Split "can I use it directly" into three separate questions
We field this one constantly: a manufacturer already holds a wheelchair test report and wants to know whether it can go straight into an FDA or a CE submission. There is no one-word answer, because the question is really three questions stacked on top of each other.
First, is the technical competence of the issuing laboratory accepted in the target market? Second, do the standards the report relies on -- and the sample configuration actually tested -- cover the scope the target market expects? Third, can the report, in the form it exists, be dropped into the submission dossier and be understood and relied upon by a reviewer or a notified body?
The answers to those three rarely line up. Most reports that come back for rework do not fail at the competence layer; they fail at the second and third. Once you separate the layers, every downstream judgement gets much easier. Below we start with the most common misconception about competence, then walk the US and EU routes separately, and close with two checklists -- one for before you ship samples, one for after the report lands.
One point up front: a laboratory accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope. It does not constitute a commitment regarding the market access outcome in any target market. Any claim of the form "our report will certainly clear FDA" or "will certainly clear CE" does not hold up.
The competence layer: what accreditation settles and what it does not
A widely repeated claim goes like this: "export reports must be issued by an accredited laboratory, or neither FDA nor CE will accept them." That sentence promotes a condition which meaningfully lowers risk into a legal threshold, and believing it costs real money -- we have seen manufacturers switch laboratories at the last minute purely to obtain a scope certificate, and others repeat work they had already finished.
The actual picture differs on each side.
On the US side, there is no statutory requirement that test data supporting a premarket submission be issued by a particular class of accredited laboratory. FDA does operate an accreditation scheme for testing laboratories, ASCA, but participation is voluntary: a laboratory may choose not to take part, and a manufacturer may use data from a laboratory that does not participate. What review looks at is still whether the data support the conclusion. Which standards that programme currently covers, and its current status, should be taken from the officially published information.
On the EU side, test data entering the technical documentation are assessed for reliability -- whether the method matches the standard, whether the equipment has metrological traceability, whether the sample is traceable, whether deviations are documented in writing. An accreditation certificate is strong supporting evidence of reliability, but the object under review is always the content of the report itself, and no certificate substitutes for completeness of content.
So the more accurate statement is this: having the laboratory's accredited scope cover the standard clauses you actually need materially reduces the chance that the report is challenged, that supplementary explanation is requested, or that retesting is required. It is not an access condition, and it promises nothing about the outcome.
That leads to a very practical verification step: check the scope, not the existence of a certificate. Accreditation certificates come with a scope annex, and the annex lists specific standards and clauses. Wheelchair standards come as a series -- stability, static strength and fatigue, brakes, electrical and control systems, information and marking all sit in different parts. What you need to confirm line by line is whether the parts of ISO 7176 you are about to test actually appear in that annex, and whether the item names correspond. Where the annex names only the series at a high level without listing individual parts, the real capability may not cover what you need. The same applies to the corresponding items under EN 12183 and EN 12184. Walking the annex line by line takes a few minutes and often removes an entire round of rework later.
The US route: pin down the product code first, then talk about the report
Many teams assume that "exporting to the US means a 510(k)", then reason backwards to "so I need a report suitable for a 510(k)". The starting point of that chain may already be wrong.
Manual wheelchairs, powered wheelchairs and mobility scooters are all regulated as medical devices in the US, but they do not share one classification entry. Manual and powered wheelchairs sit under different classification entries; the regulatory class, and whether premarket notification is exempt, must be confirmed separately for each and cannot be borrowed across. The two most common misjudgements in practice are mirror images: a manual wheelchair manufacturer sees a peer's powered product going through premarket notification and assumes the same applies to them; a powered product manufacturer hears that wheelchairs are a low-risk class and assumes the exemption covers them too. Either default can throw the whole budget and schedule off.
There is only one correct approach: take the product's actual construction, drive type and intended use, compare item by item in the FDA device classification database, land on the specific product code, then read the regulatory class attached to that code and whether it is listed as exempt. Manual and powered must each be searched once; where one family contains both manual and powered configurations, locate each configuration separately. The classification conclusion comes from the official database, and no second-hand account -- including this article -- replaces that step.
Suppose it is confirmed that premarket notification is required. The role of the report there is performance data. Review does not centre on which laboratory you used; it centres on whether the data support substantial equivalence between your device and the predicate. So the points most often queried in a report are these: the standards applied and the status of their current valid versions; whether the sample was the final production configuration; whether anything done during testing that departs from the standard was clearly written up; and how the pass or fail conclusion was reached.
The other case is that the product turns out to be exempt. A caution belongs here: the exemption removes one premarket procedure, not the obligation to do the work. Registration and listing, labelling requirements, quality system obligations, adverse event reporting and so on all remain, and those obligations can equally require you to hold evidence supporting the safety and performance of the product. What actually changes is where the report goes: it is no longer reviewed as part of a submission, and instead becomes design verification record and internal technical file, pulled out during system audits, customer audits, market surveillance and product liability disputes. Put differently, exemption reduces submission workload, not test workload. Manufacturers who drop the testing along with the submission usually get asked for it at some later point -- and by then both the sample state and the production line state have moved on.
