Planning long-term care requires more than knowing how many older people live in a country. Decision-makers need to understand who requires assistance, what type of support they receive, who provides it, whether needs are changing and where gaps are emerging. In Türkiye, much of that information exists somewhere within health, social-service, demographic or administrative systems. The harder task is turning separate information into intelligence about the long-term care system as a whole.
This matters increasingly as Türkiye ages. The population aged 65 and over reached 11.1% of the population in 2025, while more than one quarter of households included at least one older person. The Türkiye Aging, Long-Term Care and Community Support Knowledge Hub examines the wider implications of that demographic transition across financing, workforce, home care, community services and integration. Each of those issues ultimately creates a data question: what does Türkiye know about need, provision, quality and outcomes, and what remains difficult to see?
The challenge is structural. Long-term care responsibilities cross the Ministry of Family and Social Services, the Ministry of Health, provincial structures, municipalities, public and private providers, households and families. Different organisations collect information for different purposes. Health records support treatment; social-service records support eligibility and delivery; facilities maintain operational records; municipalities collect information about their programmes; national statistics describe population change. None of those datasets alone provides a complete picture of long-term care.
The stronger opportunity is therefore not simply to collect more data. It is to create an information architecture that allows Türkiye to understand need, access, capacity, quality and outcomes across the continuum of support.
Türkiye’s long-term care data challenge reflects the structure of the system
Fragmented information should not be treated as a purely technical problem. It reflects the way responsibilities are distributed. Türkiye does not operate a single comprehensive long-term care programme through which every older or disabled person enters, receives a standardized assessment and follows one nationally defined pathway.
The Ministry of Family and Social Services has responsibilities spanning older people, disabled people, social assistance and care services. The Ministry of Health oversees health care, including primary care, hospital services and home health provision. Municipalities may provide additional home, social and community services. Residential provision includes public and private organisations, while families continue to provide a large proportion of everyday support outside formal services.
Each part of this landscape has legitimate reasons to collect its own information. The difficulty arises when policymakers need to answer questions that cross organisational boundaries.
For example, how many older people discharged from hospital subsequently require sustained help with activities of daily living? How many people receiving home-based financial assistance also receive formal health or social-care services? Which provinces face the largest gap between estimated functional need and formal service capacity? How much unpaid care is absorbing demand that would otherwise appear within formal services?
These are system questions rather than individual-agency questions. Answering them requires data governance and information accountability that can connect information while preserving clear responsibility for why it is collected, who may use it and how it informs decisions.
Demographic statistics reveal pressure but not the full pattern of care need
Türkiye has a substantial demographic evidence base. The Turkish Statistical Institute, TÜİK, provides population, household, disability, health, income and other statistical information that can help policymakers understand the environment in which long-term care operates.
Population ageing is clearly visible. Türkiye had approximately 9.58 million people aged 65 and over in 2025, representing 11.1% of the population. Ageing is also geographically uneven. Provinces with older population structures face different planning pressures from younger provinces, while population projections indicate that the national proportion of older people will continue to increase substantially over coming decades.
Household structure adds another dimension. In 2025, 26.1% of Türkiye’s households contained at least one person aged 65 or over, and approximately 1.84 million older people lived alone. Those figures matter because living arrangement can influence the practical availability of informal support.
Yet age is not the same as care need. Many people remain independent well beyond 65, while some younger people require substantial long-term support because of disability or illness. A planning model based mainly on age can therefore misallocate resources.
Long-term care intelligence needs to combine demographic information with disability and functional need. The important question is not simply how many older people live in a province, but how many people experience limitations in activities such as washing, dressing, eating, mobility or managing everyday life, and what level of support those limitations create.
Functional need should become a stronger common language for planning
WHO’s assessment of long-term care in Türkiye identified the absence of standardized procedures for assessing need and clearly defined long-term care pathways as important weaknesses. That finding has consequences well beyond individual assessment.
