In December 1999 I wrote a column for Berita MMA called “Is a Browser Enough?”, and in it I quoted my own first article back at myself, written in March 1997, on the minimum computer a doctor would need to get on the Internet: a Pentium 133MHz processor, 16MB of RAM, a 33,600-baud modem. I remember being faintly amused, at the time, that this had already become quaint after less than three years. That is the correct rate of decay for a modem. Nobody expects 1997 hardware to still be current.

A few months after that first column, in August 1997, I wrote another one, about telemedicine. Malaysia had just passed the Telemedicine Act, one of the Multimedia Super Corridor’s seven flagship applications alongside Electronic Government and Smart Schools, and the government’s own blueprint gave the scale of the ambition in hard numbers: 111 government hospitals, 206 private hospitals, 9,608 doctors, 50 insurance companies, over seven million medical records, all meant to be linked by a telemedicine network built on high-definition imaging, secure networks, and remote diagnosis.

Here is the part that is not merely quaint. The modem became obsolete. The blueprint’s own numbers were superseded by newer numbers years ago, the way any statistic is. But the Telemedicine Act itself, passed by Parliament, given royal assent on 18 June 1997, has never come into force. Twenty-nine years is not a delay. It is close to the entire span of my own writing about medicine and the internet, and the specific legal instrument meant to make good on the promise I wrote about in 1997 is, as I write this in 2026, still sitting exactly where it was left.

Malaysia’s problem is no longer the absence of telehealth. It is the absence of coherence between telehealth, regulation, national health data infrastructure, reimbursement and workforce preparation.

Twenty-nine years in limbo

The Telemedicine Act 1997 (Act 564) is short — six sections, no schedule — but more carefully constructed than a single penalty clause suggests. Having now read the Act’s own text directly, it turns out to define three separate tiers of consequence, not one. Practising telemedicine without authorisation carries the harshest penalty: a fine of up to RM500,000, up to five years in prison, or both. A separate provision, Section 5, requires a doctor to obtain the patient’s written consent before any telemedicine consultation, specifically covering the right to withdraw at any time, the risks and benefits, confidentiality protections, and a bar on sharing identifiable telemedicine records with researchers without the patient’s separate consent — breaching this carries its own penalty, up to RM100,000 or two years’ imprisonment. These are not symbolic numbers. They read like a law that was meant to be enforced, in some detail, by people who had thought through what enforcement would actually look like.

It never was, because the Act itself was never switched on. Section 1(2) leaves its commencement to the Minister, by notification in the Gazette, and no such notification has ever been made. In its place, Malaysian doctors operate under self-regulation: the Malaysian Medical Council issued a temporary Advisory on Virtual Consultations during the Covid-19 movement control orders in April 2020, later replaced by a standing Guideline on Telemedicine. The MMC’s own position, as recorded in an international legal survey, is narrower than the dormant Act’s own scope in general, though not as narrow as I had first stated. Having now read the guideline itself directly, its actual position is a presumption with a stated exception, not a blanket rule: virtual consultation should generally follow an existing doctor-patient relationship established in person, but the MMC explicitly permits a virtual consultation as a first encounter too, especially in primary care, provided the doctor documents the reasoning and can justify it if challenged. The guideline is also specific about who should not be seen virtually at all — patients with cognitive disorders, those suspected of being under the influence of drugs or alcohol, and those facing a language barrier, all of whom it says should be seen in person instead.

A structured telehealth programme for stable, already-diagnosed chronic disease sits comfortably inside either version of the rule — it is precisely the continuation-of-care relationship the guideline treats as standard, so nothing here needed the exception at all. The absence of Act 564 has not, in practice, stopped telehealth from developing. As the next section shows, six separate organisations have each built their own version of it under the Medical Act, MMC guidelines and general professional standards, without Act 564 being switched on. What the dormant Act leaves behind is not a blocked industry but a framework that never grew up alongside the practice — no statutory footing for the MMC’s own guideline, no legislated consent standard, no clear enforcement pathway when things go wrong. That is a different and narrower problem, but it is a real one.

