Measuring Outcomes in Chinese Long-Term Care: Moving Beyond Beds, Visits and Service Volume

A county can report that thousands of elderly-care visits have been completed, a residential provider can show high bed occupancy and a long-term care insurance system can record increasing numbers of reimbursed services. All of those figures matter. None of them, on their own, answers the most important question: are older people safer, more independent, better supported and able to live the lives they value?

This measurement challenge is becoming increasingly important across the China Ageing, Long-Term Care & Community Support Knowledge Hub. As China develops a more coherent long-term care insurance system, strengthens national service standards and expands home, community and institutional support, it is also creating much more information about service activity. The next step is to make that information more useful.

The central policy challenge is not to abandon activity measures. Beds, visits, workforce numbers, eligibility decisions and service expenditure remain essential for understanding capacity and access. The stronger opportunity is to connect those measures with outcomes: changes in function, continuity, preventable harm, family burden, participation, quality of life and whether care intensity remains appropriate as needs change. A mature outcomes system should therefore tell decision-makers not only how much care was delivered, but what difference that care made and where the result was weaker than expected.

China has become better at counting long-term care capacity

China’s elderly-care system has developed extensive measures of infrastructure and service supply.

Local and national reporting can capture numbers of elderly-care institutions, beds, nursing-oriented beds, community facilities, trained workers and people receiving particular forms of support.

Long-term care insurance adds another layer of measurable activity through eligibility assessments, designated service institutions, reimbursed care and fund expenditure.

These measures are indispensable.

A locality cannot manage ageing demand without understanding whether sufficient capacity exists.

The problem arises when activity becomes a proxy for quality.

A filled bed does not show whether the resident maintains function. A completed home visit does not show whether the worker arrived consistently or whether the support reduced family strain. A funded service does not automatically demonstrate that the intervention was appropriate.

The wider principle of quality data, KPIs and performance metrics is therefore increasingly relevant. Measurement needs to describe both what the system does and what happens as a result.

Outputs and outcomes answer different questions

Outputs describe activity.

Outcomes describe change, maintenance or lived experience.

The distinction is simple in principle but difficult in practice because long-term care often aims to prevent deterioration rather than produce rapid improvement.

For somebody recovering after illness, a good outcome may be regained mobility and reduced formal support.

For a person with progressive dementia, maintaining comfort, preventing avoidable injury and preserving familiar relationships may represent success even where function continues to decline.

For somebody with severe permanent disability, the relevant outcome may be stable health, dignity and control over daily life rather than measurable functional improvement.

An outcomes framework therefore needs to recognise that success depends partly on the person’s starting point and likely trajectory.

Function should become one of the core measures of long-term care impact

China already places growing importance on functional assessment when determining eligibility for elderly-care support and long-term care insurance.

That creates a foundation for outcome measurement.

If functional information is collected at the point of assessment and reassessment, local systems can begin to understand whether people improve, remain stable or deteriorate over time.

Functional outcomes might include changes in mobility, eating, dressing, bathing, toileting or other activities of daily living.

However, interpretation matters.

A decline does not automatically indicate poor care.

An older person may have a progressive neurological condition or experience an unavoidable deterioration in health.

The useful question is whether support appears to be preserving ability where reasonably possible, identifying deterioration promptly and adjusting care when the person’s needs change.

Maintaining function can be a successful outcome

Health systems often value improvement because improvement is easier to demonstrate.

Long-term care requires a broader perspective.

For many older people, remaining at the same level of function for twelve months can be a meaningful achievement.

Good care may prevent avoidable deconditioning, support safe mobility and ensure that workers do not unnecessarily take over tasks the person can still perform.

This connects with the broader principle of outcomes-focused and goal-led support.

Measurement should therefore distinguish improvement, maintenance and expected decline rather than treating only improvement as success.

Operational scenario: visit completion hides declining independence

An 81-year-old woman receives publicly supported home care after a period of illness. Her care package includes morning assistance with washing, dressing and breakfast.

The provider reports excellent operational performance. Almost every scheduled visit is completed and there are very few late arrivals.

After six months, however, reassessment shows that the woman is now doing less for herself than when support began.

Further review finds that workers have gradually started completing most morning tasks for her because this is quicker and feels safer.

The visits have been delivered reliably, but the service has unintentionally increased dependence.

