Defining the NMPA Registration Test List for Wheelchairs

Correcting a common starting mistake

Teams handling a domestic registration for the first time usually copy the part titles of the GB/T 18029 series into a table, tick them off one by one, and bring that table to us asking for price and lead time. The direction is right; the order is backwards.

The direct basis for registration testing in China is the product technical requirements, and the test report has to correspond clause by clause to the performance indicators and test methods written into them. Standards are an input when drafting the technical requirements -- they are not the test list itself. That distinction sounds like word play but its effect is large: ticking off standard parts tends to buy a stack of items irrelevant to your product, while one vaguely worded indicator with no stated method leaves the testing body unable to produce a number, and the report ends up incomplete.

So defining the list really means answering two questions: which indicators belong in the technical requirements, and which of those indicators need a test report to demonstrate them. Answer both clearly and the list converges on its own.

One more thing to settle before starting: which regulatory class the product falls under for NMPA registration in China. Wheelchair products are regulated as medical devices, but the specific catalogue entry and the class must be taken from the current medical device classification catalogue and classification determination documents; a peer's older experience is not a substitute. Change the class and the registration route, the accepting authority level and the documentation requirements change with it, with the test list being only one of the affected pieces. The same logic applies to other markets: if this product is also going down the US route, whether a 510(k) submission is required or whether it is exempt must be confirmed from the official classification database for the specific product code, and cannot be inherited from a similar product.

Fix the product boundary or the list has no baseline

Before drafting the technical requirements, describe the product precisely. If the following are unsettled, every later step will need redoing.

Drive and structural type. Manual wheelchairs, powered wheelchairs and powered scooters are three different item paths and cannot be written as one. Manual products do not involve assessment in the powered drive and control system direction; powered products pull in a whole chain of additional content covering speed and deceleration, control systems, electromagnetic compatibility, batteries and chargers. For the difference in item structure, compare manual wheelchair testing with powered wheelchair testing.

Presence or absence of functional features. Is it foldable; does it have tilt-in-space and elevating legrests; does it have powered posture adjustment; are armrests and legrests detachable; is it declared suitable for use as an occupant seat in a motor vehicle. Every "yes" adds something to the list, especially posture adjustment, which pulls on both stability and strength at the same time.

Coverage of models and specifications. Domestic registrations often cover multiple specifications at once. The task is not to submit every specification for testing, but to state the differences between specifications clearly and justify which one or ones represent the whole. Difference points usually include seat width and depth, frame material and tube diameter, drive and battery configuration, and load capacity tiers. If the justification is thin, review will require supplementary testing, and at that point the schedule is beyond your control.

Intended use and claims. User weight limit, applicable environment (indoor, outdoor, road conditions), whether obstacle capability is claimed, whether extended sitting is claimed. The more you claim, the more you must demonstrate. Many companies find that a promotional sentence dropped casually into the user manual has turned into a performance indicator requiring data.

Three tests for filtering items out of the standard series

The GB/T 18029 series corresponds in technical content to the international ISO 7176 series, with parts covering static and dynamic stability, brakes, overall dimensions and mass, strength and fatigue, climatic environment adaptability, flammability of materials, electrical and control systems, electromagnetic compatibility, batteries and chargers, and marking and accompanying documentation. The applicable scope and item structure of each part can be checked against the testing standards page. When filtering, apply three tests.

First, does the applicable scope of that part cover this product. Some parts address powered products only, some manual products only, and some carry general requirements for all wheelchair products.

Second, does the product actually have the functional feature or the claim that the part assesses. With no powered drive, speed and control assessment simply does not apply; with no battery, the battery and charger direction does not apply.

Third, where an item is judged not applicable, can you give a verifiable justification. This step is easy to skip. When review sees a direction missing from the list, it does not assume you had a reason -- it asks why. Writing the not-applicable rationale into the technical requirements preparation notes is far less work than explaining it afterwards.

The table below is the usual way of deriving item directions backwards from product features, and works as a reference tool while filtering.

For powered products the items are coupled

This section is engineering analysis rather than a statistical conclusion, but it explains a class of rework that appears frequently in real projects.

For a powered wheelchair, hill-holding capability, braking behaviour and dynamic stability look like three independent items but share one load path: the friction state between drive wheels and ground, the acceleration and deceleration curves the controller commands, and the spatial position of the centre of mass. Change the tyre tread pattern once, move the battery from under the seat to the rear, or shift the controller deceleration parameter by one step, and results in all three directions move together.

The conclusion follows: a configuration change cannot be handled by retesting only "the item that changed". We have seen a customer relocate the battery to improve range and assume that resubmission needed only a battery-related report, only to find the stability and braking data no longer matched the original report -- effectively starting over. So change assessment should follow the load path, not the count of report volumes.

By the same logic, powered scooters are closer to a vehicle in structure but share the item logic of powered wheelchairs; the differences come mainly from overall dimensions, the steering mechanism and the boundary conditions imposed by the use scenario. For the item structure see mobility scooter testing.