The guideline itself is also a smaller version of the same pattern running through this whole piece. Its own final page records that it was endorsed by the MMC’s Ethics Committee on 16 October 2023 and by the full Council on 23 January 2024, and states plainly that it is scheduled for review two years after endorsement — meaning a review was due by 23 January 2026, seven months before this piece was written. The document even warns, in its own words, that changes in the law before that review “may render parts of this guideline obsolete.” As far as I can find, no updated version has been published. The statute waited twenty-nine years for commencement; now even the interim guideline appears to have passed its own review date.

The Telemedicine Act did not pass in isolation. It was one of five cyber laws Malaysia enacted together in 1997 and 1998 as part of the same MSC package: the Digital Signature Act (Act 562), the Computer Crimes Act (Act 563), the Telemedicine Act itself (Act 564), the Communications and Multimedia Act (Act 588), and a Copyright Amendment extending protection to digital works. Act 563 and Act 564, numbered one apart, passed together, are the clearest possible comparison of what happened to two siblings from the same year.

In 2026, Malaysia finally rebuilt Act 563. The Cybercrimes Bill 2026 was tabled in June, passed by the Dewan Rakyat on 1 July, and passed by the Dewan Negara on 20 July, repealing the nearly three-decade-old Computer Crimes Act outright and replacing it with a framework built for ransomware, deepfakes, digital-identity abuse and AI-generated fraud, enforced through the National Cyber Security Agency. Whatever one thinks of the new law’s specifics, the legislative machinery clearly still works: Parliament can identify an obsolete 1997 statute, debate it, and replace it within a single year when it decides to.

Act 564 got no such treatment. As far as I can find, nobody has tabled a Bill to revive it, repeal it, or replace it with something built for how telemedicine is actually practised today. It simply continues to exist, unused, the way it has for twenty-nine years. Even the reference site that catalogues Malaysia’s cyber laws today describes Act 564 in the present tense — as a law that “regulates the practice of telemedicine” and “establishes the framework for remote medical consultation” — without noting that it has never actually come into force. That is not a small slip. It is exactly the kind of confident description that lets a dormant law keep sounding operative for another twenty-nine years.

The other flagship

The unfinished telemedicine law matters more now because another part of the same 1997 vision — a lifetime digital health record for every citizen — is finally beginning to take shape. The Telemedicine Act was not the only piece of the 1997 vision left unfinished. The same Multimedia Super Corridor blueprint I wrote about that year also promised a National Multi-Purpose Card — what became MyKad, launched in 2001 — carrying an individual’s identity and, alongside it, access to services including health. The idea underneath it was womb to tomb: one identity, one continuous health record, accessible wherever a person actually sought care, for life.

That specific ambition has its own long paper trail of restatement. A 2007 academic comparison of electronic health record efforts across Malaysia and four East Asian countries already called an integrated Lifetime Health Record “fundamental” for continuity of care, and noted plainly that the aim had not yet been realised. That is nineteen years before this piece was written, describing the same unfinished promise in the same terms.

It is, finally, actually moving. As of 2026, Malaysia is expanding its Total Hospital Information System to sixteen hospitals, rolling out a cloud-based Clinic Management System to roughly 2,489 primary care facilities, and building what is now officially called a National Health Interoperability Platform, explicitly working toward a Lifetime Health Record for every Malaysian citizen. The Khazanah Research Institute has argued for years that this kind of basic infrastructure investment matters more for population health than headline technology benefiting only a small number of patients — a case that appears, this year, to finally be winning.

One design limit is already built into the current plan and worth stating plainly rather than discovering later: the Lifetime Health Record, as currently scoped, assumes a MyKad and enrolment in the national health system. Malaysia’s roughly 2.2 million registered foreign workers, who generally do not hold a MyKad, sit outside that model by design. Any telehealth service built on top of this infrastructure would inherit the same gap unless it deliberately addresses it. The equity risks this creates — structural exclusion for those without MyKad, and practical exclusion through geography, connectivity and digital literacy — are taken together in “Who this could leave out” below.

