughter may want her father to stop walking outside because she fears a fall. A son may prefer residential care because coordinating support at home has become difficult. Relatives may favour more monitoring than the person considers acceptable. In some cases, family preferences may be shaped by legitimate caregiver limits rather than by the older or disabled person’s own priorities.

Strong practice makes those perspectives visible rather than pretending they are always aligned.

The person’s wishes, the family’s capacity and legitimate safety concerns all need to be understood. Where a caregiver is reaching exhaustion, the solution should not be to describe unlimited family care as the person’s preferred choice. Conversely, a professional concern about risk should not automatically override a person’s ordinary preferences.

This is why family and advocate involvement works best when it adds perspective without removing the person from the centre of decision-making.

Autonomy becomes most important when risk increases

Person-centred care is relatively easy to endorse when the person’s preference and professional judgement point in the same direction. The more difficult test comes when they diverge.

An older person may want to continue cooking despite occasional forgetfulness. Someone with mobility difficulties may accept a greater falls risk in exchange for remaining active. A person with cognitive impairment may resist a family request for continuous location monitoring. A disabled adult may choose relationships or activities that relatives consider unwise.

Care systems can respond to these situations by progressively removing risk, but doing so can also remove independence.

Taiwan’s legal and policy landscape increasingly gives autonomy greater visibility. The Long-Term Care Services Act protects the rights and interests of people receiving long-term care and prohibits inappropriate restriction of physical freedom within long-term care services. Disability rights frameworks similarly reinforce participation and protection against improper restriction. Taiwan’s Patient Right to Autonomy Act has also established a wider legal recognition that capable individuals can make informed decisions about specified future medical treatment through advance care planning and advance decisions.

These frameworks are not interchangeable, and the Patient Right to Autonomy Act is a health-care law rather than a general long-term care decision-making statute. Together, however, they reflect a broader direction: receiving care does not remove personhood or the importance of informed preference.

For day-to-day services, positive risk-taking means understanding the benefit associated with an activity as well as the possible harm. The relevant question becomes not simply “How do we stop this happening?” but “How can this person pursue what matters to them with proportionate safeguards?”

The Positive Risk-Taking Planner can help organisations structure this reasoning. It does not determine legal capacity or replace Taiwanese law, but it provides a transferable method for making the person’s objective, the specific risk, available mitigations and agreed review arrangements visible.

Scenario: a daughter wants safety while her mother wants privacy

An 82-year-old woman lives alone and receives home-care support. After two minor falls, her daughter proposes installing cameras throughout the apartment so that family members can check on her remotely. The daughter lives some distance away and sees monitoring as a way of allowing her mother to remain at home.

The woman dislikes the idea. She accepts that falls are a concern but says she does not want relatives watching her daily life.

A person-centred response does not dismiss the daughter’s anxiety or treat the woman’s objection as an obstacle. The falls are assessed more broadly: medication, mobility, footwear, lighting, bathroom safety and the circumstances of each incident are reviewed. Rehabilitation and environmental adaptations are considered, along with less intrusive technology capable of addressing the specific risk.

From July 2026, Taiwan’s LTC 3.0 arrangements added a rental-based smart assistive-technology benefit alongside existing assistive-device and home-accessibility support, widening the potential range of equipment that eligible people can use as their needs change. The existence of technology nevertheless does not answer the ethical question of how it should be used.

The woman ultimately agrees to a more limited safety arrangement rather than continuous video monitoring. Her daughter gains greater reassurance, while the intervention remains proportionate to the identified risk.

The scenario demonstrates an important distinction. Technology can enable ageing in place, but person-centred technology is defined by the relationship between purpose, consent, proportionality and outcome, not simply by technical capability.

Technology is shifting choice from ownership towards adaptability

Taiwan’s expansion of smart assistive technology under LTC 3.0 is particularly relevant to person-centred care because functional needs change. A device appropriate after hospital rehabilitation may become unnecessary as function improves or unsuitable if disability progresses.