If different services define dependency and eligibility differently, their administrative data cannot easily be combined into a coherent national picture. One system may classify a person according to disability status, another according to clinical diagnosis, another according to income and another according to eligibility for a particular service. Each classification may be appropriate for its own purpose while remaining insufficient for whole-system planning.
A stronger intelligence model would not require every organisation to abandon its existing statutory or operational criteria. It would require enough common information to understand functional need across services.
A core dataset could, for example, distinguish:
- level and type of difficulty with everyday activities;
- cognitive or supervisory support requirements;
- living arrangements and availability of informal support;
- formal services currently received;
- significant changes in function over time; and
- geographic and socioeconomic characteristics relevant to access.
The purpose would be planning rather than creating a simplistic national score that automatically determines entitlement. Functional assessment remains person-specific. At population level, however, greater consistency would allow Türkiye to estimate need more accurately and compare it with available capacity.
Scenario: two provinces can have similar populations but different care pressures
Consider two provincial planning teams reviewing future older-person services. Both provinces have approximately the same number of residents aged over 65, so a planning model based largely on population initially suggests similar future capacity requirements.
Closer analysis reveals different realities. In the first province, a greater proportion of older residents live with adult family members, population density is relatively high and services are concentrated within accessible urban areas. In the second, more older people live alone, settlements are dispersed and a larger group reports limitations affecting mobility and everyday activities.
Service-use data adds another layer. The first province has relatively high use of community and home services. In the second, formal service use appears lower despite indicators of greater functional need. The initial interpretation could be that demand is lower. A better interpretation asks whether the apparent difference reflects unmet need, transport barriers, informal family support or limited service availability.
The planning decision changes accordingly. Rather than allocating resources solely according to the number of older residents, national and provincial decision-makers can examine need, geography, household structure and existing provision together.
This is where population needs assessment becomes operational intelligence. Data does not decide which model a province should adopt, but it makes the assumptions behind resource decisions visible and testable.
Administrative data can show provision that population statistics cannot
National surveys and population statistics provide essential information about need and inequality, but administrative systems provide another perspective: what services are actually being delivered.
Türkiye already collects substantial administrative information across public systems. The Ministry of Family and Social Services can identify people receiving specific forms of social assistance and services. Residential services maintain information about residents and provision. The Ministry of Health operates extensive digital health infrastructure. Municipalities maintain records relating to the services they administer.
Used appropriately, administrative data can help answer practical questions about service capacity, utilisation, waiting, duration and movement between services. It can also support more timely intelligence than periodic surveys alone.
However, administrative data describes the population visible to the administrative system. People who need support but receive no formal service may be absent. Family caregivers can absorb substantial care without generating a service record. Privately purchased care may be captured differently from publicly provided support. Municipal provision can vary geographically.
That distinction is fundamental. A dataset showing 10,000 service users describes 10,000 recorded users; it does not demonstrate that only 10,000 people need the service.
Strong long-term care intelligence therefore triangulates administrative records with population surveys, demographic projections, local evidence and qualitative experience rather than treating one source as the definitive measure of need.
Interoperability is ultimately about continuity, not databases
Türkiye has previous experience of trying to connect information across home-based health and social support. Inter-ministerial arrangements have sought to improve coordination and electronic information exchange between organisations involved in supporting people at home. The underlying principle remains important: people experience one life even when the services supporting them operate through separate institutions.
Interoperability is sometimes reduced to whether two computer systems can exchange information. Technical exchange is necessary, but useful interoperability requires shared meaning and operational response.
If a hospital records that an older person has significant difficulty walking, can the relevant community service understand that information? If a home health team identifies worsening social circumstances, is there an appropriate route to social support? If a municipality identifies an older resident repeatedly requesting help because of declining function, can that information contribute to assessment elsewhere where lawful and appropriate?
The goal is not unrestricted access to every record. It is purposeful interoperability and data exchange around defined care and planning functions.
Organizations examining comparable cross-system arrangements can use the Digital Transformation, AI and Cybersecurity Readiness Assessment to explore information architecture, governance and implementation dependencies. It is not a Türkiye-specific interoperability or compliance framework, but it can help structure the organisational questions that need to accompany technical integration.