This changes what the data-standard question in this piece is actually about. It is no longer a hypothetical — Malaysia is currently building the national interoperability layer a telehealth service for chronic disease would eventually need to sit on top of. The live decision is not whether to invent a standard, but whether new telehealth services are designed from the outset to plug into the National Health Interoperability Platform as it matures, or built in parallel and reconciled with it later, the way the six existing telehealth systems already described in this piece never were reconciled with each other.

What the evidence actually says now

Whatever gets built should be built on the evidence as it stands today, not on 1997’s optimism. The studies cited below cover different intervention types — telemonitoring of physiological readings, teleconsultation by video or telephone, and broader eHealth platforms — and evidence for one modality does not automatically carry over to another. For diabetes, the picture is genuinely encouraging, if modest: a 2022 systematic review and meta-analysis of 45 trials found telehealth reduced HbA1c by 0.353 percentage points compared with usual care — a real, statistically significant effect, alongside reported cost savings and higher patient satisfaction, but a modest one rather than a transformation.

For hypertension specifically, the evidence is weaker and more contested. An umbrella review of systematic reviews and meta-analyses found no clinically meaningful effect of telemedicine on blood pressure, and rated the overall certainty of the underlying evidence as low to very low. Where telemonitoring shows its clearest benefit is actually elsewhere: an earlier systematic review across four chronic disease groups found more consistent reductions in emergency visits, hospital admissions and length of stay for pulmonary and cardiac conditions than for diabetes or hypertension.

For weight and waist circumference, the evidence sits in a similar register to diabetes: real, but modest, and sensitive to what it is compared against. A 2025 meta-analysis of nursing-led telemedicine interventions found an average weight loss of 2.59 kg and a BMI reduction of 1.05 kg/m² relative to traditional care. A separate review of eHealth interventions found they performed comparably to face-to-face care, with a materially larger effect only when compared against no intervention at all — meaning the strongest evidence is for doing something remotely rather than nothing, not for remote care outperforming an in-person clinic.

None of that is a reason not to build this. It is a reason to be precise about what it is for. Among the chronic-disease applications considered here, diabetes has one of the strongest and most practically applicable evidence bases. A Malaysian telehealth programme aimed at stable diabetes has real evidence behind it. One aimed primarily at hypertension control should expect a smaller and less certain effect, and should be designed and evaluated accordingly rather than assumed to work simply because the diabetes literature does.

What already exists, scattered across six systems

None of this is hypothetical anywhere else in the country. Structured telehealth for continuing care already runs in at least six places in Malaysia, privately and publicly, without any of them being connected to each other or to a shared national record.

DoctorOnCall, founded in 2016, is Malaysia’s first dedicated telehealth platform — demonstrating commercial scalability at national scale. (Disclosure: I hold shares in DoctorOnCall as an early investor and previously served as an advisor during its startup phase; its inclusion here is based on published reporting and publicly available information.) For chronic disease patients, it issues repeat prescriptions after a teleconsultation, capped at three months, with identity verified against insurers. In 2019 it piloted DOCPod in Langkawi, a kiosk combining vital-sign screening with a video link to public health clinic doctors, referring physical-examination cases to Klinik Kesihatan Malaysia.

Sunway Medical Centre’s Telemedicine Command Centre, launched January 2021, proves the private-hospital continuation-of-care model: follow-up patients and second opinions, 70-plus doctors across 40-plus specialities, 24/7.

Subang Jaya Medical Centre, through Connected Care, shows how far the private model can go: a free advisory line, video follow-ups, home nursing, and since July 2023 SeniorConnect — remote monitoring with escalation for patients 65 and older. The structural model is right; the reach is limited to Klang Valley patients already under SJMC’s care.

On the public academic side, UMMC’s Department of Primary Care Medicine has run a teleconsultation service since April 2021 for chronic-disease patients — blood pressure and blood sugar monitoring with a single sign-on after in-person consent — demonstrating that academic primary care can implement this quietly and sustainably.

