How Thai FDA Classifies In Vitro Diagnostic Devices

In vitro diagnostic devices occupy an odd corner of Thai medical device regulation. They sit under the same act that covers surgical instruments and implants, yet the logic that determines their class has almost nothing to do with the physical device and everything to do with what happens if a test result is wrong. Manufacturers who apply the general classification rules to an IVD, rather than the diagnostic-specific ones, routinely misjudge where their product lands, and the miscalculation shows up as a budget and timeline surprise partway through registration.

One Act, No Separate IVD Statute

Thailand has never created a standalone IVD law. In vitro diagnostics are regulated as medical devices under the Medical Device Act B.E. 2551 (2008) and its 2019 amendment (B.E. 2562), and both diagnostic reagents and the instruments that read them meet the act's definition of a medical device once they are intended to examine specimens from the human body for information about a physiological or pathological state.

The Medical Devices Control Division (MDCD) of the Thai FDA enforces this, applying the ASEAN Medical Device Directive (AMDD) as the substantive classification and technical documentation standard. A classification determination reached under AMDD therefore carries direct weight in a Thai submission, though MDCD keeps discretion to layer on additional local requirements, particularly around labeling language and post-market reporting. The practical consequence is that there is no simplified "reagent" or "lab supply" shortcut around the Medical Device Act, even for products that look clinically unremarkable on their face.

Why Annex 3 Exists Separately From the General Rules

AMDD sorts devices into four classes, numbered 1 through 4 and lettered AMDD A through D, with Class 1/A the lowest risk and Class 4/D the highest. Most devices are classified under Annex 2's general rules, which weigh factors like invasiveness, contact duration, and active versus non-active status. IVDs do not use those rules. AMDD Annex 3 supplies IVD-specific criteria that apply instead, built around a different central question: not how invasive the device is, but how much harm a false result could cause.

A false positive or false negative diagnostic finding can trigger inappropriate treatment, delay urgently needed care, or create a public health hazard, and Annex 3 translates that risk into classification criteria built around the intended-use population, the condition being detected, and the clinical severity of getting the result wrong. Manufacturers who default to the general rules out of habit tend to underclassify. A clinical chemistry analyzer might look low-risk under a general-rule lens, but Annex 3 puts it at Class 2 because it feeds clinical decisions; the analyte and context of use drive the classification, not the physical instrument.

What Separates Class 1 From Class 4

Class 4, the highest-risk tier, covers devices used for HIV detection, ABO and Rh blood typing, and detection of other life-threatening infectious diseases, where an erroneous result is likely to cause death or serious irreversible harm. Blood bank and transfusion safety testing is the textbook case: a false-negative HIV result on a donor unit carries population-level consequences, and a false-positive blood type result can cause a fatal transfusion reaction.

Class 3 covers infectious disease screening in defined clinical populations where an error is serious but not immediately lethal, such as rubella, cytomegalovirus, and toxoplasma assays in prenatal monitoring, PSA testing, and oncology biomarker assays that directly steer treatment decisions. The common thread is that a missed or wrong result meaningfully changes patient management where the downstream harm is substantial. Class 2 covers non-life-threatening infectious disease detection, routine clinical chemistry analyzers, and, critically, every IVD intended for self-testing by a lay user, regardless of how low-risk the underlying analyte would otherwise be. A home glucose monitor and a consumer flu test both land in Class 2 under this rule because a lay user cannot be expected to interpret a borderline result or recognize device malfunction the way a trained lab professional would. Class 1, the lowest tier, covers general-purpose lab instruments not tied to a specific analyte, plain sample-collection devices like tubes and swabs used without added reagents, and controls or calibrators that carry no diagnostic purpose of their own. These products do not produce a result that can be directly acted on; they support testing without being the point of clinical decision-making.

The Self-Test Rule Catches More Products Than Manufacturers Expect

The automatic Class 2 floor for self-test IVDs deserves particular attention because it snags products manufacturers often assume are low-risk. It applies whenever the intended use statement, labeling, or marketing materials show the device is designed for a lay person outside a professional healthcare setting; intended use is the operative concept, so a device that could theoretically be used by a professional but is marketed to consumers gets classified by its consumer-facing use. A pharmacy-shelf rapid antigen test for influenza is Class 2 even though the same assay run in a hospital lab might otherwise sit at Class 1 or low Class 2 on analyte risk alone. The regulatory logic is that the absence of professional oversight raises the risk of misuse, misinterpretation, and delayed care, and that elevated operational risk justifies the higher floor. A manufacturer entering Thailand with a consumer diagnostic should plan on Class 2 from the outset; arguing for Class 1 treatment on the basis that the analyte itself is low-risk rarely succeeds once the labeling makes clear the device is meant for home use.

How Much Evidence Each Class Requires

Technical documentation scales with classification. Class 1 needs a technical file demonstrating conformity with general safety and performance requirements, but not clinical performance studies as a standard component; the burden here is mostly administrative, labeling confirmation, a declaration of conformity, and basic manufacturing documentation. Class 2 adds analytical validation as a standard expectation, with reviewers looking for evidence the device performs within its claimed measuring range and that precision and accuracy have been characterized under conditions representative of actual use; self-test Class 2 IVDs also need usability data showing lay users can correctly run the test and interpret the result unassisted.

Class 3 and Class 4 carry the heaviest evidence load: analytical validation covering sensitivity, specificity, precision, and accuracy, drawn from studies conducted in the actual target population rather than borrowed from a different region's reference data, since pathogen variant prevalence and patient characteristics in Southeast Asia may differ materially from a European dataset. Reference material traceability becomes a specific requirement at these two tiers; where WHO reference preparations exist for an infectious disease assay, manufacturers need to demonstrate metrological traceability to them, with calibrator values expressed in SI or WHO-established international units. Missing traceability is a common cause of deficiency notices at technical file review.

Pathways, Fees, and Where Manufacturers Get Caught Out

Registration runs through two pathways determined entirely by the classification reached above: notification and full licensing. Class 2 and Class 3 IVDs use notification, where the importer or local license holder submits a technical dossier and MDCD issues a product notification certificate; the total government fee is 49,000 THB, covering application and certificate issuance but not professional service fees or dossier preparation costs. Class 4 IVDs go through full product licensing, a more intensive review that can include a request for local bridging studies, at a total government fee of 74,000 THB, though the clinical evidence package for HIV assays, blood typing reagents, and similar high-risk diagnostics typically costs far more in time and resources than the fee itself. Class 1 IVDs follow a lighter self-declaration pathway with a correspondingly lower fee, but the self-test elevation rule means many products a manufacturer initially assumes are Class 1 will in practice land at Class 2 or above once the intended use analysis is applied.

Every IVD registration, regardless of class, must be held by a licensed importer or local license holder established in Thailand; foreign manufacturers without a Thai entity need to appoint a local representative to hold the registration on their behalf, and that local holder carries the ongoing accountability for adverse event reporting and any field safety corrective actions. Getting the classification right before the dossier is built, rather than discovering the correct class midway through review, is what keeps a registration timeline predictable. DeeMED Consulting supports IVD manufacturers through this classification analysis and the notification or licensing process that follows, including Annex 3 classification review and registration pathway planning.