The EU route: the report is part of the technical documentation, not the finish line
The logic on the EU side works differently. CE is not a matter of "being issued a certificate"; in essence the manufacturer declares that the product meets the requirements of the applicable regulation, and the test report is one piece of evidence supporting that declaration, filed inside the technical documentation.
For manual wheelchairs, powered wheelchairs and powered scooters, Europe has a corresponding harmonised standards framework, in which EN 12183 addresses manual wheelchairs and EN 12184 addresses powered wheelchairs and powered scooters. Both align closely in content with the ISO 7176 series, and many test methods are referenced directly. That is exactly why a well-planned test campaign can serve both routes at once -- provided the programme is scoped broadly at the planning stage, rather than testing to the ISO 7176 series first and going back to add EN afterwards.
The pitfalls in practice cluster in a few places.
First, harmonisation status and version correspondence. When the technical documentation cites a standard, it needs to state that the version is the current valid one and confirm the relationship of that version to the applicable regulation. A report that gives only the standard number without the version status will usually trigger a request for supplementary explanation during assessment.
Second, the report covers only part of the requirements. Applicable requirements for wheelchair products go beyond mechanical strength and stability, extending to materials, electrical safety, usability information and more. One mechanical performance report is not enough to carry an entire technical documentation file.
Third, sample and final product do not match. Testing before the design is frozen, testing a hand-built sample, or changing structural parts or the controller after testing -- all three discount the evidential weight of the report.
Fourth, risk management outputs and test items do not line up. The technical documentation expects testing to have a source: the hazards you identified should map to a corresponding means of verification among the test items. If the risk list names a hazard and the test report contains no matching verification, that gap is easy to spot during assessment.
A boundary is worth drawing here: the four points above are about what the report itself should look like. Whether a notified body is involved in this route, in what circumstances involvement is required, how to handle questions after submission, and what a system audit looks at, are separate topics we do not open here; there are dedicated articles in our technical notes section. The position of this article is simple: whether or not a third party is involved, the manufacturer has to fill these four gaps first.
One table: what each route expects from the report
| Dimension | US route | EU route |
|---|---|---|
| Role of the report | Performance data supporting substantial equivalence (where submission is required) | Evidence of conformity within the technical documentation |
| Is laboratory accreditation a threshold | Not a statutory threshold; the relevant accreditation scheme is voluntary | Not a statutory threshold; what is assessed is data reliability |
| Is there a designated laboratory list | No general designated list | No general designated list |
| Commonly cited standards framework | Mainly the ISO 7176 series | EN 12183 / EN 12184, aligned with the ISO 7176 series |
| Where the classification conclusion comes from | Query the official classification database by specific product code; confirm manual and powered separately | Apply the classification rules of the applicable regulation; the official position governs |
| Requirement on sample state | Must be the final production configuration | Must be the final production configuration and consistent with the technical documentation |
| Common reasons for pushback | Undocumented deviations; questionable sample representativeness | Incomplete item coverage; version status not stated; disconnect from risk outputs |
Setting the differences side by side helps with comparison, but what actually decides the outcome is the detail below.
How to define the model family and choose the worst-case configuration
This is the single most expensive part of the exercise and the one most likely to force rework. Manufacturers typically have a whole family: seat widths, seat heights, backrest heights, wheel sets, battery capacity tiers and controller models all differ. Testing everything is unaffordable; testing one configuration risks not covering the family.
Working from the load path gives you a broad direction. Stability-type items depend on the relationship between the centre of gravity and the support polygon, so within a family the combination with a relatively higher centre of gravity, a relatively narrower support base, and adjustable parts set to their least favourable limit positions is normally the one to cover. Strength and fatigue items depend on how load travels from the seat into the frame and then into the wheels; stress concentrations tend to appear at welds, section transitions and the locking features of adjustable parts, so configurations with thinner-walled structural members, longer spans and longer adjustment travel deserve attention. Electrical and control items depend more on the electrical architecture than on mechanical dimensions, so grouping by controller and drive solution is usually sufficient, without expanding across every mechanical size.
This needs to be stated plainly: the above is engineering analysis based on structure and load path, not a statistical conclusion. The actual worst-case configuration has to be determined against the specific product's structural calculations and design inputs, and the reason for the selection must be written into the test plan. Reviewers and notified bodies do not object to representative samples, but they will expect you to explain why this sample represents the whole family. How well that explanation is written decides directly whether you ship one unit or several.
A practical suggestion: when preparing the test plan, write the family segmentation rationale and the worst-case selection rationale as a standalone document and archive it alongside the report. When models are added later, as long as the new model falls inside the envelope already argued, the additional verification effort is much smaller. Without that document, every added model means explaining the whole thing again from scratch.
When old data can be reused and when it cannot
The other half of "can I use the report directly" is "can I test less than before". The criteria collapse into three preconditions, and reuse is only on the table when all three hold.
