Digital Health and Connected Care for Older People in Hong Kong

An older person leaves hospital after treatment for heart failure. Their medication has changed, mobility has reduced and follow-up is needed. The hospital holds detailed clinical information. A family doctor may manage ongoing health needs. An RCHE or home-support service sees what happens day to day. Family members notice changes that occur between formal appointments. The quality of the next stage depends not only on whether each party has information, but on whether the right information can reach the right person at the right time.

This is the central digital challenge within the Hong Kong Ageing, Long-Term Care & Community Support Knowledge Hub. Hong Kong already has substantial digital-health infrastructure, including its territory-wide Electronic Health Record Sharing System, commonly known as eHealth, alongside continuing development towards more connected digital healthcare. At the same time, residential homes, community organisations, private providers and families use a wide range of their own systems, devices and records.

The opportunity is significant. Better-connected information can reduce repeated history-taking, make medication changes more visible, support earlier recognition of deterioration and strengthen coordination between hospital, primary care and long-term care. But connection is not the same as integration. A technically available record may still omit the functional, social or day-to-day information another service needs. Digitalisation can also create new risks around privacy, cyber resilience, workforce capability and exclusion. For older people, the test is therefore not how many systems become digital. It is whether technology makes care more coherent, timely and understandable without replacing human judgement or personal choice.

Connected care begins with recognising that older people generate information across several systems

Older people with complex needs rarely have one complete record.

Clinical information may sit within Hospital Authority systems, primary-care records and eHealth. Residential services hold care plans, observations, medication administration information and incident records. Community providers may record visits, functional changes and family concerns. The older person and relatives may hold appointment information, paper documents or their own digital records.

Each source reflects a different purpose.

A hospital record is designed primarily around clinical care. An RCHE record needs to support continuous daily care. A home-support worker may need to know what has changed since discharge but not every detail of an inpatient episode.

This is why the broader interoperability and system-integration agenda matters. Connected care is not achieved simply by placing more information into one electronic environment. It requires enough shared structure that organisations can find and interpret the information needed for their own decisions.

Hong Kong’s eHealth infrastructure creates an important foundation

The Electronic Health Record Sharing System provides a territory-wide infrastructure through which participating healthcare providers can share specified electronic health information with appropriate patient participation and access arrangements.

For an ageing population, this foundation is particularly valuable because older people frequently receive care from more than one healthcare provider.

A person may use Hospital Authority services, attend primary care and receive follow-up from other healthcare professionals. Shared information can reduce fragmentation and support continuity when care moves between settings.

But eHealth should not be interpreted as a complete long-term care record.

Much of what determines whether an older person remains independent sits outside conventional clinical information: whether they can transfer safely, whether a spouse can still provide care, whether medication is actually being taken correctly at home or whether cognitive change is affecting everyday routines.

The stronger opportunity therefore lies in connecting clinical infrastructure with wider care pathways without assuming that health information alone describes the person’s needs.

Health records and care records answer different questions

A clinician may need diagnosis, pathology results, medication and treatment history.

A residential-care worker may need to know that the person now requires assistance when standing, becomes breathless after walking a short distance and has been advised to monitor fluid balance.

A social worker may need to understand whether family support remains sustainable.

These are all legitimate parts of one care situation, but they are not interchangeable.

Digital strategy should therefore resist the idea that integration means every professional reading the same record.

The objective is better information continuity between roles while preserving the distinct information each service needs to perform safely.

Operational scenario: discharge information is clinically complete but operationally incomplete

An 83-year-old woman returns to an RCHE after hospital treatment for pneumonia. Her discharge information records the clinical diagnosis and revised medication clearly.

During the first evening, however, staff discover that she now becomes unsteady when transferring from bed and is substantially weaker than before admission. Her daughter says she required assistance to walk on the ward during the final days of admission, but this functional change is not prominent in the information available to the home.

The problem is not necessarily missing clinical documentation.

It is that the receiving service needs a different type of information to organise everyday care safely.

The RCHE responds by increasing assistance temporarily and arranging further review. Management also examines its discharge-return process so that significant changes in mobility, cognition, continence, nutrition and assistance needs are explicitly checked rather than assumed from the medical record.

The digital lesson is important. Interoperability should be judged by whether information supports the next decision, not simply by whether data moved successfully between systems.