The rental approach introduced in 2026 for eligible smart assistive technologies creates greater scope for equipment to change alongside the person rather than locking support into one purchase. Covered categories are intended to address areas such as mobility, transfers, bathing and toileting, care beds and safety monitoring, subject to the applicable assessment and benefit arrangements.

This flexibility has a person-centred advantage: technology can potentially follow changing need.

But implementation matters. A technically sophisticated device may be unusable for somebody with limited digital confidence. Sensors may generate alerts that increase family anxiety rather than independence. Equipment may remain unused if workers and relatives are not trained. Data-generating technologies can create privacy questions that traditional mobility aids do not.

Person-centred technology therefore requires several questions before adoption:

• What outcome does the person want the technology to support?
• Does the person understand how it works and what information it collects?
• Can it be used safely within the actual home environment?
• Who will respond when the device produces an alert?
• Will it reduce unnecessary dependence or simply create another layer of monitoring?

Organisations examining these questions can use the Digital Transformation Readiness Assessment to test whether technology strategy is matched by governance, workforce capability and information controls. The framework is transferable rather than a Taiwanese regulatory standard.

Care planning should be a continuing negotiation, not a one-time allocation

People’s needs change. So do their goals, households and tolerance of risk.

A person discharged from hospital may initially need intensive assistance but recover function through rehabilitation. Someone living with dementia may need a different balance of support as cognition changes. A family caregiver who previously provided substantial help may become ill or return to employment. An assistive device may reduce the need for one form of assistance while creating a new training requirement.

Person-centred care planning therefore requires review rather than permanent translation of an initial assessment into a fixed service pattern.

The principle of support planning and review is especially important under LTC 3.0 because Taiwan is seeking stronger continuity between medical care and long-term support. If services respond to changing functional need, review needs to capture improvement as well as deterioration.

This matters for independence. A system can unintentionally create dependency when assistance continues at a level that is no longer necessary. Workers may complete tasks more quickly than supporting the person to perform them slowly. Families may understandably prefer arrangements that feel safer. Payment and scheduling systems may reward completion rather than gradual recovery.

Person-centred review should therefore ask whether the current level of support remains appropriate, what has changed, what the person now wants to achieve and whether formal care can be reduced, redirected or strengthened.

The goal is not always less care. For some people, person-centred review will identify that support needs to increase. The principle is responsiveness rather than reduction.

Scenario: rehabilitation changes what the person needs from long-term care

A 74-year-old woman experiences a stroke and enters post-acute rehabilitation. At discharge she requires substantial assistance with transfers, bathing and mobility. Her family expects that this level of support will become permanent.

Under the stronger medical and long-term care connection being developed through LTC 3.0, her transition home is treated as part of a continuing pathway rather than the endpoint of hospital treatment. Long-term care support is arranged, but rehabilitation goals remain visible within the plan.

During the following months, her transfer ability improves and she can walk short distances with appropriate equipment. If the original care package simply continues unchanged, workers may keep performing activities she is increasingly capable of undertaking herself.

Instead, review considers both safety and regained ability. Assistance is adjusted so that workers support rather than replace movement where appropriate. Home adaptations reduce environmental barriers. Her own priority is to be able to reach a nearby shop with a family member rather than remain dependent on a wheelchair for every journey.

The outcome is not measured solely through fewer care hours. It includes greater mobility, confidence and participation while maintaining enough support to prevent avoidable deterioration.

Person-centred care in this context is closely connected with rehabilitation. The service adapts to the person rather than requiring the person to remain fitted to the service package originally created at the point of greatest dependency.

Advance care planning extends autonomy into future health decisions

Person-centred long-term care also needs to recognise decisions that extend beyond everyday support.

Taiwan’s Patient Right to Autonomy Act established a framework through which adults with the required decision-making ability can participate in advance care planning and, subject to the legislation, make advance decisions concerning specified future clinical circumstances. National Health Insurance has progressively expanded payment for advance care planning consultations for defined groups, including additional groups of older people with significant illness or multiple chronic conditions.

