Medical technology companies entering Asia usually discover the same thing in the same order. The product works. The first market takes longer than planned. The second market rejects a dossier that the first market accepted. By the third, the regulatory budget has doubled and the commercial team has already published claims the evidence does not support.
None of this is caused by unusually difficult regulators. It is caused by treating Asia as a single market with a single approval, and by confusing four very different kinds of claim that a medical device company makes about its own product.
Asia is not one regulatory market, and ASEAN is not one either
The most common planning error is assuming that regional harmonisation means regional approval. It does not. The ASEAN Agreement on Medical Device Directive commits member states to transpose a shared set of requirements into national law, covering essential principles of safety and performance, risk-based classification, conformity assessment, product registration, licensing of the party responsible for placing a device on the market, technical documentation and post-market obligations including adverse event reporting and field safety corrective actions.
What the directive does not create is a single regional filing. Each market still assesses and registers devices under its own authority, and the depth of implementation varies considerably between member states. Singapore and Malaysia operate mature frameworks. Several other markets have transposed parts of the directive while continuing to develop guidance, inspection capacity and post-market systems.
The World Health Organization’s model regulatory framework explains why this pattern is normal rather than exceptional. It recommends that countries build device regulation in steps — starting with basic controls such as enacting law and resourcing an authority, then expanding into establishment inspection and fuller post-market surveillance — and it treats reliance on the assessments of trusted authorities as a legitimate mechanism for regulators working with limited resources. Regulatory maturity is a trajectory, not a switch.
The Southeast Asian map at a glance
| Market | Authority | Framework basis | Planning note |
|---|---|---|---|
| Singapore | Health Sciences Authority (HSA) | Health Products Act; risk-based classification | Most detailed published guidance in the region, particularly for software and AI-enabled devices |
| Malaysia | Medical Device Authority (MDA) | Medical Device Act 2012 (Act 737); AMDD-aligned | Accepts MDSAP reports and certificates as quality system evidence for registration and establishment licensing |
| Thailand | Thai FDA | National rules revised toward AMDD | Common submission documentation used for higher-risk classes |
| Indonesia | Ministry of Health | AMDD-aligned national regulation | Expect a domestic review in its own right rather than automatic acceptance of foreign approvals |
| Philippines | Philippine FDA | Transitioning toward full AMDD implementation | Requirements have been moving; confirm current circulars before dossier preparation |
| Vietnam | Ministry of Health | Partially aligned national rules | Classification determines both pathway and timeline; low-risk classes move fastest |
Beyond ASEAN, the picture is different again. China’s NMPA, Japan’s PMDA together with its health ministry, Singapore’s HSA and South Korea’s MFDS all sit on the management committee of the International Medical Device Regulators Forum, while India’s regulator participates as an affiliate member. That matters practically: technical documentation structured around IMDRF conventions tends to travel better between submissions than documentation built around a single national template.
Software and AI have their own clock
For companies whose product is software rather than hardware, Singapore is the market that most clearly signals where the region is heading. HSA regulates software medical devices across the total product life cycle, with published expectations spanning quality management systems, pre-market registration, dealer licensing, change notification, post-market management, cybersecurity and machine-learning-enabled devices. The guidance has been revised repeatedly, most recently at the end of 2025, with expanded treatment of AI-enabled products, clearer handling of significant versus minor changes, and stronger cybersecurity expectations.
The commercial consequence is easy to underestimate. A hardware device is approved and then sold. A machine-learning product is approved with a defined intended use, a defined model and a defined change-control boundary — and every meaningful retraining or intended-use expansion has to be assessed against that boundary. Companies that plan a quarterly model-improvement cadence without planning a matching change-management process will eventually ship an update that their registration does not cover. Hospitals are increasingly alert to this, which is why procurement teams now ask vendors how model changes are governed, not just how accurate the model is.
Reliance is quietly changing how sequencing works
Quality system auditing has historically been the most duplicated cost in Asian market entry: the same factory, the same processes, audited repeatedly for different markets. That is beginning to change. Malaysia’s Medical Device Authority was recognised as an affiliate member of the Medical Device Single Audit Program with effect from 16 September 2025, and now accepts MDSAP reports and certificates as evidence of quality management system compliance in support of establishment licensing and device registration applications.