Sarawak Heart Centre, an MOH facility, formally evaluated TELEG, its geriatric telemedicine service, across 148 patients — the one example here with published outcome data and a validated acceptance instrument behind it, not just a service description. It is the evaluated public-sector model.

There is a sixth example, and it is the most consequential one, because it already has the reach none of the other five do. MySejahtera was built as Malaysia’s Covid-19 check-in app and was, at its peak, installed on roughly 30 million of the country’s 32 million people. Rather than retire it after the pandemic, MOH deliberately repurposed it from mid-2022 onward into a platform for chronic disease management and immunisation tracking — the same pivot this piece has been arguing for, already attempted, on the one app already installed on almost every phone in the country.

It is further along than a stated intention. In 2023, MySejahtera added a Home Assessment Tool for remote monitoring of patients with chronic illness — daily symptom reporting, vital sign uploads, direct messaging with a healthcare provider. A peer-reviewed 2024 study of over 4,400 patients found overall uptake of only 39%, with older patients and non-Malaysian citizens significantly less likely to use it. The platform with the scale to test national reach at the same time exposed exactly the equity gap this piece has already described.

MySejahtera’s enormous reach makes it important, but not necessarily as the telehealth platform itself. Its real value may be as infrastructure: identity verification, patient authentication, access to the national health record, and a trusted channel already on almost every phone in the country. Clinical services could be delivered by different providers through interoperable systems, with MySejahtera providing the patient-facing layer rather than hosting the clinical logic. The important thing is the data architecture, not which app happens to be on the patient’s phone. Its own history is also a caution: it carries the credibility damage of having been a mandatory pandemic check-in tool, which the minister overseeing its pivot acknowledged directly at the time, and a stated plan from 2022 to add the Pink Book maternal record to the app has not been confirmed as delivered as of the most recent account available.

Put side by side, these six are proof of concept, not a system. Two private hospitals, a teaching hospital, a state facility, a commercial platform, and now a national government app have each independently built some version of this, with no shared data standard connecting any of them to another, let alone to the National Diabetes Registry already discussed. The gap this piece has been describing is not whether telehealth for chronic disease can work in Malaysia. It already does, in at least six places, one of them installed on almost every phone in the country. The gap is that nobody has joined them up.

What the next step should actually look like

The six examples above are not a foundation to build on. They are evidence of the actual problem: six separate organisations have each independently concluded that structured telehealth for continuing care is worth doing, and each has built its own answer, with no shared data model connecting any of them. Adding a seventh standalone system would simply repeat the mistake one more time. Malaysia does not need another platform. It needs the ones it already has to talk to each other.

The backbone already exists, or will shortly. Malaysia is currently building a National Health Interoperability Platform and working toward a Lifetime Health Record for every citizen. That should be the layer telehealth services eventually plug into — not another independent database with another patient portal sitting alongside it. The live design decision is not whether to build from scratch, but whether new telehealth services are designed from the outset to connect to the national interoperability layer as it matures, or built in parallel and reconciled later, the way the six systems already described never were.

The model for chronic disease should be asynchronous first, not video-consultation first. A diabetic patient does not necessarily need a monthly video call. For many stable chronic-disease patients, what may matter more than a scheduled video consultation are longitudinal readings — blood glucose, HbA1c, blood pressure, weight — alongside medication adherence, symptom reports, and alerts when something is going wrong. The clinician reviews, communicates, and intervenes when the data warrant it. That model — monitor, review, communicate, escalate when necessary — is potentially far more scalable than reproducing the traditional clinic consultation over a screen, and it is what the evidence for remote telemonitoring already supports. Physical consultation remains available, and should be easy to reach, whenever the clinical situation requires it. Escalation to in-person care should never be treated as a failure of the telehealth system. The important clinical skill this demands is recognising when remote monitoring is sufficient and when it is no longer sufficient — and the system should make that transition easy rather than hide it.

The workforce delivering this does not need to be doctors at every step. Nurses, pharmacists, diabetes educators and other allied health professionals can handle protocol-driven monitoring, patient education and routine follow-up, with escalation to a doctor when readings or symptoms cross a defined threshold. A digital system that simply transfers the existing doctor bottleneck online will not solve the access problem it is meant to address.