This is significant because long-term care increasingly supports people through frailty, dementia, multiple chronic illnesses and the later stages of life. Preferences about medical intervention, place of care and what the person considers an acceptable quality of life can become increasingly important.

Advance care planning should not be reduced to a document completed shortly before death. Its deeper contribution is to create structured conversation while the person is able to express preferences clearly.

For long-term care services, the operational requirement is not to replace clinical processes but to know when relevant advance decisions exist, ensure information can follow the person appropriately and avoid repeatedly placing families in the position of trying to reconstruct wishes during a crisis.

This becomes increasingly relevant as LTC 3.0 pursues continuity extending towards end-of-life care. Respect for autonomy should not stop when a person crosses from community care into hospital treatment or from home care into residential support.

Residential care tests whether person-centred principles survive institutional routines

Residential services face a particular challenge because collective living requires routines. Meals, medication, staffing, bathing, activities and overnight support all need organisation. Efficiency matters when many people require assistance from the same workforce.

Yet routines can gradually become the organising principle of residents’ lives.

A person-centred residential service asks where standardisation is genuinely necessary and where choice can remain. Residents may differ in when they want to get up, what they eat, how they spend time, which relationships matter, what privacy means to them and how much assistance they want.

The relevant principle of person-centred planning for older people becomes particularly important for residents with significant dependency because the more assistance somebody needs, the easier it is for organisational routines to dominate.

Families can help services understand history and preferences, particularly where communication becomes difficult, but residents should not be treated as passive recipients simply because they live in an institution.

Quality oversight therefore needs to look beyond whether required care tasks were completed. It should also ask whether people retain meaningful control over ordinary aspects of daily life.

Person-centred care depends on the workforce having discretion

Policy can promote autonomy, but frontline interactions determine much of what people actually experience.

A worker deciding whether to wait while somebody buttons their own clothing, a care manager exploring what outcome matters to a family, a day-care employee adapting an activity around somebody’s interests or a residential worker supporting a resident to make an ordinary choice all exercise professional judgement.

This means person-centred care has workforce implications.

Workers need communication skills, knowledge of rights, understanding of rehabilitation and confidence in proportionate risk management. They also need enough continuity to know the person. High turnover and fragmented scheduling make individualised support harder because workers repeatedly begin with limited knowledge.

Time matters as well. Task-focused models can unintentionally reward speed. Helping somebody perform an activity themselves may take longer than completing it for them. Listening to preferences may disrupt a pre-planned schedule. Supporting community participation can be operationally more complex than providing care inside the home.

The challenge is therefore not solved simply by telling workers to “be person-centred”. Service design needs to give them enough information, capability and discretion to act accordingly.

This also requires boundaries. Individualisation cannot mean promising any service at any time regardless of resources, worker safety or public funding rules. Person-centred practice is the disciplined negotiation of what matters to the person within legitimate constraints, not the absence of constraints.

Scenario: a service timetable begins determining a disabled person’s employment

A 42-year-old disabled man receives assistance with personal care and preparing for the day. He works several days each week and needs support earlier on those mornings. A service reorganisation moves his visits into a standard scheduling window that is easier to staff but would regularly make him late for work.

On paper, the same quantity of care remains available. Functionally, the change threatens employment and independence.

A person-centred review reframes the issue. The purpose of morning support is not simply to complete personal-care tasks; it enables the man to participate in employment and community life. Scheduling therefore becomes part of the outcome.

The provider examines whether worker deployment can be adjusted, whether the pattern can be made predictable enough to support workforce planning and whether alternative arrangements are available on the days when an earlier visit cannot be delivered. The man is involved in designing the solution rather than simply being notified of a new timetable.

This does not mean every requested time can always be guaranteed. Workforce constraints are real. The difference is that the impact on the person is visible in the operational decision.