The care plan is reframed around maintaining ability. Workers allow more time for the woman to complete parts of washing and dressing independently, provide assistance only where needed and record changes in function rather than simply confirming task completion.

At the next review her overall support remains necessary, but her level of personal participation has stabilised.

The scenario illustrates why output data need an outcome layer. Visit completion demonstrates reliability. It does not demonstrate whether the support is helping the person retain ability.

Care intensity is itself an important outcome measure

One useful way to understand long-term care outcomes is to examine how much assistance a person needs over time.

If rehabilitation, adaptation or better care planning allows somebody to move from intensive support to a lighter package, that change has value for the individual and the wider system.

Conversely, increasing care intensity may be entirely appropriate where dependency has grown.

The measure becomes useful when it is interpreted alongside function.

A sudden increase in support without corresponding evidence of changing need may indicate that a service has become overly dependent on substitution rather than enablement.

A failure to increase support despite clear deterioration may indicate under-provision.

Outcome measurement can therefore help test whether service intensity remains proportionate to assessed need.

Long-term care insurance creates a powerful new measurement opportunity

China’s transition from geographically limited long-term care insurance pilots towards a more coherent national framework significantly strengthens the potential for systematic measurement.

The developing system uses nationally aligned disability assessment, designated long-term care service institutions and a more standardised national service catalogue.

That creates greater consistency around what is being funded.

It also creates a stronger basis for comparing how publicly supported care is delivered across different areas.

Service quality evaluation is becoming increasingly connected with agreements between long-term care insurance administrators and designated care institutions, including links with service payment and continued participation.

This matters because long-term care insurance does more than finance care.

It can create an information architecture through which eligibility, service use, expenditure and quality begin to connect.

The opportunity is to avoid using that architecture solely for payment control.

It can also support understanding of outcomes.

Fund integrity and person outcomes need to be measured together

Long-term care insurance needs strong financial controls because public insurance funds must be protected from inappropriate claims, unnecessary service or fraudulent activity.

But fund integrity and care outcomes should not become separate agendas.

A service can be financially compliant while still producing weak outcomes.

Conversely, a high-cost service may be appropriate where somebody has severe dependency and requires intensive support.

The strongest evaluation asks two related questions.

Was the service legitimately delivered and appropriately funded?

Did the support correspond to the person’s assessed needs and contribute to a reasonable outcome?

This prevents quality evaluation from becoming either purely financial or detached from resource stewardship.

Standardisation can improve comparability without eliminating local variation

China’s elderly-care and long-term care systems have historically developed through considerable local experimentation.

That has allowed innovation but has also made comparison difficult where assessment criteria, service definitions and payment arrangements differ.

Nationally aligned assessment and service standards can improve comparability.

However, outcome expectations still need to account for geographic and population differences.

A rural county with dispersed households and limited specialist capacity should not automatically be expected to produce identical service patterns to a wealthy urban district.

The appropriate comparison is often whether similar groups of people achieve comparable levels of safety, access and independence despite different delivery models.

Case-mix adjustment is essential if outcomes are used to compare providers

Raw outcome data can punish providers that accept people with the greatest needs.

An institution specialising in severe dementia may record more falls, hospital transfers and deaths than a facility supporting mainly independent residents.

A home-care provider working with people discharged after serious illness may show more deterioration than a service focused on light household support.

This does not automatically mean quality is worse.

Outcome comparison therefore needs some understanding of starting need, including functional dependency, cognition, health complexity and other relevant risk factors.

The objective is not to create a perfect statistical model before any comparison is possible.

It is to avoid presenting unadjusted numbers as though providers serve equivalent populations.

Person-reported outcomes add something administrative data cannot

Administrative systems can measure many things efficiently.

They cannot fully determine whether somebody feels respected, has meaningful choice or remains connected to family and community.

Those outcomes require the person’s own perspective wherever possible.

Questions can explore whether support allows the older person to continue routines they value, whether workers communicate respectfully and whether the person feels involved in decisions.

This aligns with the wider theme of service-user feedback and co-production.

Person-reported information should not replace objective measures of safety or function.

It completes the picture.

A service can be clinically safe but experienced as highly restrictive. Another may be warm and relational while failing to manage serious health risks. Good quality requires visibility of both.