Digital care planning can make changing needs more visible

Electronic care records can strengthen long-term care when they make trends easier to recognise.

A worker recording reduced appetite, greater assistance with transfers or repeated night-time distress may see only one episode. A digital record viewed longitudinally can reveal that the pattern has developed over several weeks.

This is one of the strongest opportunities within digital care planning.

The value lies not in replacing paper with screens, but in improving visibility.

Good digital records can support:

• clearer chronology of changing needs;
• faster access to current care information;
• more consistent handover between workers;
• better identification of recurring incidents or deterioration;
• easier review of whether planned interventions are working; and
• stronger continuity when staff or services change.

The benefit disappears, however, if recording becomes so burdensome that workers spend less time observing the person or begin entering minimal information simply to complete mandatory fields.

Digital transformation should remove duplication rather than reproduce it electronically

Long-term care organisations often accumulate documentation over time.

A new field is added after an incident. Another checklist follows an inspection. A separate spreadsheet is introduced for management reporting. Eventually the same information may be recorded in several places.

Moving those processes onto a digital platform does not automatically improve them.

If a care worker still enters the same information three times, the organisation has digitalised duplication.

A stronger digital approach begins with the workflow.

What information is genuinely needed? Who uses it? Can one reliable entry support several legitimate purposes? Which controls exist only because older systems could not communicate?

The Digital Transformation Readiness Assessment can help organisations structure similar questions around digital strategy, workforce capability, information governance and operational readiness. It is not a Hong Kong regulatory instrument, but its underlying principle is relevant: technology should redesign unnecessary work rather than merely transfer it onto a screen.

Connected records are valuable only when data quality is reliable

Digital systems can spread poor information more efficiently as easily as they can spread good information.

An incorrect medication entry, outdated telephone number or old mobility status becomes more consequential when several organisations rely on it.

This makes data quality and reliable metrics a core part of connected care.

Services need clear responsibility for keeping important information current.

Not every field requires constant updating, but information affecting immediate safety, treatment or care needs should have an identifiable source and review process.

Connected care therefore increases rather than reduces the importance of information governance.

More information can create new clinical and operational risk if nobody knows what is current

One of the paradoxes of digital integration is that access can become easier while interpretation becomes harder.

A professional may see several medication lists created at different points in time. A care worker may find an older assessment alongside a newer care plan. Family information may conflict with the formal record.

The system needs ways to distinguish current information from historical information.

This is especially important during transitions, when many details change at once.

Connected systems should therefore support chronology and provenance: when information was recorded, by whom and in what context.

Without that clarity, more data can increase ambiguity.

Operational scenario: one person has three versions of the same medication list

An older man with diabetes, hypertension and heart failure attends several healthcare services and receives home support.

After a specialist review, one medicine is discontinued and another adjusted. The family receives an updated list. The home-support worker continues to see an older list kept in the flat, while another version appears within a separate service record.

No individual has acted negligently. The problem is version control.

The worker notices the discrepancy and escalates it rather than trying to determine independently which prescription is correct. The family is asked to replace the outdated paper copy once the current regimen is confirmed through the appropriate healthcare route.

The provider then reviews how medication changes are incorporated into its own records so that old versions are clearly superseded.

The scenario demonstrates why digital connection alone cannot guarantee medication safety. Information needs an authoritative current state and a clear process for resolving conflicts.

Primary healthcare creates an important bridge between digital health and prevention

Hong Kong’s direction of travel towards stronger primary healthcare increases the importance of longitudinal information outside hospitals.

Older people living with diabetes, hypertension, chronic respiratory disease or other long-term conditions benefit when deterioration can be identified before emergency treatment becomes necessary.

Digital health can support that objective by improving access to relevant history, test information and follow-up.

But prevention also depends on information that may be generated outside formal clinical encounters.

A community worker notices reduced appetite. A daughter reports increasing confusion. An RCHE records repeated breathlessness during ordinary activity.

Connected care becomes more powerful when these observations can trigger appropriate clinical attention rather than remaining isolated within social or residential care records.

Information sharing should follow a defined care purpose

The case for greater connectivity does not justify unrestricted access to personal information.

Older people’s records can include highly sensitive health, family, mental-health and social information.

The useful governance question is therefore not “How can everybody see everything?”

It is “What does this professional or service need to know to perform this particular role safely?”

This principle supports both privacy and usability.