Data should follow the person across transitions where possible
Long-term care needs often become visible during transitions. An older person may enter hospital independently and leave with reduced mobility. Someone with dementia may begin requiring substantially more supervision after an acute illness. A family caregiver may reach the point where existing arrangements can no longer be sustained.
These transitions cross organisational boundaries, making them particularly vulnerable to information loss. Clinical discharge information may explain diagnoses and medication while providing little insight into whether the person can safely prepare food, use the bathroom or move around the home. Social-service information may describe support arrangements without being visible to health professionals.
For planning purposes, transition data can reveal recurring pressure points. If large numbers of older people require new support after hospitalisation, that has implications for home care, rehabilitation and family support. If delayed transitions repeatedly occur because community capacity is unavailable, the issue becomes visible as a system constraint rather than a collection of individual cases.
This connects data architecture directly with health and social-care coordination. Information exchange should support the immediate person while aggregated intelligence helps Türkiye understand where pathways repeatedly become difficult.
Scenario: discharge data becomes a planning signal
A hospital in İzmir notices that a growing number of older patients are medically ready for discharge but families are uncertain whether they can manage safely at home. Individual teams resolve cases as best they can, sometimes through home health services, family support, rehabilitation or other local provision.
Initially, the hospital records each discharge as a separate operational issue. There is no regular analysis of why these cases become difficult.
A structured review changes the picture. The hospital begins recording a small set of relevant factors: pre-admission living arrangement, functional change during admission, support available at home, onward service required and whether the planned support was available when needed. The purpose is not to create another complex assessment but to identify recurring patterns.
Several months of data show that the largest difficulty is concentrated among people who lived alone before admission and experienced a new mobility limitation. The problem is not primarily medical discharge decision-making; it is the interface between hospital recovery and community support.
Shared appropriately with relevant provincial and service partners, this evidence can inform capacity planning. Rehabilitation, short-term home support or earlier assessment may offer greater value than repeatedly resolving the same problem at the end of each hospital stay.
The operational lesson is that good intelligence converts repeated individual difficulties into a visible system pattern.
Measuring capacity requires more than counting facilities and places
Long-term care planning often begins with infrastructure because infrastructure is measurable. Türkiye can count residential facilities, beds, service centres and other formal capacity. Those figures are useful, but capacity is multidimensional.
A residential service may have an available place without having the workforce or specialist capability required by a particular person. A home-care programme may technically cover a district while having insufficient staffing to provide the required intensity. A day service may exist but remain inaccessible to someone without transport.
Effective capacity intelligence therefore combines infrastructure with workforce, utilisation, accessibility and complexity. For home and community services, the unit of capacity may be staff hours, geographic reach or caseload rather than beds.
Information about private provision also matters. National planning that captures public capacity but only partially understands private purchasing can misread both availability and household financial exposure.
The wider principle is that capacity should describe what the system can actually deliver, to whom and under what conditions. That is more useful than an inventory of physical assets alone.
Workforce intelligence should connect staffing with future demand
Türkiye’s Twelfth Development Plan recognises the need to strengthen human resources and professional competence in care services. Delivering that objective requires more detailed workforce intelligence than an aggregate count of employees.
Long-term care planners need to understand professional roles, care-worker capacity, geographic distribution, turnover, age profile, skills and the relationship between staffing and the dependency of people being supported. They also need visibility of informal caregiving because family capacity continues to absorb a substantial part of demand.
Workforce information becomes more valuable when combined with demographic and service data. A province with rapid ageing and low formal workforce density presents a different future risk from one with similar population ageing but stronger existing capacity.
Organizations exploring this relationship can use the Predictive Workforce Risk Module to structure analysis of vacancy, turnover, retention and service-continuity risk. It does not forecast Türkiye’s national workforce requirements, but the underlying principle is relevant: workforce information becomes strategic when it is connected to service demand and continuity rather than reported in isolation.
Quality intelligence needs to move beyond activity
Data becomes especially valuable when it helps distinguish service volume from service quality. A long-term care system can expand the number of contacts, places or beneficiaries without necessarily improving people’s outcomes.