Outcome measurement needs alternatives when people cannot report easily

Some older people cannot participate in conventional surveys because of dementia, communication difficulty or severe illness.

That should not make their experience invisible.

Providers can use adapted communication, observation, family input and other evidence to understand whether the person appears comfortable, engaged and consistently supported.

Proxy reporting needs care because relatives may have different preferences from the older person.

The aim should therefore be to gather the best available evidence without automatically replacing the person’s perspective with somebody else’s.

Outcome systems become more inclusive when they allow several routes through which experience can be understood.

Family outcomes deserve measurement because long-term care changes family life

China’s elderly-care system continues to depend substantially on families.

Formal services therefore create effects beyond the individual receiving care.

A reliable home-care package may allow an adult daughter to remain in employment. Respite may reduce exhaustion for an older spouse. Residential care may reduce physical burden while creating new emotional or financial pressures.

Family impact should not become the primary measure of whether care is appropriate, because the older person remains central.

But ignoring family outcomes can hide an important part of system performance.

Measures of carer burden, confidence and ability to continue caring can help local systems understand whether formal services are genuinely supporting the wider care network.

Hospital use can provide a useful outcome signal if interpreted carefully

Unplanned hospital admissions, emergency visits and repeated transfers can indicate gaps in long-term care.

A service that recognises deterioration early, manages medication well and coordinates effectively with healthcare may prevent some avoidable escalation.

However, low hospital use should never become an objective in isolation.

Some older people need hospital treatment and should receive it promptly.

A provider could create apparently favourable data by delaying appropriate transfer.

The more meaningful measure therefore examines patterns: repeated potentially preventable admissions, rapid readmission after discharge or transfers associated with recurring failures in the same area of care.

Operational scenario: low hospital transfer rates create false reassurance

An elderly-care institution reports fewer hospital transfers than other facilities in the same city and initially presents this as evidence of strong clinical management.

A deeper review examines residents who later required urgent treatment.

Several cases show that staff had observed deterioration for many hours before healthcare input was sought.

The institution’s low transfer rate therefore reflects two different factors: some genuine prevention, but also a culture in which staff are reluctant to escalate because hospital transfers are viewed as evidence of service failure.

Management changes the indicator.

Instead of targeting lower transfers, the provider reviews whether escalation was appropriate, timely and consistent with the person’s condition and wishes.

Repeat hospital use is examined separately to identify potentially preventable patterns.

The quality conversation becomes more sophisticated.

The scenario illustrates a central risk in outcome measurement: once a metric becomes a target, behaviour may shift towards improving the number rather than improving care.

Outcome measures need balancing indicators

No single indicator should define long-term care quality.

Reducing falls could be achieved by preventing people from walking. Reducing hospital use could be achieved by delaying escalation. Reducing care hours could reflect successful reablement or unsafe under-support.

Outcome frameworks therefore need balancing measures.

For example, lower falls rates can be interpreted alongside mobility and participation. Reduced support intensity can be considered alongside functional status and satisfaction. Lower hospital use can be examined alongside deterioration, mortality and clinical review.

This prevents one objective from being pursued at the expense of another.

Organisations examining comparable performance systems can use the Quality Dashboard Builder to structure balanced sets of quality, workforce and outcome indicators. It is not a Chinese long-term care measurement framework, but the principle of avoiding single-metric governance is directly relevant.

Measurement should support care decisions, not merely retrospective reporting

Outcome data create the greatest value when they influence decisions while care is still being delivered.

If functional decline, weight loss, repeated falls or increased family concern become visible only in an annual report, the measurement system is too slow to improve the person’s experience.

Providers need ways of identifying meaningful change at operational level and escalating it for review.

The principle is simple: measurement should shorten the distance between a change in need and a change in support.

This is particularly important in long-term care because deterioration can be gradual. A small reduction in mobility, appetite or participation may not appear serious on its own. A pattern across several weeks can indicate that the current plan no longer fits.

Provider dashboards should connect quality, workforce and outcomes

Long-term care outcomes rarely have a single cause.

An increase in falls may relate to frailty, medication, environment, staffing, supervision or poor recognition of changing mobility. Increased hospital transfers may reflect deteriorating health, weak care coordination or an appropriate response to a more complex resident population.