Biocompatibility: scope it, do not run the full battery

What genuinely enters biocompatibility evaluation on a wheelchair is the parts in prolonged skin contact: cushion fabric, backrest, armrest covering, handrims, legrest support surfaces. The ISO 10993 series starts its evaluation from the classification of contact site, contact nature and contact duration; frame the contact clearly and the scope shrinks considerably. Structural parts and covers with no body contact generally stay outside the evaluation.

A frequent pitfall here: the same model changes fabric or covering material supplier between batches, the material grade changes, and the technical requirements are left untouched -- so the sample used for registration testing no longer matches later production parts. The evaluation route does allow material characterisation and assessment of existing data first, with testing added only where genuinely necessary, but that reasoning has to be captured in verifiable documents, not mentioned once in meeting minutes.

What existing export test data is good for

Plenty of companies export first and register domestically afterwards, and already hold reports issued to EN 12183 or EN 12184. Reuse is conditional on standard version, product configuration and sample batch all corresponding; if any one fails to line up, the work has to be redone. What can usually be carried across directly is not the report but the failure experience inside it: which weld cracked first, which adjustment mechanism developed play after fatigue. Use that to guide sample selection and pre-testing for the domestic submission. The full framework for judging cross-route report reuse is covered in a separate article and is not repeated here.

Registration testing can be self-tested or outsourced

This section starts by correcting a position that is outdated yet still circulating: that registration testing must be sent to a testing body holding the relevant qualification. Under the current Provisions for the Administration of Self-Testing for Medical Device Registration, a registration applicant may carry out registration testing itself or commission a testing body to do it; the two routes run in parallel. The direct consequence of following the old position is budgeting and scheduling entirely around outsourced testing, giving away a share of work that could have been absorbed internally.

The precondition for self-testing is demonstrable capability. Self-testing is not a matter of buying the equipment and running the test. The applicant needs equipment, personnel, methods and a quality management system matched to the items being self-tested; the applicant is responsible for the results; and the self-test report together with the corresponding capability evidence must be included in the submission and is subject to review and possible on-site inspection. Where the capability evidence falls short, the self-test report will not stand up during review, which costs more time than outsourcing from the start.

The key point for outsourcing is scope, not the mark. The commissioned party must issue the report within its qualification-accredited scope; for items outside that scope, the report may not be relied upon at submission even if the equipment can perform the test. So once you have the testing body's qualification annex, compare coverage item by item against the intended list, rather than merely confirming that CMA and CNAS marks exist.

The two can be mixed. A common arrangement keeps routine items within internal capability in-house and outsources fatigue, electromagnetic compatibility, and battery and charger items, which demand more of equipment and facilities. That combination is permissible, but two things need watching: the version of the technical requirements and the sample state must be identical across the self-tested and outsourced portions; and the same item must not end up with two sets of data telling different stories, which leads straight to a review query.

SUNGO Mobility Testing Lab is accredited by CNAS, CMA and IAS (USA), with laboratories in Shanghai and Hefei. To be explicit: an accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope and does not constitute a commitment regarding the market access outcome in any target market. Whether a product completes registration depends on the overall quality of the submission and the review opinion; the test report is one part of that, and we do not promise that any given report will be relied upon.

When choosing a testing body, beyond checking that the qualification scope covers the intended items, confirm whether it can work through the technical requirements clause by clause and whether it can give a failure diagnosis during pre-testing rather than only a pass-or-fail statement. For companies registering for the first time, that last point usually affects total lead time more than price does.

What to have ready before samples ship

Once the list is fixed, whether the work starts on schedule is decided by submission preparation. The table below follows the order we actually check at sample intake. It applies equally to companies on the self-testing route, except that the material has to form verifiable internal records rather than being handed over with the sample.

Where schedules typically stall

Purely in terms of time, climatic environment adaptability and fatigue items occupy test stations for longer, while strength and impact items may need spare samples because samples get damaged. Placing both groups at the tail end of the registration plan leaves no buffer if anything goes wrong. The sensible approach is to launch the long-lead items as soon as the technical requirements are finalised, and to run documentation and marking corrections in parallel with testing.

The other common time sink is change. If the technical requirements are amended during review and the amendment touches the basis on which a completed item was judged, the report has to be reissued. Reducing that risk is mainly a matter of working ahead: write the indicators and methods rigorously while drafting the technical requirements, and avoid unworkable phrasing such as "carried out with reference to the relevant standards".

How we can help

SUNGO Mobility Testing Lab has long handled testing for manual wheelchairs, powered wheelchairs, powered scooters and related mobility aids, with laboratories in Shanghai and Hefei. In the run-up to registration we can help you map the applicable item scope, derive from product features the performance directions that belong in the technical requirements, assess which parts of existing export reports are convertible, and arrange pre-testing before formal submission to reduce the chance of rework. Specific parameters and acceptance limits are governed by the current valid version of the standard text, and the actual programme is confirmed item by item against your product structure and target markets. A fuller view of what we cover is on the services page.

To confirm an item list or arrange sample submission, call +86 132 4819 8029, or request a quote directly. Send us your product documentation and we will come back with item recommendations and a lead time assessment.