One, methods correspond. The standard behind the old report and the standard required by the target route must correspond item by item in test method, loading arrangement and judgement principle. The ISO 7176 series aligns closely with EN 12183 and EN 12184, and the GB/T 18029 series corresponds closely in technical content to the ISO 7176 series, which gives reuse a basis; but alignment is not equivalence, and you still have to compare item lists against the current valid versions to catch requirements unique to the target route.
Two, samples correspond. The sample used in the old testing must match the product now being submitted in configuration, key components, software and control strategy, and production batch. Where structural parts have been changed, or the controller or battery solution swapped, the affected items essentially have to be redone. Traceable sample identification must be findable in the old report; "same model" as a bare assertion will not carry it.
Three, the format is acceptable. The same data can face different expectations across markets regarding the issuing party, report format and language. The domestic market has its own requirements for who issues a test report, so even where the technical content is reusable, the report itself usually has to be reissued to domestic requirements rather than handing over the export copy.
Of the three, method correspondence is the one most often overstated: manufacturers see similar item names in two standards and conclude the whole package is reusable. The safer approach is to have the laboratory take the old report and the target route's item list and produce a clause-by-clause mapping with three verdicts -- directly usable, supplementary testing needed, full retest needed -- and build the budget from that. If you want a first look at how the standards relate to one another, see our standards index. Specific test parameters and acceptance limits are governed by the current valid version of the standard text.
The material needed before samples ship is usually more than expected
Plenty of projects stall at the intake stage, not because the product is inadequate but because incomplete documentation makes it impossible to fix the test plan. We suggest assembling the following before you even contact a laboratory.
Product level: complete description of the product structure, model list with a differences comparison, key component list (frame material and specification, wheel sets, motor and controller, battery and charger solution), assembly or exploded drawings, nameplate and label artwork, and a final or near-final version of the user manual.
Intended use level: intended use description, target user population, use environment limitations, contraindications and warnings. This directly drives the selection of applicable standards and also affects consistency in the submission documents downstream.
Sample level: samples must be in the final production state, with serial numbers or traceable identification. Some destructive tests consume the sample, so the quantity to ship needs confirming in advance. If both export routes are to be served, settle this in one pass rather than shipping in two batches whose samples are not fully identical.
Existing documentation: if some testing has already been done, hand over the old reports. Some items can be judged as usable existing data and need not be repeated; but that judgement belongs to the laboratory, based on the standard requirements and sample consistency, and cannot be self-declared.
After the report arrives, check these as well
Receiving the report is not the end of it. Run this checklist before the report goes into the submission dossier.
| Check | Why it matters |
|---|---|
| Does the scope annex go down to the specific parts tested | Where only the series is named without the individual parts, coverage needs re-confirming |
| Do all tested items fall inside the accredited scope | Items outside scope are more easily challenged on evidential weight |
| Do the sample identification and the frozen product state agree | Report, technical documentation and user manual must be mutually consistent |
| Does the standard citation state the current valid version | A bare standard number with no version status commonly triggers a request for clarification |
| Is traceability information given for the test equipment | Metrological traceability is a basic element of laboratory competence and a focal point in data reliability review |
| Are deviations from the standard method documented in writing | A deviation is not the problem; an undocumented deviation is |
| Does the wording of the conclusion match the basis for judgement | A report conclusion must not assert beyond the coverage actually tested |
| Are the Chinese and English versions consistent | Export submissions often need an English report, and translation discrepancies invite avoidable questions |
The first two rows were added deliberately, for the reason given earlier: most manufacturers only look for an accreditation mark and rarely turn to the scope annex to compare line by line. The items that actually get questioned during assessment are exactly the ones in that annex that do not line up.
Conclusion: there is no universal report, but there is a plan that works first time
Back to the original question. There is no single wheelchair test report that satisfies every market, because different markets expect different things from the issuing party, the coverage and the presentation. But there is a way of working that compresses duplicate testing cost to almost nothing: lay out all target market requirements at the test planning stage, scope the items, select the samples and fix the report format to the broadest of them, then test once and produce several outputs.
The expensive alternative is to run one campaign for a single market, discover months later that another market is on the table, and find the samples dismantled, the structure changed and the original deviations unrecorded -- leaving nothing to do but start over. We see this often, and it is nearly always avoidable at the planning stage.
How we can support you
SUNGO Mobility Testing Lab is the dedicated wheelchair and mobility aid testing lab within our group, with laboratories in Shanghai and Hefei, accredited by CNAS, CMA and IAS (USA), covering manual wheelchairs, powered wheelchairs, mobility scooters, walkers and crutches. To restate the point once more: an accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope; it does not constitute a commitment regarding the market access outcome in any target market, nor does it represent any review or assessment conclusion in that market.
We can help you confirm the applicable standards, compare the accredited scope against your item list clause by clause, justify the family segmentation and worst-case configuration, prepare and run the test programme, and issue Chinese and English reports in the form export submissions normally expect. For the item breakdown, see powered wheelchair testing and manual wheelchair testing; a broader view of what we cover is on the services page.
If you are planning export testing, or you already hold a report and are unsure whether it is sufficient, send us the product documentation and we will start by telling you what has to be added and what can be saved. Call +86 132 4819 8029, or request a quote online.