Digital Care Records in Hong Kong: Building Continuity Across Health, Rehabilitation and Community Support

An older person is discharged from a Hospital Authority hospital after a fall. During admission, medication has changed, physiotherapy has identified new mobility risks and outpatient follow-up has been arranged. Back at home, a community service is supporting daily living while a family member coordinates appointments. The quality of that transition depends not simply on whether information exists, but whether the right people can access an accurate, current and understandable version of it when decisions must be made.

This is becoming an increasingly important issue across Hong Kong’s health and long-term-care environment. The territory has established a substantial electronic-health infrastructure while residential homes, rehabilitation services, NGOs and community providers also maintain their own operational records. The wider Hong Kong Ageing, Long-Term Care & Community Support Knowledge Hub examines how those interfaces influence continuity, workforce, quality and independence.

The central challenge is that a digital health record and a digital care record are not the same thing. Clinical information about diagnoses, prescriptions and investigations is essential, but day-to-day care also depends on mobility, communication, nutrition, routines, family involvement, environmental risks, rehabilitation goals and what matters to the person. A connected care system therefore needs both stronger digital infrastructure and disciplined decisions about what information should move between services.

The opportunity for Hong Kong lies not in creating one enormous record containing everything about everyone. It lies in creating reliable information continuity across organisational boundaries while preserving purpose, consent, professional judgement and accountability.

Hong Kong has moved beyond a standalone electronic-health record model

Hong Kong’s electronic-health framework has been developing for more than a decade. The Electronic Health Record Sharing System provided a statutory basis for participating healthcare providers to share specified electronic health information for healthcare purposes.

That framework entered a new phase on 1 December 2025 when amendments renamed the governing legislation as the Electronic Health System Ordinance and introduced changes intended to support the continued development of eHealth.

The revised framework strengthens the concept of a personal eHealth account and streamlines how health information can be deposited and accessed through participating healthcare providers. Citizens retain control over important aspects of sharing, while the system is designed to make clinically relevant information more complete and more readily available across healthcare encounters.

The amended legislation also creates a power for specified healthcare providers to be required, through future subsidiary legislation, to deposit specified important health information into personal eHealth accounts. That power should not be confused with a universal requirement already applying to every provider and every dataset. Implementation of particular mandatory deposit requirements depends on further specification.

This distinction illustrates an important principle for digital transformation: infrastructure can be established before every operational use is compulsory or universal.

Health records and care records solve different problems

Electronic health information is particularly valuable for diagnosis and clinical treatment. It may include prescriptions, allergies, laboratory information, radiology findings, immunisation records and other healthcare data.

Long-term support involves additional information that may never sit naturally within a clinical record.

A residential home may need to know that a resident becomes disoriented in unfamiliar surroundings, requires a particular approach when transferring from bed to chair and prefers a daughter to be contacted before non-urgent appointments. A rehabilitation service may be working towards walking safely to local shops. A community worker may know that a person who appears independent during a clinic visit struggles significantly in their home environment.

These records answer different but complementary questions.

• A health record helps explain the person’s clinical condition and treatment.
• A care record explains how that condition affects everyday support.
• A rehabilitation record captures functional goals, progress and adaptation.
• A community record adds information about the person’s home, relationships and daily circumstances.
• Family and service-user information may explain preferences and changes that formal assessments have not yet captured.

Digital integration therefore requires more than transferring clinical information into social-care systems. Organisations need to decide what information has operational value in each setting.

Residential care already depends on structured records

Hong Kong’s residential-care framework demonstrates how extensively formal records already support everyday practice.

The Code of Practice for Residential Care Homes for the Elderly includes requirements and templates associated with individual care planning, personal health and nursing records, medication administration, individual drug records, accident reporting, restraint use, complaints and other aspects of resident care.

These records are not merely administrative evidence. They support decisions across shifts and professional groups.

An individual care plan should help staff understand current needs. Medication records support safe administration. Accident and incident records create evidence for review. Health and nursing documentation provides continuity as needs change.

The move towards digital care planning can strengthen this architecture, but only when digital systems make current information easier to understand and act upon.

Replacing a paper form with an electronic replica provides limited benefit if employees still have to search through numerous screens to identify what changed overnight.

Digital records should reduce fragmentation rather than digitise it

Care organisations can operate several systems simultaneously: a care-planning platform, medication system, workforce application, incident database, rehabilitation record and financial or administrative system.

Digitisation therefore does not automatically create integration.