Where similar conflicts recur across many people, the issue should move beyond individual complaint handling. It may indicate that workforce capacity or scheduling rules are systematically reducing choice. Person-centred governance turns repeated individual experience into evidence about service design.

Data should show whether people are achieving outcomes, not only receiving services

Long-term care systems naturally generate administrative data: assessments, service utilisation, payments, workforce activity and provider information. These data are necessary for financial control and capacity planning.

They reveal less about whether support is helping people live the lives they value.

A person-centred evidence model needs to connect service activity with outcomes. Depending on the service and individual, these may include maintaining function, remaining at home, reducing caregiver strain, participating in community life, regaining mobility, sustaining relationships or experiencing greater confidence in daily activities.

Not every outcome can be reduced to a single national indicator. Individual goals vary. Yet this does not mean person-centred outcomes are too subjective for governance.

Services can examine whether goals are identified, whether people participate in reviews, whether plans change when circumstances change and whether recurring feedback reveals restrictions created by service design. Complaints and family feedback can add further perspective without substituting for the person’s own experience.

The Quality Dashboard Builder offers organisations a way of combining different evidence sources rather than relying on one activity measure. In Taiwan or another international setting, its value lies in the analytical principle: quality becomes clearer when utilisation, experience, outcomes and risk are considered together.

This is also where service-user feedback and co-production become governance evidence rather than optional engagement exercises.

Local variation can create unequal degrees of choice

Taiwan’s national long-term care framework operates through locally developed service networks. Municipalities and counties or cities therefore matter greatly to the person-centred experience.

Dense urban areas can support multiple providers and specialised services because enough demand exists within a relatively small geography. Rural, remote and island communities face different conditions. Travel time, workforce availability and smaller populations can make some service models difficult to sustain.

Formal national entitlement can therefore coexist with local differences in practical choice.

This does not mean every locality can or should offer an identical provider market. Geography matters. The governance requirement is to understand where variation becomes inequity and to develop alternative models where conventional service delivery is unrealistic.

Technology may extend specialist input. Transport can connect people with community provision. Flexible or multi-purpose services may work better where demand cannot support several separate organisations. Workforce development may need to be targeted geographically.

The important point is that person-centred policy cannot assume that choice operates in the same way everywhere. For somebody living in a remote area, continuity with one reliable service may be more meaningful than theoretical choice between providers that cannot realistically reach them.

Co-production requires more than asking whether people are satisfied

Person-centred care concerns individual relationships, but it also has implications for how systems develop.

People receiving long-term care and family caregivers hold operational knowledge that administrative data cannot reproduce. They know where assessment processes are confusing, where transport prevents access, where service times do not match ordinary life and where transitions between health and long-term care create repeated explanations or gaps.

Collecting satisfaction scores after services have been designed captures only part of this knowledge.

Co-production moves participation earlier. People can contribute to the design of information, service models, technology, quality indicators and improvement priorities. This does not require every policy decision to be made by public vote. Government still has responsibility for legislation, budgets and equitable allocation of public resources.

It does require meaningful mechanisms through which lived experience can influence those decisions.

The distinction becomes particularly important during LTC 3.0 because the system is changing. New technology, stronger medical-care integration and expanding eligibility can all create unintended operational consequences that become visible first to people using services.

A mature feedback system therefore asks not only whether people liked a service, but what their experience reveals about how the system itself is functioning.

Governance needs to distinguish individual preference from systemic performance

Person-centred care can become difficult to govern because individual preferences differ. One person wants extensive family involvement; another wants greater privacy. One values technology; another avoids it. One prioritises independence despite risk; another prefers more support and reassurance.

Good governance does not attempt to standardise those preferences.

It standardises the expectation that preferences are sought, recorded, considered and reviewed.

This creates a different assurance model. Decision-makers can examine whether assessments capture meaningful goals, whether people receive accessible information, whether service plans reflect agreed outcomes and whether restrictions have a clear rationale. They can look for patterns in complaints, missed choices and local access differences.