Activity remains important. Ministries and service organisations need to know how many people receive support and how resources are being used. But mature performance intelligence also asks what happens to people receiving that support.
Relevant outcomes depend on the service. They may include maintaining functional ability, avoiding preventable deterioration, supporting safe transitions, reducing caregiver strain, improving participation or enabling someone to remain at home where that reflects their preferences and needs.
Quality intelligence should also connect outcomes with safety and experience. Complaints, incidents, falls, medication concerns, hospital use and family feedback can reveal problems that raw activity figures cannot.
The wider Impact Insights framework on outcomes frameworks and indicators is relevant here because measurement should follow the purpose of the service. Türkiye does not need one universal outcome measure for every form of long-term care. It does need enough consistency to determine whether services are maintaining independence, protecting people and responding to changing need.
Scenario: a dashboard shows improvement until the data is disaggregated
A provincial team reviews a community programme supporting older people at home. The headline dashboard appears positive. Service reach has increased, average waiting time has fallen and reported satisfaction is high.
Instead of stopping at the aggregate results, the team examines the information by age, district, living arrangement and level of functional difficulty. A different pattern emerges. People in the provincial centre are accessing the programme quickly, while older residents in several outlying districts wait considerably longer. People living alone are also more likely to discontinue support early.
The overall averages were accurate; they were simply incomplete.
Further review identifies several contributing factors. Travel time limits workforce capacity in dispersed areas, transport difficulties affect attendance at community services and some older people living alone struggle with the administrative steps required to maintain contact.
The service can now respond to an identified access problem rather than a general perception of rural disadvantage. It might redesign routes, use mobile provision, simplify contact arrangements or work with municipalities and other local partners. Subsequent data can show whether the changes narrow the difference.
This illustrates the value of data-led equity planning. National and provincial averages remain useful, but disaggregation reveals who benefits from a service and who remains less visible within apparently successful performance.
Dashboards are useful only when they change decisions
Digital systems make it increasingly easy to produce dashboards. The danger is confusing presentation with intelligence.
A dashboard containing dozens of indicators can create the appearance of sophisticated governance while leaving decision-makers unclear about what requires action. Strong performance reporting starts with the decisions leaders need to make and works backwards to the information required.
For a provincial long-term care system, a manageable dashboard might combine measures of population need, access, capacity, workforce, quality and outcomes. National reporting may require a different level of aggregation. Service-level teams need more immediate operational information.
The reporting rhythm matters as much as the measures. Some information requires monthly review; demographic projections may be reconsidered much less frequently. Serious quality or safety signals may require immediate escalation rather than waiting for the next scheduled report.
The Quality Dashboard Builder offers organisations a practical way to structure indicators and reporting arrangements. It does not prescribe measures for Türkiye, but its central discipline is transferable: each measure should have a purpose, an accountable audience and a response when performance changes.
Without that operating rhythm, data collection can become an administrative burden rather than a governance tool.
Local intelligence should travel upwards as well as national policy travels downwards
National ministries need comparable information to understand long-term trends, distribute resources and assess whether policy objectives are being achieved. Yet national datasets can miss the practical reasons behind variation.
Provincial directorates, municipalities, health organisations and service providers encounter patterns before they become visible nationally. They see recurring referral problems, unmet transport needs, changes in family capacity, new workforce pressures and groups who do not engage with existing services.
A learning system needs a route for those observations to influence wider decisions.
This is partly quantitative. Repeated delays, unmet referrals or changing utilisation can be aggregated. It is also qualitative. A service-user narrative may reveal that a formally available service is practically inaccessible. Staff feedback can explain why a referral process generates repeated failure. Family experience can expose hidden workload that administrative data does not record.
Strong intelligence therefore combines statistical evidence with the translation of practice into evidence. The objective is not to allow anecdote to replace data, but to use operational experience to explain what the numbers mean.
Privacy and trust set boundaries around useful integration
Better long-term care intelligence does not require the creation of an unrestricted national record containing every detail about a person’s health, finances, family and daily life.