Providers therefore need to see outcome data alongside the operational conditions that may explain them.

A useful management view might connect:

• functional change and care intensity;
• falls, pressure injury and other safety indicators;
• hospital use and care transitions;
• workforce turnover, continuity and skill mix;
• complaints, experience and family feedback;
• service utilisation and expenditure.

The purpose is not to create a large collection of indicators for its own sake.

It is to help leaders see whether several weak signals are pointing towards the same underlying problem.

Long-term care insurance is moving towards more structured service-quality evaluation

China’s developing long-term care insurance framework increasingly connects designated service providers with formal service agreements, information requirements and quality evaluation.

That creates an important accountability mechanism.

Designated long-term care service institutions need professional service teams, internal management arrangements and the ability to connect with healthcare-security information systems. Quality evaluation can then influence ongoing provider management and the relationship between service performance and payment.

The opportunity is significant because the insurance system can create a more consistent flow of information across home, community and institutional long-term care.

However, quality evaluation should avoid becoming focused exclusively on whether claims were technically correct.

Financial integrity is essential, but outcome measurement can add another question: whether reimbursed care remains appropriate to the person’s assessed dependency and produces reasonable results.

A nationally defined service catalogue improves the measurement baseline

The establishment of a national long-term care insurance service catalogue helps create greater clarity about the types of services that can be supported through the system.

This strengthens measurement because areas are increasingly working from a more common definition of long-term care activity rather than entirely different local service descriptions.

Standardisation does not remove the need for local implementation.

Service availability, payment arrangements and provider capacity still vary.

But a more consistent service vocabulary makes it easier to examine how different forms of support relate to outcomes.

For example, systems can begin asking whether particular combinations of home-based care, rehabilitation-oriented support or institutional care are associated with different trajectories for people with comparable functional needs.

This kind of analysis should be used cautiously, but it can gradually strengthen policy design.

One-person-one-record approaches can strengthen longitudinal outcome measurement

Long-term care becomes easier to understand when information follows the person over time rather than being fragmented into separate service episodes.

China’s designated long-term care provider arrangements increasingly emphasise electronic service records for individual beneficiaries.

This creates the possibility of longitudinal measurement.

Instead of asking only what happened during a particular visit, the system can examine how the person’s condition, service use and care intensity changed across months or years.

The wider principle of digital records and information governance is therefore relevant.

Longitudinal records can help identify whether care is stabilising need, whether reassessment is happening at the right time and whether people repeatedly move between services without continuity.

They also create stronger requirements around data quality, privacy and access.

Interoperability matters because outcomes cross organisational boundaries

An older person’s long-term care outcome may depend partly on events outside the elderly-care provider.

Hospital admission can change mobility. Primary-level healthcare may alter medication. Rehabilitation may reduce dependency. Family circumstances may change the amount of informal care available.

If every service holds an isolated record, the outcome story becomes fragmented.

The broader challenge of interoperability and system integration therefore matters for measurement as well as care coordination.

Not every organisation needs unrestricted access to every record.

But relevant information needs to move sufficiently for decisions to make sense.

A long-term care provider cannot interpret increased dependency accurately if it does not know that the person recently experienced a stroke or hospital admission.

Outcome measurement can reveal weaknesses at transitions

Transitions are a useful place to examine system performance because they involve several organisations and often expose gaps that routine service measures miss.

An older person may leave hospital with reduced mobility and enter home care. If function continues declining because rehabilitation never starts, each organisation may still report that it completed its own activity.

The hospital completed discharge. The home-care provider completed visits. The insurer funded eligible support.

Yet the overall outcome is weak.

Measurement across transitions therefore needs to ask what happened after the handover, not merely whether the handover occurred.

Operational scenario: every organisation meets its activity target but the person deteriorates

A 76-year-old man returns home after hospital treatment for a fractured hip. He is eligible for long-term care support because his function has reduced significantly.

Home-based care begins promptly and scheduled visits are delivered reliably.

The hospital discharge is recorded as successful, and the long-term care provider reports full service completion.

Three months later, reassessment shows that the man remains almost entirely dependent for transfers and rarely leaves his chair.

A review finds that responsibility for continued rehabilitation was never clearly connected with the long-term care plan. Care workers have been safely helping him transfer, but the overall system has focused on supporting dependency rather than restoring function where possible.