Organisations examining their own structures can use the Governance Maturity Assessment to test whether information from frontline practice reaches decision-makers and leads to action. It is not an assessment of compliance with Taiwan’s system; its relevance lies in making accountability and learning more explicit.

This distinction between standardising process and standardising lives is fundamental. Public long-term care needs consistency in eligibility, financial control and basic quality expectations. Person-centred practice requires flexibility in how those structures translate into individual support.

The future relationship is increasingly one of partnership rather than dependency

Taiwan’s long-term care reforms have progressively shifted the role of public support. The system is no longer concerned only with responding when families can no longer cope. Community care, prevention, rehabilitation, assistive technology and stronger health-care integration increasingly position long-term care as infrastructure that can help people maintain ordinary life.

That evolution changes the relationship between citizens and services.

The older model of dependency asks what the service will do for the person. A more person-centred model asks what the person wants to continue doing, what barriers are preventing it and what combination of formal care, rehabilitation, technology, family support and community infrastructure can help.

This approach should not obscure severe dependency. Some people require extensive assistance, continuous supervision or residential support. Person-centred care does not make those needs disappear. Its importance may actually increase as dependency grows because the person has fewer opportunities to exercise control without deliberate support from others.

Nor should choice be used to shift responsibility from the state or provider to the individual. Offering a menu of inadequate options is not empowerment. People can exercise meaningful choice only where services have sufficient capacity, information is accessible and public funding arrangements make appropriate support realistically obtainable.

The stronger opportunity for Taiwan lies in combining system expansion with this deeper understanding of citizenship.

International learning: autonomy has to be designed into the system

Taiwan’s person-centred trajectory reflects its own long-term care legislation, family structures, public financing arrangements, National Health Insurance system, disability-rights framework and local service infrastructure. Those institutions cannot be transferred directly to another country.

The experience nevertheless highlights several broader principles.

Choice depends on capacity. National entitlement means less when the relevant service cannot be accessed locally. Autonomy depends on information because people cannot influence decisions they do not understand. Family involvement is strongest when it supports rather than replaces the person’s voice. Technology creates genuine independence only when its use remains proportionate and understandable. Review matters because person-centred care cannot remain fixed while the person changes.

Most importantly, person-centredness has to exist at several levels simultaneously.

Frontline workers need to listen and adapt. Providers need enough flexibility to respond. Local systems need realistic service options. National policy needs funding and eligibility arrangements that permit different forms of support. Governance needs evidence capable of showing whether individual outcomes are being achieved.

The transferable lesson lies less in any particular Taiwanese programme than in this alignment. Person-centred care becomes durable when the system around the frontline relationship is designed to make it possible.

Conclusion

Taiwan’s move into Long-Term Care 3.0 creates an important opportunity to deepen person-centred care at the same time as the long-term care system expands. Wider eligibility, stronger medical-care integration, community support, rehabilitation, assistive technology and ageing-in-place policy can all increase the range of ways in which people are supported. Their value, however, depends on whether those resources remain connected to the lives people actually want to lead.

The central challenge is therefore not simply to provide more choice. It is to make choice meaningful. That requires accessible information, realistic local options, continuing review, proportionate approaches to risk, respectful family partnership and a workforce able to exercise judgement rather than merely complete tasks. It also requires governance capable of distinguishing service activity from personal outcomes.

As care needs become more complex, autonomy should not become progressively less relevant. People requiring substantial support may need more deliberate help to express preferences, maintain relationships and retain control over ordinary decisions. Technology and family involvement can strengthen that control, but they can also diminish it if convenience or protection automatically overrides the person’s perspective.

Taiwan’s next stage of long-term care development will therefore be shaped not only by how many services it can build, but by the relationship those services establish with citizens. A mature system does more than care for people safely. It uses public policy, professional practice and community infrastructure to help people remain authors of their own lives for as long, and in as many ways, as possible.