Too little access can fragment care. Excessive access can intrude unnecessarily and make systems harder to navigate.

Role-based information design is therefore central to mature interoperability.

Consent and participation need to remain visible in digital health

Older people should understand, as far as reasonably possible, how their information is being used and shared within applicable arrangements.

Digital systems can make consent feel abstract because information moves invisibly.

A person may understand that a doctor is treating them but be less clear about which information another organisation can access electronically.

Good digital communication therefore needs plain explanation rather than assuming that enrolment or technical authorisation automatically produces meaningful understanding.

This becomes particularly important for people with cognitive impairment, where participation and decision-making ability may vary according to the issue.

Families are important information partners but should not become unofficial data controllers

Family members often hold vital information.

They know which medicines the person actually takes, what changed after discharge, which symptoms have developed and how the person functioned before illness.

Digital tools may make family involvement easier through appointment information, messaging or shared monitoring where appropriate.

But family access should reflect the older person’s wishes and lawful arrangements rather than becoming automatic simply because relatives provide care.

An older person may want a daughter involved in appointments while preferring some information to remain private.

Connected care should strengthen chosen family partnership without assuming that dependence eliminates confidentiality.

Digital inclusion is part of care quality, not a separate technology policy

A digital system can make access easier for many people while creating a new barrier for others.

Older adults vary widely in digital confidence, eyesight, dexterity, language, cognition and access to devices.

Some use smartphones and online banking confidently. Others depend on family support or prefer telephone and face-to-face contact.

The wider digital inclusion agenda therefore needs to remain central to Hong Kong’s connected-care strategy.

A digital-first pathway should not silently become a digital-only pathway where essential access is concerned.

The strongest model uses technology to widen access while maintaining realistic alternatives for people who cannot or do not wish to use it independently.

Operational scenario: digital access improves convenience for most users but excludes one group

A community service introduces online appointment and reminder functions to reduce telephone workload. Uptake is strong among many older people and families, and missed appointments fall.

Staff later notice that a smaller group of older users continues telephoning repeatedly because they cannot navigate the digital process. Several rely on older mobile phones, while others struggle with written instructions.

The service could treat these users as exceptions who need to adapt.

Instead, it maintains telephone support for those who need it and simplifies the digital instructions. Staff also ask whether a trusted family member should be involved where the older person wants that support.

The digital service remains the preferred route for many users, but the operating model recognises that equal access does not require everybody to use the same channel.

The lesson is that successful digital transformation should reduce friction without transferring that friction onto people with lower digital confidence.

Workforce capability is one of the main determinants of whether digital care works

Digital transformation changes jobs.

Care workers need to record information accurately and understand alerts. Nurses may need to review information across several systems. Managers need to interpret dashboards rather than accept them unquestioningly. Social workers and community teams may increasingly work with digital information generated outside their own organisation.

This creates training needs beyond basic system navigation.

Workers need to understand data quality, confidentiality, cyber risk and how digital information should influence professional judgement.

The broader digital skills and workforce adoption agenda is therefore fundamental. A sophisticated system used inconsistently by staff can create more risk than a simpler system that everybody understands.

Digital competence should be role-specific

Not every employee needs the same depth of digital knowledge.

A frontline worker may need to record observations, identify alerts and escalate concerns. A manager may need to interpret trends and understand data limitations. A senior leader may need to govern investment, cyber resilience and supplier risk.

Training should therefore reflect what each role is expected to decide.

Generic digital-awareness programmes are useful foundations, but they do not replace competence within the actual workflow.

Connected care should reduce administrative burden so that professional attention can return to people

The strongest argument for digital health is not simply faster information.

It is better use of human attention.

If a nurse spends less time searching for records, more time becomes available for clinical judgement. If an RCHE worker does not repeatedly enter the same information, more attention can remain on residents. If a family does not need to retell the same history at every contact, the care journey becomes less burdensome.

Technology should therefore be judged partly by what unnecessary work it removes.

A connected system that generates more screens, more alerts and more duplicate documentation may technically modernise care while making it operationally harder.

Remote monitoring can extend visibility between formal contacts

Much of long-term care takes place between appointments.

An older person may see a doctor periodically while changes in mobility, sleep, appetite, blood pressure or daily routine develop gradually at home. Remote monitoring can help make some of those changes visible sooner.