One of the most common structural risks is that the same information appears in multiple places. A mobility status may be recorded in an assessment, risk plan, rehabilitation note and daily-care record. If one changes and the others do not, workers face competing versions of reality.

The stronger design principle is to establish which record is authoritative for each important type of information and how dependent records are updated when circumstances change.

This moves digital records and data governance from an IT issue into care operations.

Every additional system should answer a practical question: what decision does this record help somebody make, and how will users know whether it is current?

Scenario: discharge information arrives but the care plan does not change

An older resident returns to a care home after hospital treatment for pneumonia. Her discharge information shows a medication change and identifies increased weakness following several days in bed.

The nurse reviews the hospital information and updates the medication record. However, the resident’s electronic care plan still describes her as independently mobile within the home because the medication and care-planning systems are separate.

During the evening shift, a care worker checks the care plan rather than the nursing note and encourages the resident to walk to the bathroom as usual. She becomes unsteady and requires assistance.

The immediate problem is resolved, but the incident exposes a record-design weakness. Relevant information was available, yet it had not been translated into the record used by the worker making the day-to-day decision.

The home redesigns its hospital-return workflow. A return from hospital now prompts a structured reconciliation covering medication, mobility, nutrition, cognition, wound care, appointments and any temporary increase in supervision or assistance. The responsible worker records which care-plan sections were reviewed and which remained unchanged.

The operational lesson is important. Digital continuity is not achieved when one system receives information. It is achieved when information reaches the decision point where practice changes.

Interoperability needs a defined operational purpose

Interoperability is often described as the ability of systems to exchange data. In care services, the more important question is whether exchanged information can be used safely.

A technically successful interface may transfer a diagnosis without explaining how it affects support. Another system may import a medication list but fail to show that a medicine has been temporarily suspended. A rehabilitation score may transfer correctly yet be meaningless to workers who do not know how to interpret it.

Strong interoperability and system integration therefore requires shared meaning as well as shared data.

Organisations need agreed definitions, timestamps, source information, responsibility for correction and clear rules for dealing with conflicting entries.

Without these controls, interoperability can make information more available while making certainty harder to achieve.

eHealth can strengthen clinical continuity without replacing provider records

Hong Kong’s eHealth environment offers an important platform for healthcare continuity. Participating healthcare professionals can access relevant sharable information where the statutory and consent framework permits it, reducing dependence on patients carrying fragments of clinical history between providers.

That is particularly valuable for people living with multiple long-term conditions who interact with hospitals, primary healthcare and specialist services.

However, a care organisation should not interpret access to eHealth as a replacement for its own records.

A residential home remains responsible for documenting the care it provides. A rehabilitation team still needs its own assessment and progress evidence. A community organisation still needs information about agreed support, risk and outcomes.

The stronger model connects external clinical information with local care documentation while preserving clear responsibility for each.

Consent architecture matters when information moves between systems

The revised eHealth arrangements retain a structured consent framework. Healthcare recipients can join eHealth and, where applicable, provide sharing consent allowing relevant healthcare providers to access sharable information. Substitute decision makers may act within the statutory arrangements where the person is unable to provide the relevant consent themselves.

Following the December 2025 changes, the system also supports a streamlined mechanism intended to make it easier for private healthcare providers to deposit information in personal eHealth accounts, while citizens retain control over sharing with relevant providers.

For care organisations, the wider lesson is that technical capability does not itself establish permission.

A system may be capable of sharing a record with ten organisations. That does not mean all ten should receive it.

Consent also needs to be understandable. People should not encounter digital information sharing solely through lengthy forms completed at admission. They need clear explanations of what information is being used, why access matters and how choices operate.

This becomes particularly important for older people, people with intellectual disabilities and people experiencing cognitive change. Digital participation should not depend on sophisticated technical understanding.

Rehabilitation requires records that show change over time

Rehabilitation illustrates why care records should not simply describe current deficits.

A useful rehabilitation record needs to capture goals, progress, barriers, intervention and changing levels of support. It should help different professionals understand not only what a person can or cannot do but what they are working towards.

For example, a physiotherapist may record that an older person can walk a defined distance with assistance. A community worker may observe that the person can manage the same distance at home but becomes anxious outdoors. An occupational therapist may identify an environmental barrier. Family members may report that confidence varies substantially during the day.

Bringing these perspectives together allows support to remain outcomes focused rather than becoming a collection of disconnected professional notes.

The digital record should make trajectory visible. Is independence improving? Is support reducing? Has progress stalled? Has a new risk emerged?

This is very different from using the record primarily as evidence that scheduled visits occurred.