Türkiye’s data environment operates within legal requirements including the Personal Data Protection Law No. 6698. Health and social-care organisations also have their own professional and administrative responsibilities concerning information.
System integration should therefore be purpose-led. Information needed to coordinate an individual’s care may differ from information required for national planning. Planning data can often be aggregated or de-identified. Staff should not receive access to information merely because technical integration makes access possible.
Good trust, transparency and ethical data use also requires clarity with the public. People should be able to understand, at an appropriate level, why information is collected and how it contributes to care or planning.
This matters particularly as predictive analytics and artificial intelligence become more capable. The ability to infer future risk from large datasets does not remove the need for proportionality, transparency and human judgement.
Predictive intelligence could support earlier planning, but it should not become automated entitlement
Türkiye’s growing data infrastructure creates future opportunities for more anticipatory planning. Demographic projections, service-use trends, workforce information and health data could help identify where demand is likely to grow and where existing capacity may become insufficient.
At population level, predictive modelling can support scenarios. Policymakers could examine how different rates of disability, family availability or community-service expansion affect future demand for formal care. Provincial authorities could identify districts where ageing, workforce constraints and geographic access combine to create higher future pressure.
At individual level, greater caution is required. Predictive models can potentially identify people at increased risk of hospitalisation, functional decline or service instability, but predictions are probabilities rather than facts. Data may reproduce existing inequalities if groups with poor service access are underrepresented in historical records.
The distinction is critical. Predictive intelligence can help direct attention; it should not automatically determine whether a person deserves support.
Organizations testing different capacity assumptions can use the Digital Twin Scenario Modeler to examine interactions between workforce, demand, quality and service stability. Such scenario modelling is an analytical aid rather than a forecast of Türkiye’s national system, but it demonstrates how multiple variables can be considered together instead of extrapolating one trend in isolation.
Scenario: national planners test what population ageing means for actual service demand
A national planning team is preparing a ten-year long-term care capacity outlook. Population projections clearly show a growing older population, but simply applying today’s service-use rate to future population numbers would produce a weak forecast.
The team develops several scenarios. One assumes current patterns of family caregiving continue. Another models reduced availability of unpaid family care as household structures and labour-force participation change. A third assumes stronger prevention and home-based support reduce some demand for residential provision while increasing community workforce requirements.
The scenarios produce different infrastructure and workforce implications even though they use the same demographic projection.
Planners then compare the models with provincial evidence. Areas with high proportions of older residents, population dispersal and lower formal service capacity may require different development from large metropolitan areas where private provision and specialist health services are more available.
No model can determine exactly how many people will require each service ten years ahead. Its value lies in showing which assumptions drive the result. If future capacity depends heavily on families continuing to provide the present volume of unpaid care, that dependency becomes explicit rather than remaining hidden inside the forecast.
Intelligence improves planning not by eliminating uncertainty but by making uncertainty manageable.
Better data can strengthen the relationship between funding and need
Financing decisions become more defensible when they can be related to population need and service outcomes. This is particularly important in Türkiye because long-term care funding currently spans general government expenditure, social assistance, health financing, municipal resources, household contributions and private purchasing rather than operating through one comprehensive long-term care financing mechanism.
Data can help policymakers understand where expenditure is concentrated and how that compares with need. It can reveal whether resources follow demographic pressure, whether some areas rely disproportionately on families and whether expansion in one service reduces or increases demand elsewhere.
It can also improve evaluation of reform. If Türkiye develops new financing arrangements, expands community provision or introduces further long-term care insurance mechanisms, baseline information will be essential. Without knowing current access, need, household burden and outcomes, subsequent change is difficult to evaluate.
This is why information architecture should develop alongside financing reform rather than afterwards. New funding arrangements can create opportunities to standardize reporting and outcome measurement, but they can also create additional data silos if each programme develops independently.
People and families need to be visible in the intelligence system
Administrative data can become institution-centred. It records what ministries fund, what facilities provide and what professionals do. Long-term care, however, exists to support people’s everyday lives.