Depending on the purpose, this may involve physiological monitoring, activity sensors, telecare alerts or other connected devices. The potential benefit is not continuous surveillance for its own sake. It is earlier recognition of meaningful change.

The broader remote monitoring and telecare agenda is therefore particularly relevant to Hong Kong’s ageing population, where technology may help extend professional visibility without requiring every concern to trigger a hospital visit.

But remote monitoring only becomes connected care when somebody knows what to do with the information.

An alert is useful only when it changes a decision

Digital systems can produce large numbers of alerts.

Some identify genuine deterioration. Others reflect temporary variation, device error or thresholds set too sensitively.

This creates an operational requirement for escalation logic.

Who receives the alert? How quickly should it be reviewed? What constitutes urgent escalation? What happens outside normal service hours? Which professional has enough context to interpret the signal?

Without answers to these questions, remote monitoring can create a new queue rather than better care.

The strongest models therefore design the response pathway before expanding the technology.

Operational scenario: a sensor detects change but not the reason for it

An 80-year-old man with chronic respiratory disease lives alone and receives community support. A connected monitoring arrangement tracks selected physiological readings and daily activity patterns.

Over several days, the system shows reduced movement and a gradual deterioration in one monitored parameter. An alert is generated.

The data suggest that something has changed, but they do not explain why.

A clinician reviewing the information considers whether the pattern requires assessment. At the same time, a community worker reports that the man has eaten less and has seemed unusually tired during recent visits.

Bringing those sources together produces a stronger picture than either could provide independently. The man is contacted and assessed before his condition deteriorates further.

The scenario illustrates the real value of connected monitoring. The device did not diagnose the problem. It created an earlier signal that became meaningful when combined with professional judgement and contextual information from everyday care.

Connected care needs escalation pathways that cross health and long-term care

Remote information can lose value when the organisation receiving it cannot act beyond its own boundary.

An RCHE may identify a concerning trend but need clinical advice. A home-support worker may notice deterioration but not have direct access to the professional responsible for ongoing healthcare. A family may receive a device alert but be unsure whether to contact a doctor, emergency service or community provider.

The pathway therefore matters as much as the sensor.

For connected care to support earlier intervention, organisations need clear routes through which information can reach the professional or service capable of changing the outcome.

This is especially important for older people with multiple conditions, where responsibility may already be spread across several teams.

Technology should support ageing in place without transferring excessive responsibility to families

Digital tools are often presented as a way to help older people remain at home longer.

That can be valuable, but the model needs careful design.

A daughter receiving movement alerts throughout the day may technically have greater visibility of her father’s safety while also acquiring a new monitoring workload. Families can become anxious about whether every unusual reading requires action.

Technology can therefore shift responsibility rather than remove it.

A stronger approach defines what families are expected to do, what professional support remains available and which alerts require formal escalation.

Connected care should strengthen family confidence rather than quietly turn relatives into unpaid monitoring centres.

Consent and proportionality become more important as monitoring becomes continuous

A blood-pressure reading taken once a day and a camera observing a living space continuously raise very different privacy questions.

Older people should not lose privacy simply because technology makes monitoring technically possible.

The stronger principle is proportionality.

What problem is being addressed? What is the least intrusive technology capable of addressing it? Who can access the information? How long is it retained? Can the person withdraw agreement where applicable?

The wider digital safeguarding and technology-enabled risk agenda is relevant because connected care can create harm as well as prevent it if surveillance expands beyond its original purpose.

Residential services face different digital opportunities from people living at home

RCHEs can use connected technology across a shared service environment.

Digital care records, medication systems, sensor-based alerts and communication tools can strengthen consistency and help managers identify trends across residents.

The scale can also create additional complexity.

An alerting system that works for one person at home may generate dozens of notifications when applied across a large home. Staff need to know which alerts are urgent, which can wait and how responsibilities are divided during different shifts.

Technology therefore needs to fit staffing and workflow.

A system that increases visibility but overwhelms workers with non-actionable information may reduce rather than improve safety.

Digital tools can support workforce productivity without becoming a substitute for staffing

Hong Kong’s long-term care workforce pressures create strong incentives to use technology more efficiently.

Digital scheduling can reduce administrative time. Mobile records can allow workers to document care closer to the point of delivery. Automated reminders can support follow-up. Shared information can reduce repeated telephone calls between services.

These gains matter.