Clinical governance is not a design detail to be settled later. Before any system scales, the following questions need explicit answers: who holds clinical responsibility for a patient enrolled in remote monitoring? Who reviews an abnormal reading flagged at 2 am, and within what timeframe? What is the liability when an alert is generated but not acted upon? Who is the data controller, and on what consent model is secondary use permitted? Each of the six systems already operating has necessarily answered these questions locally, but those arrangements will not automatically generalise to a national framework. Getting them right before scaling is less expensive than getting them wrong after.

Clinician workflow is where digital health projects most commonly fail. If a doctor has to log into a separate portal, manually scroll through hundreds of readings, and then duplicate documentation into their existing electronic medical record, the system will eventually be ignored. The information needs to arrive summarised, with abnormal trends flagged and routine readings filtered. Interoperable longitudinal data would also allow increasingly capable decision-support tools, including AI-assisted summarisation, to flag clinically relevant change without displacing clinician responsibility. A platform built on interoperable data from the start is the precondition for any of that to work.

Start with diabetes, build the architecture for more. Among the applications considered here, diabetes has one of the strongest and most practically applicable evidence bases, the most standardised measurements, and the most established monitoring workflow. It is the logical first clinical module. But the architecture should be built from day one so that hypertension, hyperlipidaemia, weight, maternal care and others can be added as separate modules without requiring the whole system to be rebuilt. The risk of specifying all conditions upfront is that the first version becomes so complicated that it never actually launches. Healthcare IT projects have a consistent tendency to become enormous before anyone has shown that clinicians or patients will use them. Build one module properly, demonstrate that the workflow works for real clinicians and real patients, then expand.

On the National Diabetes Registry: interoperability matters, wholesale migration is a separate question. The NDR holds records for more than 1.6 million patients, but its annual cross-sectional audit was never designed to track individuals over time. One research team had to manually link records for patients who happened to appear in at least two separate audits, precisely because continuous patient-level tracking was not part of the original design. A new system built for longitudinal monitoring should connect to the NDR where that is useful, but importing 1.6 million historical records of variable quality is a separate question from building interoperability. Interoperability does not mean copying every historical datum into a new database. The right starting point is to ask what information is clinically useful, whether it is current, whether patient identity can be reliably matched, and whether the data quality is adequate — not to import simply because the data exist.

Maternal and child care illustrate the modular expansion this architecture should eventually support — and where the womb-to-tomb ambition raised earlier in this piece actually starts. The evidence for antenatal telehealth is comparable to the adult chronic-disease conditions: a meta-analysis of 35 studies covering more than 16,000 pregnancies found telehealth follow-up broadly equivalent to in-person care. Malaysia already runs paper versions of both, in the Pink Book for pregnancy and the Child Health Record Book from birth to age six.

A properly designed system would also generate much of the data needed for audit and feedback as a by-product of routine care. Every reading arrives already timestamped and attributable — which is close to what quality-improvement literature calls audit and feedback: giving a clinician a live view of their own patients against a guideline target, without any payment attached to the number. Cochrane’s own review, drawing on more than 140 trials, finds a real but modest average effect — 4.3% absolute improvement in professional compliance, with wide variation. A platform built around integrated monitoring generates this feedback as a structural by-product, rather than as a separate initiative someone has to build and justify afterwards.

Remote monitoring is clinical work and should be paid as such. Reviewing blood pressure readings, glucose measurements, laboratory results and patient messages takes time even when there is no conventional consultation. One practical option would be a sessional model — a clinician committing a fixed block of hours each week to remote chronic-disease review, paid on its own schedule separate from the per-consultation fee — which reflects what the work actually is without importing the per-encounter structure that already fails in-person primary care. A companion piece has set out how Malaysia’s private GP consultation fee has been frozen since 1992; trying to pay for telehealth monitoring through the same structure would import that problem directly into the new service before it has even started.