The response is not to criticise the workers for completing the agreed care.

Instead, the local pathway is redesigned so that post-discharge functional goals, rehabilitation input and long-term care intensity are reviewed together for similar cases.

The scenario demonstrates why outcomes can expose fragmentation that activity measures conceal. Every component can perform its assigned task while the person still experiences a poor overall trajectory.

Reassessment should become an outcome checkpoint

Functional reassessment is often treated primarily as an eligibility or funding process.

It can also provide valuable outcome information.

When a person is reassessed, the system can examine whether their condition changed, whether care intensity changed appropriately and whether the original goals remain relevant.

This is particularly useful where long-term care insurance eligibility is linked to functional impairment.

The same assessment architecture that determines access can therefore help understand change over time.

Care is then less likely to become an indefinite package that continues without questioning whether the level or type of support still fits.

Outcome measurement should identify both under-support and over-support

Long-term care quality can fail in opposite directions.

Under-support leaves people without enough assistance to remain safe.

Over-support can unnecessarily reduce independence and increase cost.

Outcome data can help identify both.

Repeated falls, weight loss or family exhaustion may indicate insufficient support.

Stable function combined with steadily increasing care hours may justify review of whether workers are doing too much for the person.

The wider principle of just enough support and least restrictive practice is relevant because good long-term care should provide the amount of assistance required without automatically replacing abilities that remain.

Equity outcomes need to show who is not benefiting equally

National averages can conceal substantial variation.

China’s urban and rural areas differ in provider supply, transport, workforce availability, fiscal capacity and access to specialist services.

Outcome measurement should therefore be capable of identifying whether people with comparable needs experience systematically different results depending on where they live.

This does not mean every locality needs to deliver care in the same way.

Rural models may use different combinations of county institutions, township services, village support and family involvement.

The relevant question is whether geographic variation produces avoidable differences in safety, access, continuity or independence.

Access is itself an outcome of system design

A service can demonstrate strong outcomes for the people who receive it while still operating within a system where many eligible people cannot obtain support.

Outcome measurement therefore needs an access dimension.

Useful questions include whether eligible people can find a designated provider, how long services take to begin and whether certain geographic areas consistently have weaker provider coverage.

This becomes increasingly important as long-term care insurance expands.

Formal entitlement has limited value if service capacity is unavailable locally.

Rural outcomes need to be interpreted through geography

Distance can shape long-term care outcomes in ways that are less visible in urban systems.

A rural provider may deliver fewer direct contacts because workers travel further. Specialist input may require remote consultation or periodic county-level visits. Families may provide a larger share of daily support because formal services are thinner.

Outcome measurement should therefore avoid rewarding only high service volume.

A lower-volume rural model may still achieve good results if people remain safe, family burden is manageable and escalation is reliable.

The stronger comparison focuses on what is achieved with the available model rather than assuming identical service intensity is always preferable.

Family burden should be interpreted carefully alongside formal-care expansion

One objective of a more developed long-term care system is to reduce unsustainable reliance on unpaid family care.

However, a simple fall in family care hours is not always the right outcome.

Some families want to remain deeply involved.

The relevant distinction is between chosen involvement and unsupported obligation.

Outcome measures can therefore examine whether relatives feel able to continue their role, whether care interferes excessively with employment or health and whether formal services provide dependable relief where needed.

This creates a more nuanced measure of how long-term care redistributes responsibility between families and formal systems.

Provider comparison should support improvement rather than simplistic ranking

Comparative data can be powerful.

A provider that sees substantially higher falls, turnover or hospital transfers than comparable services has a reason to investigate.

But public or administrative ranking can distort behaviour if providers focus on protecting scores rather than understanding outcomes.

Case mix, service type and reporting culture all matter.

The stronger use of comparison is diagnostic.

It identifies unusual variation and prompts questions.

It does not assume the number itself explains the cause.

Low incident rates can sometimes indicate weak reporting

Outcome and safety data depend heavily on reporting culture.

An organisation with very few recorded incidents may be exceptionally safe.

It may also discourage staff from recording events.

The same applies to complaints.

A provider with no complaints could have highly satisfied users, or people may not understand how to complain or may fear consequences.

Measurement therefore needs triangulation.