But they do not mean that digitalisation removes the need for sufficient human care.

A sensor can indicate that someone has not moved. It cannot necessarily reassure a frightened person, help them wash or understand why they have stopped eating.

The strongest workforce case for technology is therefore augmentation: remove avoidable administration, improve coordination and direct scarce professional attention towards situations where human judgement and relationships are most valuable.

Operational scenario: a digital rota improves efficiency but weakens continuity

A home-support provider introduces automated scheduling to reduce travel and fill available workforce capacity more efficiently.

Initial results are positive. Travel time falls and fewer shifts remain unallocated.

Several older people, however, begin receiving support from a larger number of different workers because the scheduling system prioritises geographic efficiency above continuity.

Complaints increase among people with dementia and among families who rely on familiar workers to notice subtle changes.

The organisation therefore adjusts the scheduling rules.

Continuity is added as an important constraint rather than treating travel efficiency as the only optimisation objective. The provider then reviews travel time, continuity, punctuality and complaints together.

The lesson is that digital optimisation needs to reflect the outcome the service is actually trying to achieve. A technically efficient rota can still produce poorer care if the model ignores relationships.

Cyber resilience is becoming a direct care-quality responsibility

The more services depend on connected systems, the more cyber incidents can affect care delivery.

A ransomware attack, unavailable cloud system or compromised account can interrupt access to medication records, care plans, schedules or contact information.

This moves cybersecurity beyond the information-technology function.

The broader cyber security and digital resilience agenda therefore belongs within operational governance.

Managers need to understand which services depend on digital systems, what happens if those systems become unavailable and how essential information can still be accessed safely.

A digital transformation programme that improves efficiency but creates a single point of operational failure has not strengthened resilience.

Business continuity needs to include digital downtime

Connected care should have a credible degraded-mode operation.

If the electronic care record is unavailable, how will workers know essential medication or mobility information? If the scheduling platform fails, how will home visits be prioritised? If remote monitoring stops, who needs to be contacted?

The answers may include secure offline information, defined escalation arrangements and tested recovery procedures.

The principle is simple: essential care should not become impossible because one technology platform is temporarily unavailable.

Technology suppliers become part of the care ecosystem

Digital transformation introduces organisations that may not previously have been considered part of long-term care delivery.

Software vendors, cloud providers, device manufacturers and analytics companies can all influence continuity and data security.

Providers therefore need more than technical procurement criteria.

They need to understand service reliability, support arrangements, data ownership, interoperability, cyber controls and what happens if a supplier changes its product or exits the market.

This is particularly important where a digital system becomes embedded in day-to-day care.

The more operationally essential the technology becomes, the more supplier governance becomes a care-quality issue.

Data portability matters because older people should not become trapped inside one provider’s technology

Care arrangements change.

An older person may move from home support to residential care, change healthcare provider or begin receiving a different combination of services.

If essential information cannot move with them because it is locked into one proprietary system, digitalisation creates a new boundary.

Interoperability therefore has a strategic dimension as well as a technical one.

Services should be able to preserve relevant information continuity when providers or platforms change, subject to appropriate privacy and governance controls.

Quality dashboards can connect digital information with management action

Digital systems increasingly allow leaders to see incident trends, care outcomes, workforce measures and operational performance in one place.

This can strengthen oversight, particularly across larger organisations.

But the value of a dashboard depends on whether it supports decisions.

Organisations examining similar governance questions can use the Quality Dashboard Builder to structure relationships between quality, workforce and operational evidence. It is not a Hong Kong reporting requirement, but it illustrates an important principle: connected data should help leaders identify relationships and ask better questions rather than simply generate more metrics.

Predictive analytics should remain a prompt for professional review

As datasets become larger, systems may increasingly attempt to identify who is at higher risk of hospital admission, falls, deterioration or service breakdown.

This can support earlier intervention if used carefully.

Prediction is not certainty.

A model can identify statistical risk without understanding the individual’s preferences, family context or recent change. It can also reproduce bias if the underlying data reflect unequal access or inconsistent recording.

The appropriate role is therefore decision support.

A risk score may tell a team where to look more closely. It should not automatically determine what happens to the person.

Connected technology should support independence rather than redefine independence as self-service

Digital services can unintentionally equate independence with the ability to use technology without help.

That is too narrow.