Who this could leave out

Four groups deserve explicit attention in design, not as afterthoughts. Patients in rural areas and those with poor connectivity face the most obvious barrier; the design response is assisted access — through community health workers, kiosk-style touchpoints, or low-bandwidth alternatives — rather than assuming smartphone ownership and a reliable data connection. Older and digitally less confident patients are more likely to need monitoring and less likely to use a self-managed app; the MySejahtera data already showed that uptake among those 65 and older was significantly lower, and interfaces need to be designed for this group, not retrofitted after launch. Patients with language barriers will be poorly served by a platform built around a single interface language; multilingual support is not a nice-to-have but a patient-safety requirement for any system handling escalation alerts. And foreign workers and others without MyKad sit outside the current Lifetime Health Record design entirely — an alternative identity pathway needs to be built into the architecture from the outset. A pilot that only reaches confident, urban, MyKad-holding patients would improve care for exactly the group least in need of the improvement, while leaving unchanged the access problem it was meant to solve.

Training the workforce that will actually use this

None of this works without doctors who are actually ready to practise this way, and that readiness cannot be assumed. A platform can be built well and still fail because the people using it were never taught how.

Malaysia’s own gap has been diagnosed directly, recently, in print. A January 2025 paper in The Lancet Regional Health — Western Pacific found that where Malaysian medical curricula touch on digital health at all, it is typically limited to electronic medical records and basic telemedicine awareness, well short of comprehensive training, and that no national framework exists to say what digital health competencies should be taught or how they should be assessed. The paper’s own recommendation is specific: the Malaysian Medical Council’s medical education committee should convene a working group to build one, potentially adapting an existing international competency framework rather than starting from nothing.

Some basic digital-health competency belongs in undergraduate and postgraduate training, and housemanship is one natural place to introduce it — it is already the mandatory, structured period every Malaysian medical graduate passes through regardless of which university trained them, rotating through six postings including medicine, surgery, obstetrics and gynaecology, orthopaedics, paediatrics, and one alternative posting, under direct Malaysian Medical Council oversight. But the emphasis should be on principles rather than platform-specific skills: patient selection for remote care, the limitations of what a virtual assessment can and cannot detect, consent, confidentiality, documentation, escalation pathways, and the appropriate use of digital tools. Telehealth will change quickly, and a compulsory module built around today’s specific platform risks becoming outdated before the cohort that received it has finished their first year of practice.

Housemanship only reaches new graduates, and most of the doctors who will actually staff this platform in its first years qualified long before any of this existed. That gap is real and documented, not assumed; the more encouraging finding is that the barrier is rarely willingness. A New Zealand study found widespread interest in digital health upskilling among practising clinicians, with time, institutional leadership and educator shortages as the actual barriers, not resistance to the idea itself. For these doctors — who will actually staff any new system in its first years — the more important vehicle is CPD.

Malaysia’s MMC Continuing Professional Development system already requires practising doctors to accumulate points across categories that include formal courses and structured work-based learning. A telehealth competency module could become a recognised CPD category without creating a new training bureaucracy. The format that has worked best elsewhere for exactly this population is short, self-paced online units — the kind US professional bodies built specifically for faculty who were themselves uncomfortable with telemedicine, introduced because the pandemic forced existing staff and new trainees to learn the same skills at the same time. Malaysia has the luxury of doing this in advance and deliberately, rather than in a rush during the next crisis. The same principle applies to the broader workforce: nurses, pharmacists and allied health professionals involved in protocol-driven monitoring will need their own parallel pathways through their respective professional bodies, not as an afterthought once the medical training is settled.

