Medical Device Classification in Malaysia: Class A, B, C and D Explained

Classification looks like an administrative step. It is actually the decision that determines how much your registration will cost, how long it will take, and how much technical evidence you will need to produce. Getting it wrong at the start is expensive to unwind later.

Why classification drives everything

Malaysia applies a risk-based classification system under the Medical Device Act 2012 (Act 737), broadly aligned with the international harmonised approach used across ASEAN and by the former Global Harmonization Task Force. Devices fall into four classes, A through D, in ascending order of risk.

The class you land in determines the depth of conformity assessment a Conformity Assessment Body must perform, and therefore the volume of technical evidence you must produce, the fees you will pay, and the time the process will take. A Class A device and a Class D device are not the same project with a different label — they are different projects.

The four classes

Class A — Low risk

Generally non-invasive devices, or devices that contact intact skin only, with no measuring or sterile function. Typical examples include examination gloves, surgical retractors, hospital beds, wheelchairs and stethoscopes.

Class A carries the lightest evidence burden. Note, however, that Class A devices which are supplied sterile or which have a measuring function attract additional scrutiny, because those attributes introduce risks the base classification does not account for.

Class B — Low to moderate risk

Typically invasive devices used short-term, or devices in contact with breached skin. Examples include hypodermic needles, suction equipment, and many single-use surgical instruments.

Class C — Moderate to high risk

Devices with a significant potential to cause harm through malfunction or misuse: infusion pumps, orthopaedic implants, ventilators, and devices that administer medicine or energy in a manner that carries meaningful risk.

Class D — High risk

Devices that are life-supporting, life-sustaining, permanently implantable, or otherwise carry the highest potential for harm — heart valves, implantable defibrillators, coronary stents. Class D attracts the deepest assessment and the largest technical documentation requirement.

In vitro diagnostic devices are classified separately

IVDs — tests performed on samples taken from the body — use a parallel A-to-D classification with its own logic. IVD classification turns on the consequence of an incorrect result, for both the individual and public health. A test for a transmissible agent that could cause a life-threatening condition sits at the top of the scale; a general laboratory reagent with no specific diagnostic claim sits at the bottom.

Applying the general medical device rules to an IVD is a recurring and avoidable error.

What drives the classification decision

Classification is determined by applying a defined set of rules to the device's characteristics. The factors that matter most are:

• Intended purpose — as stated by the manufacturer, not as a distributor imagines it might be used. This is the anchor for the whole analysis.
• Duration of contact — transient, short-term or long-term.
• Degree of invasiveness — non-invasive, body-orifice invasive, or surgically invasive.
• Whether it is active — does it depend on a source of energy other than the body or gravity?
• Which part of the body it affects — devices in contact with the central circulatory or central nervous system classify higher.
• Whether it delivers or exchanges energy or substances with the patient.

Intended purpose is the anchor. Change the intended-use statement and you can change the class — which is why the statement must be written carefully and then applied consistently across labelling, IFU and dossier.

Grouping: the decision people forget

Classification tells you how deep the assessment goes. Grouping tells you how many registrations you need — and it has a direct effect on cost.

Rather than registering every individual product code separately, related devices may be registered together under defined grouping categories:

A distributor with 80 product codes may need far fewer registrations than they assume — or considerably more, if the codes do not genuinely satisfy the family criteria. This analysis routinely changes project budgets by a significant margin in either direction, and it is worth doing properly before you commit.

The three mistakes we see most

1. Copying the classification from another market. A device that is Class IIa in Europe or Class II in the United States does not automatically map to a specific Malaysian class. The systems are related but not identical, and the mapping must be reasoned, not assumed.
2. Classifying on the product, not the intended purpose. Two physically similar devices with different stated intended uses can classify differently. The intended-purpose statement governs.
3. Over-classifying to be safe. Deliberately claiming a higher class to avoid argument commits you to a deeper assessment, a larger dossier, higher fees and a longer timeline — permanently. It is not a conservative choice; it is an expensive one.

What to do next

If you are at the start of a registration project, resolve classification and grouping before anything else. Every downstream decision — which CAB to approach, what evidence to gather, what the project will cost, when you can realistically launch — depends on it.

Our MDA registration service begins with exactly this analysis, and we provide a written classification rationale you can defend in your submission. If you would like a view on where your device sits, send us the intended-use statement and product literature.