An older person may remain highly autonomous while choosing telephone contact, asking a family member to assist with an app or preferring face-to-face support.

The wider technology and digital support for older people should therefore be judged by whether it expands control and access, not whether it moves everyone onto the same digital channel.

Connected care succeeds when technology adapts to people rather than requiring people to reorganise their lives around technology.

Digital governance needs to sit alongside clinical and service governance

As technology becomes embedded in ordinary care, responsibility for it cannot remain confined to information-technology teams.

A system outage can affect medication information. Poor data quality can alter a care decision. An incorrectly configured alert can either miss deterioration or generate excessive intervention. A scheduling algorithm can change continuity. A permissions model can affect privacy.

These are care-governance issues because they influence what happens to people.

Leadership therefore needs visibility of digital risk alongside workforce, incidents, complaints and service outcomes. This does not require senior decision-makers to become technical specialists. It does require them to understand which technologies are operationally critical, what risks those systems create and where accountability sits when something goes wrong.

Organisations examining similar questions can use the Governance Maturity Assessment to structure thinking around accountability, escalation and assurance. It is not a Hong Kong digital-health framework, but the principle is transferable: once technology influences care decisions, its governance belongs within mainstream organisational oversight.

Connected care requires governance across organisational boundaries

The greatest benefits of digital health often arise precisely where organisational responsibility changes.

Hospital discharge, referral to community services, transition into residential care and changes in primary healthcare all create points where information needs to move.

These transitions are also where accountability can become unclear.

If information does not arrive, who notices? If two records conflict, who resolves the discrepancy? If a digital interface fails, which organisation is responsible for ensuring that the care pathway continues?

Interoperability therefore requires operating agreements as well as technology.

Systems need defined responsibilities for information transfer, exception handling and escalation. Otherwise digital connection can make an interface appear integrated while frontline staff continue relying on telephone calls and manual workarounds whenever the automated pathway breaks.

Information continuity should be measured from the receiving service’s perspective

A sending organisation may regard a transfer as successful because information left its system.

The receiving organisation may experience something very different.

The document may arrive too late, contain information that is difficult to interpret or omit the functional detail required to organise immediate support.

A useful measure of digital integration therefore asks whether the recipient could act safely and efficiently on the information received.

This moves evaluation away from technical transmission and towards operational usability.

For older people, that distinction is critical. A record is only valuable if it supports the next stage of care.

Outcome measurement should include what digitalisation changes for people and workers

Digital programmes are often evaluated through adoption measures.

How many users enrolled? How many organisations connected? How many electronic records were created? How many remote-monitoring devices were deployed?

These measures show scale, but they do not show whether connected care improved the experience or outcome.

A stronger evaluation asks whether digitalisation reduced repeated history-taking, improved medication continuity, identified deterioration earlier, reduced unnecessary administrative work or made it easier for older people to navigate care.

It should also examine negative effects.

Did digital access become harder for people with lower confidence? Did staff spend more time documenting? Did alert volumes increase without improving intervention? Did families acquire additional responsibilities?

Technology should be judged by the care system it produces, not simply by utilisation.

Digital inequality can become health and long-term care inequality

The consequences of digital exclusion become more significant as more services move online.

If appointments, information, communication or follow-up are easier to access digitally, people who cannot use those routes may experience slower or more complicated access.

This can compound existing inequalities.

An older person living alone, somebody with cognitive impairment, a person with limited literacy or an individual without reliable family support may need more help to navigate digital services than somebody with a confident relative managing technology on their behalf.

The equity test is therefore not simply whether a digital service is technically available.

It is whether people can use it effectively without requiring resources they do not possess.

Language and accessibility matter as much as technical availability

Digital inclusion also depends on how information is presented.

Small text, complex menus, unfamiliar terminology and multi-stage authentication can turn a technically accessible system into a practical barrier.

For some older people, voice support, simplified navigation or assistance from staff may be more important than adding additional functionality.

Digital design for ageing therefore needs to consider eyesight, hearing, dexterity, cognition and language alongside conventional measures of usability.

A sophisticated platform that only confident users can navigate independently may widen rather than reduce differences in access.

Artificial intelligence is likely to increase the value and complexity of connected data

As Hong Kong’s digital-health infrastructure develops, artificial intelligence may increasingly be used to analyse large volumes of information, identify patterns and support decision-making.