Scenario: three professionals record three different goals

A man recovering from a stroke receives physiotherapy, occupational therapy and community support after returning home. Each service has a separate electronic record.

The physiotherapy goal is to improve walking distance. Occupational therapy is concentrating on safe meal preparation. Community workers have been instructed mainly to prevent falls and therefore complete many tasks for him rather than allowing him to practise them.

Every team is acting reasonably from its own perspective, but the records do not communicate a shared objective.

At review, the man explains that his priority is to prepare breakfast independently and walk to a nearby café. That goal requires contributions from all three services.

The teams agree a coordinated plan. Physiotherapy focuses part of its work on the distance and endurance required for the café journey. Occupational therapy addresses kitchen organisation and safe preparation. Community support changes from task completion to graded assistance, recording where the man succeeds and where additional support remains necessary.

The digital record does not need to merge every professional note. It does need a shared layer showing the person’s central goals, relevant risks, current level of independence and agreed responsibilities.

This scenario shows why digital integration should follow the person’s pathway rather than organisational structures alone.

Community services need records designed for mobile work

Residential environments provide relatively stable access to records. Community care is different.

Workers may move between homes throughout the day, operate without immediate access to a manager and encounter changes in the person’s condition during relatively short visits.

A digital community record therefore needs to provide concise, current information on a mobile device without overwhelming the worker with years of history.

Important information may include current support needs, significant health risks, communication requirements, recent changes, emergency contacts and specific tasks or observations required during the visit.

The workflow also needs to allow staff to record meaningful changes quickly.

If a worker notices that an older person is markedly more confused, the system should support escalation rather than simply recording “visit completed”. If a family carer reports that eating has deteriorated, that information needs to reach whoever can assess whether the support plan should change.

Digital systems can therefore improve community responsiveness when they are built around decisions rather than documentation volume.

Structured data and professional narrative both matter

Digital systems often favour structured fields because they are searchable and measurable. That can strengthen consistency. A falls field completed in a standard format is easier to analyse across a service than hundreds of free-text notes.

But structured data has limitations.

Human support contains context that cannot always be reduced to a dropdown menu. “Refused meal” tells a different story from “declined lunch because daughter had just brought favourite food”. “Agitated” can conceal pain, fear, environmental overload or frustration.

Good digital design therefore combines structured information with concise professional narrative.

The structured layer helps organisations identify patterns. Narrative explains meaning.

This balance is essential for strong data quality and performance measurement. Standardisation should increase comparability without stripping information of the context needed for safe interpretation.

Record quality depends on workforce capability

A digital platform cannot compensate for weak observation, judgement or recording practice.

Workers need to know what is significant, what should be escalated and how to distinguish fact from interpretation. Managers need to recognise when records indicate deterioration. Clinicians need to explain changes in terms that non-clinical workers can act upon.

This creates a strong relationship between records and digital workforce capability.

Training should go beyond system navigation. Employees need competence in:

• recording contemporaneously and accurately;
• identifying significant changes in need;
• distinguishing observation from assumption;
• using escalation functions correctly;
• checking whether critical information has been updated; and
• understanding when another professional needs to be involved.

Managers also need assurance that digital records reflect actual practice rather than becoming a parallel administrative task completed retrospectively.

The Digital Transformation Readiness Assessment can help organisations examine whether technology adoption is matched by workforce capability, governance and operational readiness rather than being treated as a software implementation project.

Family contributions need structure and boundaries

Families frequently hold important information that formal services do not see.

A daughter may recognise that her father’s confusion usually precedes an infection. A spouse may know that apparent refusal of care reflects hearing difficulties. A parent may understand a disabled adult’s communication signals better than a new worker.

Digital records can make it easier to capture this knowledge, particularly when several services are involved.

But family information should not automatically become unquestioned fact.

Workers need to distinguish family observations from professional assessments and from the person’s own views. Where family members have digital access to information, permissions need to reflect the individual’s preferences and applicable authority.

The record should strengthen partnership without turning relatives into default decision-makers simply because they are available.

Scenario: family reporting identifies deterioration before formal review

An older woman receives community support following discharge from rehabilitation. Her son notices through regular visits that she is eating less and spending longer in bed. Individual care visits continue to be completed, but no single worker has seen enough change to recognise a pattern.

The son reports his concern to the community service. A worker reviews recent digital records and notices several apparently minor observations: unfinished meals, increased fatigue and two occasions when the woman declined her usual walk.