What would actually need to happen

1. Decide the fate of Act 564, one way or the other. A law that has waited twenty-nine years for a commencement notice is not a law in reserve. It is a standing question nobody has answered — and the answer matters even though telehealth has developed anyway. Act 564 provides no operative telemedicine-specific statutory enforcement framework; regulation therefore rests primarily on existing professional rules and general law. None of that has stopped six organisations from building working systems. But those systems rest on softer ground than their designers may realise. Either bring Act 564 into force, updated for how telemedicine is actually practised in 2026, or formally retire it and put the MMC’s own guideline on a firmer statutory footing instead. Leaving it exactly as it is, for a thirtieth year, is itself a decision — just an unstated one.
2. Plug into the National Health Interoperability Platform — do not build around it. Malaysia is already building the national interoperability layer and Lifetime Health Record that any serious telehealth service will eventually need to sit alongside. The decision that needs to be made now, explicitly, is whether new telehealth services are designed from the outset to connect to that national layer as it matures, or built in parallel and reconciled later — the way the six existing systems described in this piece never were reconciled with each other, or with the National Diabetes Registry. The second path produces a seventh silo. The first produces something that can actually scale.
3. Start with diabetes, but build the architecture for more from day one. Among the chronic-disease applications considered here, diabetes has one of the strongest and most practically applicable evidence bases, and the most standardised monitoring workflow — it is the right first module. But the system should be designed so that hypertension, hyperlipidaemia, weight, maternal care and other conditions can be added as separate modules without rebuilding from scratch. Do not mandate all conditions at launch; healthcare IT projects that try to do everything at once consistently fail to do anything well. Demonstrate the workflow works for real clinicians and real patients first, then expand. And whichever module launches first should report from its first cohort who it is failing to reach — rural patients, older patients, those without reliable connectivity — rather than treating equity as a phase-two concern.
4. Create a sessional fee schedule for remote monitoring, separate from Schedule 7. Reviewing blood pressure readings, glucose measurements, laboratory results and patient messages is clinical work even when there is no conventional consultation. Trying to pay for it through the same frozen per-consultation fee that has already created major sustainability concerns in private primary care — as set out in a companion piece — would import a thirty-three-year-old problem directly into a new service. One practical starting point would be a sessional schedule — paid per block of monitoring time rather than per encounter — which reflects what the work actually is. It also opens the role to nurses, pharmacists and allied health professionals doing protocol-driven monitoring alongside doctors, each paid appropriately through their own professional structures.
5. Build digital health competency into training at both ends, with principles before platforms. A 2025 Lancet paper found Malaysia has no national digital health curriculum framework. The MMC’s medical education committee should build one, focusing on principles — patient selection, limitations of remote assessment, consent, documentation, escalation — rather than platform-specific skills that will change quickly. For new graduates, some of this belongs in housemanship. For the doctors, nurses, pharmacists and allied health professionals already practising, the primary vehicle is CPD: short, self-paced online modules through the relevant professional bodies, deliverable without requiring time that nobody has. Build this deliberately, before the next crisis forces it to be done in a rush.
6. Define clinical governance, data governance and accountability before scaling. A national system needs explicit answers — before launch, not after — to questions the six existing systems have each resolved only locally: who is the data controller? What consent model governs secondary and research use? Who holds clinical responsibility for a patient enrolled in remote monitoring, and who is accountable when an alert goes unactioned? What are the cybersecurity obligations for a system carrying clinical data at national scale? How is patient identity matched reliably across systems? Without answers, interoperability creates shared risk without shared accountability.

In short

I wrote about this once, in 1997, at the very beginning of both this column and this technology. The computer I described that year is a museum piece, exactly as it should be. The telemedicine vision I described that year has, in its fragmented way, actually arrived — six times over, in six separate places, none of them connected to each other. Its own legal framework, Act 564, received Royal Assent in 1997 and has spent the years since sitting uncommenced. Its sibling law, passed the same year one Act-number apart, was rebuilt from scratch this year. Nothing about Act 564 required the same neglect. It simply received it.

What has been missing is not the technology, and it is not the will of individual institutions — six of them have demonstrated that independently. It is coherence: a national framework that connects what already exists, trains the people who will use it, gives the whole enterprise a legal footing that has now been waiting twenty-nine years for someone to decide what to do with it, and — most urgently — closes the gap between telehealth, regulation, national health data infrastructure, reimbursement and workforce preparation. Malaysia has spent years digitising individual pieces of healthcare. The next step should not be to digitise another piece in isolation. It should be to make the pieces talk to each other.