Incident levels should be considered alongside staff feedback, family experience, inspection evidence and other quality information.

A mature system avoids assuming that the lowest number is automatically the best number.

Outcome indicators can influence provider behaviour, so incentives need careful design

Once performance measures are linked with payment, continued designation or public reputation, organisations have a strong incentive to improve those measures.

That can be beneficial.

It can also create gaming or unintended behaviour.

If maintaining low care intensity is rewarded, providers may hesitate to increase support when needs deteriorate. If hospital avoidance becomes a target, escalation may be delayed. If functional improvement is heavily rewarded, providers may prefer people with greater recovery potential.

Payment and quality systems therefore need a balanced set of indicators and professional review.

No metric should create an incentive to deny appropriate care.

Outcome governance needs to ask why variation persists

Measurement becomes valuable when somebody is responsible for responding to what it shows.

A local system may identify that one district has consistently higher hospital transfers, weaker continuity or slower service initiation than others.

The governance question is what happens next.

Leaders need to determine whether the difference reflects population need, workforce shortages, provider practice, geography or another factor.

If variation persists despite improvement activity, the response may need to move from provider-level correction towards service redesign or resource reallocation.

Organisations considering comparable assurance questions can use the Governance Maturity Assessment to examine whether data, accountability and escalation are sufficiently connected. It is not a China-specific governance tool, but the principle is relevant.

Outcome data should support local planning as well as provider oversight

Long-term care outcomes can reveal more than the performance of an individual organisation.

They can also show whether the wider service system is configured appropriately.

If several providers in the same county report rising dependency, repeated delayed discharges or growing family strain, the underlying issue may not be provider performance. It may indicate insufficient rehabilitation capacity, weak community services, workforce shortages or a mismatch between local demand and available support.

This is where outcome measurement becomes strategically valuable.

Local governments can use aggregated evidence to understand whether investment is producing the intended effect across a population rather than examining each service in isolation.

Patterns can help identify where additional home-care capacity is needed, where community rehabilitation should be strengthened or where institutional services are supporting people whose needs could potentially be met differently.

The stronger system therefore treats outcomes as intelligence for planning, not simply as evidence for judging providers.

Measurement needs to connect with resource allocation

Outcome evidence becomes particularly powerful when it informs how limited resources are distributed.

A locality may spend heavily on institutional capacity while outcome evidence shows that many people entering residential care have moderate needs that could potentially have been supported at home with better community infrastructure.

Another area may invest extensively in home care while repeated hospital admissions show that people with high medical complexity lack sufficiently coordinated healthcare input.

Neither pattern can be understood through expenditure alone.

Outcome evidence helps decision-makers ask whether money is producing the right type of capacity.

This does not mean funding should automatically move towards whichever service produces the lowest cost per person.

High-dependency care will remain expensive.

The relevant question is whether the level and type of spending correspond to need and whether alternative investment could produce stronger independence, safety or continuity.

Outcome measurement should influence provider improvement before funding withdrawal

Where performance is weak, the first response should not always be immediate exclusion from publicly supported service arrangements.

Some providers can improve if the underlying problem is identified clearly and corrective action is monitored.

Outcome evidence can help distinguish an isolated poor result from a persistent pattern.

A provider with an unexpected rise in falls may need to review staffing, medication and mobility support. A repeated pattern across several reporting periods, combined with weak corrective action, raises a more serious accountability question.

This is where the wider principle of quality improvement plans and action tracking becomes relevant.

Improvement should have clear ownership, timescales and evidence of whether the underlying outcome has changed.

Operational scenario: outcome data trigger service redesign rather than provider blame

A city district notices that several home-care providers serving older people with moderate functional impairment show an unusually high rate of increasing care hours after the first six months of support.

The initial assumption is that providers may be creating unnecessary dependency.

A closer review compares assessment records, service plans and local rehabilitation capacity.

It finds that most providers are responding appropriately to increasing need, but people rarely receive short-term rehabilitation or reablement after illness before long-term care packages are expanded.

The common problem therefore sits upstream of individual provider behaviour.

The district introduces a stronger pathway linking functional reassessment with access to rehabilitation for people whose dependency has recently increased.

Providers continue to increase care where necessary, but they also gain a route for testing whether some lost function can be restored before support becomes permanently more intensive.