For a manufacturer already holding MDSAP certification, this removes a duplicated audit rather than a registration requirement. It is a reminder that the smartest sequencing decision is often made before any dossier is written: choose the audit and documentation format with the widest downstream acceptance, then file, rather than filing market by market and rebuilding the evidence each time.
The four-claim test
Regulatory strategy fails most visibly not at submission but in marketing. Medical technology companies routinely blend four separate claims into one sentence, and each requires completely different evidence.
- The regulatory claim. “Registered with authority X for intended use Y in market Z.” Evidence: the registration itself, held against a specific intended use and a specific market. It is not transferable between markets, and it is not a quality ranking.
- The technical performance claim. “Detects condition A with stated sensitivity and specificity in population B.” Evidence: validation studies, test reports, and honest disclosure of the population the device was validated in.
- The clinical outcome claim. “Improves patient outcomes.” Evidence: clinical investigation or published research designed to answer that question. Regulatory clearance does not, by itself, establish this. Neither does a customer testimonial.
- The commercial or operational claim. “Largest installed base in the region”, “first system of its type deployed in Asia”, “highest documented throughput.” Evidence: verifiable counts, audited records, independent documentation.
Almost every credibility problem in medical technology marketing comes from borrowing the authority of one claim to support another — most often using a regulatory approval or a commercial milestone to imply clinical superiority. Regulatory approval means a product met defined safety and performance requirements for a stated purpose. It is a floor, not a verdict on comparative effectiveness.
Where independent recognition fits — and where it does not
The fourth category is the one companies handle worst, usually by asserting scale claims with no verifiable basis. Yet operational and commercial milestones are often genuinely notable: a manufacturing volume, a deployment footprint across multiple countries, the scale of a screening programme a device supported, or a documented first-of-its-kind installation. These are objective, countable facts, and they deserve to be documented as such.
This is the appropriate role of independent record recognition. Organisations such as Asia Record document measurable achievements by companies and institutions, and reviewing what record recognition in Asia actually assesses is a useful exercise for any medtech leadership team tempted to publish an unverified superlative. A company that becomes an Asia Record holder has had a defined, measurable achievement assessed against submitted evidence. Medical technology companies with a documented operational milestone can apply for Asia Record recognition in the same way as any other business, and the application discipline itself is valuable: it forces a company to state precisely what was measured, how, and under what conditions.
What record recognition does not do is substitute for anything in the first three claim categories. It is not a regulatory approval, not a quality system certification, not an accreditation, and not evidence of clinical effectiveness. A record documents that something measurable happened at a stated scale. Presenting that documentation as proof of better medical outcomes is precisely the error the four-claim test exists to prevent, and regulators across the region treat implied clinical claims seriously regardless of the wrapper they arrive in.
A sequence that avoids rework
- Fix the intended use statement first. Classification, evidence requirements and permissible marketing language all flow from it. Changing it later means re-doing all three.
- Classify in each target market separately. Alignment to a common directive does not guarantee an identical class, and class drives cost and timeline more than any other variable.
- Choose the quality system audit with the broadest acceptance before commissioning it, not after.
- Build the technical file in a structure that travels, then adapt to national templates, rather than the reverse.
- Appoint the local responsible party deliberately. In most markets this entity holds or supports the registration; changing it later is an administrative event with commercial consequences.
- Design post-market processes before first sale. Vigilance reporting and corrective action obligations begin immediately, and gaps here surface during inspection, not during submission.
- Route every external claim through the four-claim test before it reaches a website, a tender response or a conference slide.
The mistakes that cost the most
Three patterns account for a disproportionate share of stalled Asian market entries. The first is assuming a foreign approval — from a European or North American authority — guarantees acceptance elsewhere; several Asian markets conduct their own review regardless. The second is treating regulatory affairs as a filing function rather than a product-planning function, which is fatal for AI-enabled products where the release cadence and the change-control boundary must be designed together. The third is building a market-entry plan around timelines quoted in commercial guides rather than around current published requirements, which move more often than annual planning cycles assume.
Underneath all three sits a simpler point. In medical technology, credibility is cumulative and specific. It is built by stating exactly what was approved, exactly what was measured, exactly where it was validated, and exactly what remains unproven — and by never letting one category of evidence stand in for another.
This article provides general information on regulatory and business practice in the medical technology sector. It is not regulatory, legal or medical advice, and requirements change; companies should confirm current obligations with the relevant national authority or a qualified regulatory affairs adviser.