Potential applications include identifying people at elevated risk of deterioration, summarising longitudinal records, supporting scheduling or highlighting unusual combinations of clinical and operational data.

These possibilities are significant, but they should be distinguished from established routine practice.

The strongest role for AI in older people’s care is likely to be augmentation rather than autonomous decision-making.

A system may identify a pattern that deserves review. A professional still needs to determine whether that pattern is meaningful for the individual.

Older people are not simply collections of risk variables. Their choices, functional ability, family situation and changing goals remain essential to interpretation.

Algorithmic confidence should never be mistaken for certainty

Predictive systems can present outputs numerically, which can make them appear objective.

But predictions depend on the information used to create them.

If one group has historically had less access to services, its data may reflect unequal access rather than lower need. If care records are inconsistent, a model may learn patterns from incomplete information.

Governance therefore needs to ask what data sit behind the prediction, whether outcomes are being checked and whether automated recommendations create unintended differences between groups.

Human oversight becomes more important, not less, as analytical systems become more sophisticated.

Procurement decisions increasingly determine future interoperability

Every major digital purchase creates a long-term architectural choice.

A provider choosing a care-record platform, monitoring system or scheduling product is not only buying current functionality. It is also deciding how easily information may connect with other services later.

Procurement therefore needs to consider interoperability, data portability and supplier dependence alongside immediate cost and features.

A low-cost system that creates years of isolated data may prove more expensive strategically than a product designed around open exchange and practical portability.

This is particularly relevant to ageing services because older people frequently change care settings over time.

Digital strategy should avoid creating new organisational silos

One of the ironies of digital transformation is that every organisation can modernise while the system remains fragmented.

A hospital develops excellent electronic records. An RCHE adopts a sophisticated care platform. A community provider introduces mobile documentation. A family uses its own monitoring application.

If those technologies cannot support meaningful information continuity, each organisation has become more digital without the person experiencing more connected care.

The future challenge is therefore architectural as much as technological.

Digital investment should increasingly be considered against the pathway that older people travel rather than only the needs of the organisation purchasing the system.

International learning lies in building connected infrastructure without assuming one universal record

Countries approach health and long-term care data through very different institutional structures.

Some have highly centralised national records. Others operate through insurers, regional systems or numerous independent providers.

Hong Kong’s eHealth infrastructure is shaped by its own healthcare arrangements and cannot simply be transplanted into another system.

The transferable lesson lies in creating reliable information continuity while preserving clear purpose and accountability.

Integration does not necessarily require every piece of health and social information to be stored in one place.

It requires the important information to be available when it affects a decision, understandable to the person receiving it and governed appropriately.

Other systems can adapt that principle without replicating Hong Kong’s institutional mechanism.

The future direction is from digital records towards genuinely connected care

Hong Kong has already established important foundations for digital health.

The next stage is more demanding because it concerns relationships between systems rather than the digitalisation of individual services.

Clinical records need to connect more effectively with transitions. Community observations need routes into healthcare decision-making. Remote monitoring needs defined escalation. Care technology needs to reduce rather than multiply administrative work.

Digital development also needs to remain accessible to older people who use technology differently or require human support.

The strategic opportunity is therefore not simply more digital care.

It is a care system in which technology makes organisational boundaries less burdensome to the person moving across them.

Conclusion

Hong Kong’s digital-health infrastructure gives it a strong foundation for improving care continuity as its population ages, but the next stage of development cannot be measured simply by the number of records, devices or platforms connected. Older people experience care across hospitals, primary care, residential services, community support and family life, and digital systems need to help those separate environments behave more coherently.

The strongest connected-care model combines reliable information with clear accountability. Records need to be current and useful to the receiving service. Remote monitoring needs an actionable response pathway. Digital tools should reduce administrative burden without replacing professional judgement or human relationships. Privacy, cyber resilience, accessibility and digital exclusion need to remain part of quality governance rather than being treated as secondary technical concerns.

Artificial intelligence and predictive analytics may extend what connected data can achieve, but their value will depend on transparency, evidence and appropriate human oversight. Technology should help professionals see more clearly, not create an invisible layer of decision-making around older people.

Hong Kong’s longer-term opportunity is therefore to move from digitally enabled organisations towards genuinely connected care. The measure of success will be whether older people experience fewer information gaps, safer transitions, earlier support and greater control as they move through an increasingly complex care system.