Because those entries can be viewed together, the service identifies a trend rather than isolated events. The issue is escalated for appropriate health assessment and the support plan is temporarily adjusted.

The organisation then examines its recording workflow. Workers had captured useful observations, but the system did not automatically highlight cumulative deterioration.

It introduces a simple escalation prompt where repeated changes in appetite, mobility or cognition occur within a defined period, while retaining professional review rather than allowing the system itself to determine clinical significance.

The value of the digital record in this scenario is not prediction for its own sake. It is the ability to assemble distributed observations into an earlier and more coherent picture of the person’s changing condition.

Medication information is a critical continuity test

Medication provides one of the clearest tests of whether digital records genuinely improve continuity.

A person may receive a prescription through a healthcare provider, have medicines dispensed elsewhere and rely on residential or community staff for administration or prompting. Hospital admission can change the regimen again.

Every transition creates potential discrepancy.

Residential-care documentation therefore needs to reconcile external prescribing information with the record used for administration inside the home.

Digital integration can reduce transcription and improve visibility, but governance still needs to define which source is authoritative and who verifies changes before practice is altered.

A system should never encourage an unqualified employee to interpret conflicting medication information simply because both versions are displayed electronically.

Digital records should support governance as well as frontline work

Well-designed digital records create organisational intelligence as a by-product of care delivery.

If workers record falls consistently, management can identify patterns by time, location or service. If care-plan reviews are structured, overdue reviews can be monitored. If rehabilitation goals are recorded consistently, organisations can analyse whether progress is being achieved. If repeated hospital transfers occur shortly after discharge, pathway weaknesses may become visible.

This information can strengthen quality monitoring systems, provided managers avoid confusing data volume with insight.

A useful dashboard might connect:

• care-plan currency and overdue reviews;
• medication discrepancies after transitions;
• incidents associated with incomplete information;
• changes in functional outcomes;
• hospital transfers and readmissions;
• record-access or system-availability problems; and
• service-user feedback about continuity.

The Quality Dashboard Builder can help organisations structure this type of assurance, translating operational records into a smaller number of indicators for management and governing-body oversight.

Scenario: a dashboard reveals a transition problem hidden inside individual records

An NGO operates several community services supporting older people after hospital discharge. Each service records assessments, changes and unplanned hospital returns correctly, but data is reviewed mainly case by case.

A new dashboard brings transition information together. It shows that one service has a noticeably higher number of unplanned hospital returns within the early period after discharge.

Managers initially assume that its users simply have greater clinical complexity. Further review shows something more specific. Medication discrepancies are taking longer to resolve, and follow-up rehabilitation appointments are being missed more frequently because information about them is not always transferred into the community record.

The provider works with relevant pathway partners to improve discharge reconciliation and introduces a structured transition section in its digital care record. It also records whether medication, rehabilitation follow-up, equipment and family communication have been checked.

Subsequent monitoring looks not only at whether the new fields are completed but whether discrepancies and avoidable returns reduce.

The example demonstrates the difference between digital documentation and digital assurance. Individual records tell the story of one person. Aggregated data can show whether the same weakness affects an entire pathway.

Cyber resilience is part of record continuity

Dependence on digital records creates a new form of operational risk. If the system becomes unavailable, workers may lose access to medication details, emergency contacts, current care plans or recent changes.

Cybersecurity therefore matters not only because personal data must be protected but because essential records must remain available when care is being delivered.

Organisations should connect digital-record design with cybersecurity and digital resilience.

Important controls include role-based access, secure devices, timely removal of leavers’ accounts, backup arrangements, incident response and a realistic downtime process.

A downtime plan should answer practical questions. How will workers obtain current medication information? Which critical records are available offline? How are updates made during the outage reconciled after systems return?

Paper contingency arrangements may still have a legitimate role where they protect essential continuity rather than simply replicating the whole electronic system.

Technology procurement should start with care pathways

Care-record systems are often purchased by comparing functions: care plans, medication modules, rostering, analytics, mobile access and integrations.

A better starting point is the service pathway.

Organisations should map where information currently breaks down, which workers need information in which settings, what duplicated recording exists and which transitions create the greatest risk.

Only then should they decide what technology is required.

This avoids a common problem in digital transformation: purchasing a powerful platform and then redesigning operational practice around the software’s structure rather than around the person’s pathway.

Providers considering significant change can use the Digital Twin Scenario Modeller to explore how changes in workflow, capacity and service interfaces might affect operational stability before implementing wider redesign. It is not a Hong Kong regulatory tool, but it can support structured scenario testing.