Over time, the district monitors both care intensity and functional outcomes.

The scenario demonstrates why measurement should lead to diagnosis before judgement. Unusual provider data can reveal a system weakness rather than poor provider behaviour.

Outcome measures should be co-designed with the people whose lives they describe

Professionals naturally tend to measure what systems can record easily.

Older people may value different things.

A service may focus on falls, hospital admissions and activities of daily living while the person cares most about being able to visit a local market, prepare tea independently or remain in the same neighbourhood.

Both perspectives matter.

System measures need enough consistency for comparison, but individual outcomes should also reflect personal priorities.

This connects with the wider principle of co-production, choice and control.

At service level, older people and families can help identify which dimensions of quality matter most. At individual level, care planning can translate broader outcome domains into personally meaningful goals.

Individual goals and population indicators need to coexist

Long-term care systems need standardised measures because governments and insurers cannot plan from thousands of entirely different personal goals.

At the same time, standardisation should not reduce every person to the same outcome template.

The solution is a layered approach.

Population indicators can examine broad domains such as function, safety, service intensity, hospital use, continuity and experience.

Individual care plans can then define what success means within those domains for a particular person.

For one person, maintaining the ability to walk to the bathroom independently may be the key functional goal. For another, the most important outcome may be continuing to participate in family meals despite severe mobility limitation.

The system measure and the personal goal answer different but complementary questions.

Outcome evidence can strengthen accountability to families without transferring responsibility to them

Families often want to know whether formal care is making a meaningful difference.

Providers can support that accountability by explaining changes in function, risk, care intensity and agreed goals rather than communicating only when something goes wrong.

This can strengthen trust.

However, families should not be expected to become informal auditors of service quality.

Formal providers and public authorities retain responsibility for assuring the services they deliver or fund.

Family feedback is valuable evidence, not a substitute for governance.

Measurement needs to recognise cultural and regional differences in expectations

China’s size and diversity mean that expectations of family involvement, community support and formal services vary across regions and households.

An outcome framework should therefore avoid assuming that one model of independence fits everyone.

For some older people, living with adult children may be the preferred outcome. For others, maintaining a separate household may be central to autonomy.

The goal of measurement should not be to impose a uniform lifestyle.

It should assess whether the person receives sufficient support to live in a way that is safe, dignified and consistent with their preferences where reasonably possible.

Digital measurement can improve timeliness but also create surveillance risk

Digital care records, sensors and connected devices can generate increasingly detailed information about older people’s routines and service use.

This can support earlier identification of deterioration.

Changes in mobility, sleep, medication use or service engagement may become visible before a serious event occurs.

But greater measurement is not automatically better.

Continuous monitoring can become intrusive if data are collected without a clear purpose or if older people have little understanding of how information is used.

The principle of digital safeguarding and technology-enabled risk is therefore relevant.

Outcome measurement should remain proportionate.

Systems need enough information to improve care without treating older people’s homes and daily lives as unlimited data environments.

Artificial intelligence may help identify outcome patterns, but governance remains essential

As data volumes increase, artificial intelligence and advanced analytics may eventually help identify patterns that are difficult to detect manually.

Systems might highlight people whose functional decline, service intensity and hospital use suggest increasing risk.

They may also identify providers whose outcome patterns differ substantially from comparable organisations.

These applications remain emerging and require caution.

Algorithms can reproduce poor assumptions or amplify weaknesses in the underlying data.

A prediction should therefore support professional review rather than automatically determine service reduction, eligibility or provider sanction.

Outcome governance remains a human responsibility even when analytical tools become more sophisticated.

Goodhart’s law is a practical risk in long-term care measurement

When a measure becomes a target, organisations may begin optimising the number rather than the outcome it was intended to represent.

Long-term care is particularly vulnerable because many outcomes are complex and influenced by factors beyond provider control.

If low care intensity is rewarded, services may resist increasing support. If high independence scores are prioritised, providers may avoid people with severe dependency. If complaint numbers are treated negatively, organisations may discourage people from raising concerns.

The solution is not to abandon measurement.

It is to use balanced indicators, professional interpretation and regular review of unintended consequences.

Measures should remain servants of quality rather than becoming the definition of quality.

National reporting should focus on trends that support system learning

As China develops a more coherent long-term care framework, national reporting can help identify broad patterns across provinces and localities.