Türkiye’s intelligence framework should therefore include information about experience and outcomes that matter to people using support. That includes whether they feel listened to, whether services are reliable, whether they can remain connected to family and community and whether support preserves as much independence as possible.
Families also provide information that the formal system may otherwise miss. A person can appear stable in administrative records while a spouse or daughter is providing unsustainable levels of support. Understanding caregiver burden can therefore provide an early signal that the current arrangement may not remain viable.
Qualitative evidence is particularly valuable where formal service use is low. People who never receive a service cannot complete a service-user satisfaction survey. Population research, community engagement and caregiver evidence can reveal unmet need beyond administrative boundaries.
The objective is not to measure every aspect of life. It is to prevent a technically sophisticated intelligence system from knowing everything about service activity while understanding too little about whether support is working for the person.
Türkiye can build a long-term care learning system rather than another data repository
The strategic opportunity is larger than creating a central database. Türkiye can develop a long-term care intelligence function that connects evidence to policy, resource allocation, quality improvement and local service design.
Such a model would require several layers. Nationally, common definitions and indicators could support strategic planning and comparison. Provincial and municipal intelligence could identify local need and access patterns. Providers would retain operational data required to manage quality and capacity. Population surveys would provide visibility of people outside formal services.
The layers would not need to contain identical information. They would need enough alignment to answer shared questions.
Governance would determine how those questions are reviewed. A national indicator showing widening geographic variation should prompt analysis rather than simply publication. Provincial performance showing persistent unmet need should reach decision-makers able to influence capacity. Successful local models should generate evidence strong enough to inform wider development.
This is the difference between performance reporting and an operating intelligence rhythm. Data acquires value when it repeatedly moves through interpretation, decision, action and review.
International experience supports common data principles, not a single model
Countries with mature long-term care systems use different information structures because their financing and administrative arrangements differ. Social-insurance systems can generate standardized claims and assessment data through mechanisms that Türkiye does not currently share. Tax-funded systems may obtain information through local public authorities. Other countries remain fragmented across health, welfare and private provision.
Türkiye therefore does not need to reproduce another country’s database architecture.
The transferable principles are more useful. Long-term care planning benefits from standardized understanding of functional need, visibility of formal and informal care, geographic disaggregation, linkage between capacity and outcomes, and clear governance over how information is used.
WHO’s contemporary long-term care work similarly treats information systems, monitoring and evaluation as core components of a functioning long-term care system. International measurement is increasingly moving towards indicators that compare need with actual receipt of formal long-term care rather than counting services alone.
For Türkiye, the central adaptation is institutional. Better intelligence must connect organisations whose responsibilities remain legitimately different. Integration of information does not require the Ministry of Health, Ministry of Family and Social Services or municipalities to become one organisation. It requires enough shared evidence to understand the same population and coordinate decisions around it.
Conclusion
Türkiye’s long-term care data challenge is not a shortage of information in the narrow sense. Demographic statistics, health records, social-service information, residential-care records, municipal data and household evidence already describe important parts of the system. The strategic gap lies in turning those separate views into a coherent understanding of need, access, capacity, quality and outcomes.
As population ageing accelerates, that intelligence will become increasingly important. Planning based primarily on age or existing service use risks overlooking functional need, unmet demand, geographic inequality and the contribution of unpaid family care. Better interoperability can improve continuity for individuals, while stronger population intelligence can show where workforce, funding and community infrastructure need to develop.
The strongest direction is not unrestricted data centralisation. It is purposeful alignment: common measures where comparison matters, appropriate information exchange where continuity requires it, disaggregated evidence where inequality may be hidden, and governance that connects every important indicator with a decision. Local experience should inform national strategy just as national priorities shape local delivery.
Türkiye’s future long-term care system will ultimately be judged by the support people receive rather than the sophistication of its datasets. The purpose of better intelligence is therefore practical: to see need earlier, understand variation more clearly, allocate capacity more intelligently and learn whether policy is producing better everyday outcomes. Data becomes valuable when it helps the system act.