Artificial intelligence may change how records are used

The next phase of digital records is likely to involve greater automation and analytical support.

AI systems could potentially summarise long care histories, identify repeated changes in mobility, highlight patterns preceding hospital admission or help professionals find relevant information more quickly.

These uses remain an emerging area rather than an established substitute for professional review.

The quality of any AI output will depend heavily on the underlying record. If documentation is incomplete, biased or out of date, automated analysis can amplify those weaknesses.

AI also creates questions about explainability, confidentiality, supplier access and human accountability.

A summary generated automatically may save time, but somebody still needs to determine whether it accurately reflects the person. A risk alert may prompt review, but it should not become an automatic restriction on the person’s independence.

The stronger future model will use technology to reduce search and administrative burden while keeping consequential care decisions with appropriately accountable people.

Service-user access can rebalance the record

Digital records have traditionally been written primarily by professionals and organisations. Personal eHealth accounts create a different expectation: individuals increasingly expect to access and manage information relating to their own health journey.

That principle has wider relevance to care records.

Where appropriate, giving people clearer visibility of their own support plans, goals and review outcomes can strengthen accuracy and accountability. They may identify mistakes, challenge assumptions or explain that a recorded goal no longer matters to them.

This aligns digital records with co-production, choice and control.

Access needs to be designed carefully for people who require communication support or whose digital confidence is limited. Digital empowerment should not become digital exclusion.

Organisations should therefore retain accessible alternatives and support people to engage with records in ways that match their needs.

Digital inclusion matters for workers and service users

Hong Kong is highly connected technologically, but access and confidence are not uniform.

Some older people may be comfortable using eHealth applications and mobile services; others may rely heavily on relatives. Some staff may adopt new technology quickly; others may need substantial support. People with sensory, cognitive or communication needs may face additional barriers.

A digital care system should therefore measure adoption rather than assume it.

Low engagement may indicate poor design, accessibility problems, insufficient training or lack of trust.

The goal should not be to eliminate every non-digital interaction. It should be to use digital systems where they improve continuity while preserving human routes for people who need them.

International learning: continuity needs a shared information layer

Many care systems internationally are trying to connect hospitals, primary healthcare, rehabilitation and long-term support. The institutional architecture differs considerably, and Hong Kong’s eHealth system cannot simply be transplanted into other jurisdictions.

The more transferable lesson lies in creating a shared information layer around transitions.

People should not lose continuity simply because organisational responsibility changes. Yet achieving this does not require every provider to own the same system.

What matters is agreement on the critical information that must follow the person, who maintains it, how changes are reconciled and where responsibility sits when records conflict.

Digital integration therefore depends as much on governance and workflow as on technical interoperability.

The next stage should connect health data with functional and personal outcomes

Hong Kong’s eHealth development strengthens the infrastructure for clinical continuity. The next opportunity for the wider care system is to ensure that digital transformation also captures functional, rehabilitation and person-centred outcomes.

A strong connected record should help answer not only what conditions a person has, but whether they can remain at home, participate in their community, achieve rehabilitation goals, manage medication safely and maintain relationships that matter to them.

These outcomes sit across organisational boundaries.

The challenge is therefore to build interfaces that allow different services to contribute relevant information without creating one uncontrolled universal record.

That requires agreed data standards, clear access, reliable reconciliation and strong provider governance.

Conclusion

Hong Kong has created a significant digital-health foundation, and the Electronic Health System reforms effective from December 2025 provide further capacity for personal eHealth accounts and information sharing across healthcare. For long-term care, however, stronger continuity will depend on connecting that clinical infrastructure with the records used in residential homes, rehabilitation and community support.

The key issue is not simply digitisation. A paper-based fragmentation problem remains fragmented when reproduced across five electronic platforms. Better records need clear ownership, accurate updates, usable workflows, meaningful consent and information that reaches the worker or professional making the next decision. Medication, mobility, rehabilitation goals, family knowledge and changes in daily functioning all need to travel appropriately through the pathway.

Hong Kong’s strongest opportunity is therefore to build digital continuity around the person rather than around individual organisations. Health information should support care decisions; care records should reflect functional and personal outcomes; aggregated data should help leaders identify system weaknesses; and technology should reduce administrative friction without replacing professional judgement.

As eHealth, interoperability and artificial intelligence develop, the quality of the underlying record will become increasingly important. Connected care will ultimately depend not on how much information Hong Kong can digitise, but on whether reliable information follows the person safely enough to support better decisions, stronger independence and more coherent care.