This may include variation in functional outcomes, access, service intensity, provider capacity or family burden.

The purpose should be to understand where implementation is producing different results and why.

National comparison is particularly useful when it identifies structural questions.

Why do some areas achieve stronger home-care continuity? Why does one region rely much more heavily on institutional provision? Why do similar beneficiaries receive very different levels of formal support?

Those questions can inform policy refinement without assuming that every variation represents failure.

Outcome frameworks need stability as well as evolution

Measurement systems lose value if definitions change so frequently that trends cannot be followed over time.

China therefore needs enough stability in core outcome domains to build longitudinal evidence.

At the same time, the framework should evolve as the long-term care system develops.

New service models, technologies and funding arrangements may require additional measures.

The strongest approach maintains a stable core while allowing selected indicators to develop around emerging priorities.

This helps decision-makers distinguish genuine change in outcomes from change produced simply by altering the measurement method.

The 15th Five-Year Plan period creates an opportunity to embed outcomes into system development

China’s 2026–2030 policy direction is moving elderly care towards greater national coherence, stronger functional assessment, wider long-term care insurance coverage, more community provision and improved service quality.

This creates a particularly important opportunity to build outcome measurement into the system as it develops rather than adding it later.

If assessment, service records, insurance claims and provider-quality information become increasingly standardised, they can support a more integrated view of what happens to people over time.

The strongest model would connect:

• initial functional assessment and eligibility;
• the type and intensity of services received;
• changes in function and wellbeing;
• hospital use and major safety events;
• family and person-reported experience;
• reassessment and subsequent service adjustment.

That does not require one national database containing every detail of an older person’s life.

It requires sufficient interoperability and common definitions for relevant evidence to inform decisions.

Outcome measurement should become part of everyday improvement culture

The most mature use of outcomes occurs when frontline teams understand why the measures matter.

If staff experience measurement only as an external reporting requirement, data quality is likely to be weaker and learning limited.

Workers should be able to see how information about falls, function, continuity or participation connects with the care they provide.

This helps convert measurement from bureaucracy into practice intelligence.

The broader principle of embedding learning into day-to-day practice is therefore directly relevant.

Outcome information becomes most valuable when teams use it to ask what they should do differently tomorrow, not only what they need to report next month.

What China’s outcomes transition offers international systems

China’s long-term care system is developing within administrative, financing and cultural conditions that differ from those of many other countries. Its emerging measurement architecture therefore should not be copied mechanically.

The underlying lessons are widely relevant.

First, service volume is necessary to understand capacity but insufficient to demonstrate value.

Second, long-term care outcomes need to recognise maintenance and expected decline as well as improvement.

Third, functional assessment can become a powerful longitudinal tool when it is connected with reassessment and service use.

Fourth, provider comparison requires context because dependency and case mix strongly influence outcomes.

Fifth, insurance systems can create useful measurement infrastructure, but financial control should not displace human outcomes.

Sixth, person and family experience adds information that administrative data cannot supply.

Finally, outcome evidence is most powerful when it leads to diagnosis, improvement and system redesign rather than simplistic ranking.

Conclusion

China’s long-term care system is becoming increasingly measurable, but the most important next step is to make measurement more meaningful. Beds, facilities, visits, insurance claims and service expenditure all provide essential information about capacity and activity. They do not, by themselves, show whether older people remain safer, more independent, better supported or able to live with dignity.

A stronger outcomes framework should connect functional change, care intensity, safety, hospital use, continuity, family burden and lived experience. It should recognise that maintaining ability can be a positive outcome, that deterioration is not automatically evidence of poor care and that provider comparisons need to reflect dependency and local context. Long-term care insurance, digital records and more consistent assessment standards create important opportunities to build this evidence longitudinally.

The decisive governance question is what happens once variation becomes visible. Outcome data should lead to review, improvement and better resource allocation rather than existing primarily for reporting or ranking. Where several providers show the same weakness, the system itself may need redesign.

During the 15th Five-Year Plan period, China has an opportunity to move from measuring how much long-term care it provides towards understanding what that care achieves. That transition would strengthen accountability while keeping the purpose of the system clear: not simply to deliver more services, but to support longer lives with as much function, autonomy, safety and connection as possible.