NMPA Usability Guidance: How to fulfill the requirements

The NMPA Usability Guidance affects many manufacturers of medical devices and IVDs, as well as manufacturers of combination products, who want to market their products in China.

This article clarifies which devices must comply with the NMPA requirements regarding usability and what these requirements are.

1. Whom the NMPA Usability Guidance concerns

a) Included and excluded device classes

The guidance document of the Chinese regulatory authority NMPA has been applicable since October 8, 2024. It mainly concerns class II and III devices, including IVD and drug-device combination products. However, IVD reagents remain outside the scope of application.

b) Grandfathering

When approvals/registrations are renewed, the NMPA does not insist on additional data from usability engineering.

On the other hand, “grandfathering” does not apply to substantial changes to the users, the use scenarios, or the user interface. In this case, the manufacturers must at least evaluate the changed “aspects” by usability engineering.

2. Decisions to determine the NMPA requirements

The NMPA’s usability guideline requires manufacturers to check which requirements of its Usability Guidance must be met. These requirements depend on the answers to the following three questions (see Fig. 1):

1. Is the device a “high use risk medical device”?
2. Is the device novel (or is there a predicate device)?
3. Is there sufficient data available from a predicate device?

The following sections discuss these three decisions.

2.1 Decision on whether the device is a high use risk medical device

Manufacturers must first determine whether their device is a “high use risk device.” In this context, the NMPA defines three risk classes (see Tab. 1).

Class	Severity of Possible Harm	Recall Level
High Use Risk	Incorrect use may lead to serious injury or death	First
Moderate Use Risk	Incorrect use may lead to minor harm	Second
Low Use Risk	Incorrect use is unlikely to lead to harm	Third or none

The classification must be based on the “use risks,” not on the general risks that the device can cause. However, only “high risk medical devices” can also be “high use risk medical devices” (see Fig. 2).

For devices already on the market, or in the case of comparable devices already on the market, classification can also be based on “recall levels” instead of the severity of possible harm (see Tab. 1).

In most cases, high use risk medical devices are critical IVD or medical devices that

• are used in a completely new way,
• require a steep learning curve,
• are used by non-professional users, or
• are highly complex to use.

2.2 Decision on whether the device is novel

The second step is to decide whether a device is novel or whether there is a comparable predicate device. The NMPA assesses the equivalence using several characteristics (see Fig. 3).

For all attributes that are not equivalent, the manufacturers must conduct a usability test (see section 2.3).

2.3 Decision on whether sufficient data are available

The NMPA insists that the manufacturer proves equivalence regarding usability, i.e., provides sufficient evidence. This includes the following data, for example:

• Results of the usability validation of the predicate device
• If these are not available, an analysis of the post-market usability problems of the predicate device
• Analysis of the post-market usability problems of similar devices

Depending on the comparison result (see “Sufficient Evidence” diamond in Fig. 1), carrying out a usability validation/a usability test in China may be necessary. Tab. 2 describes which activities are required.

Comparison with the predicate device	No new “use risks”	New “use risks“
No differences to the predicate device	No usability validation necessary	Usability validation necessary for new risks
Differences to the predicate device	Usability validation is necessary for the differing aspects	Usability validation necessary for new risks and for the differing aspects

Suppose the predicate device based on the attributes in Fig. 3 differs (e.g., an additional user group) but poses no new use risks. In that case, the manufacturer must conduct a usability test with this user group in China.

The first line of Tab. 2 distinguishes whether there are new “use risks.”

3. Requirements of the NMPA Usability Guidance

3.1 Requirements for usability testing

In its Usability Guidance, the NMPA specifies many requirements similar to those of the FDA and IEC 62366-1, but they are not identical. For example, it requires:

1. Tests must be carried out in China
2. Representative test environment
3. At least 15 representative users from each “user group”
4. Testing of all critical tasks
5. Experienced usability testers, among whom no “Stakeholder” such as developers may be

The NMPA generally allows data from other countries to be used. However, approval in the home country is mandatory for approval in China, and the NMPA expects

• a report on “post-market use problems” for similar devices and
• a report highlighting the differences between the requirements in China and the other country and specifying which data is still needed.

3.2 Requirements for the analyzes

Report on Post-Market Use Problems

The report on known post-market problems must include

• what information was searched for (e.g., on which devices),
• when (date, period), how (e.g., search terms), and where (e.g., databases) the search was conducted,
• what the results of the search are (sources, evaluations, conclusions).

Report for Comparison

This report is comparable to other assessments of equivalence, as known from the FDA’s predicate devices.

The “Report for Comparison” should also state:

• The objective of the evaluation
• The compared devices
• The selection of the predicate device
• Data suggesting equivalence
• The description of the evaluation path (see Fig. 1)
• Conclusions
• Information about the person who performed the evaluation

Usability Test Report

The contents of the Usability Test Report correspond to what the FDA and IEC 62366-1 expect:

• Test object (device)
• Test subjects
• Test conditions, test environment
• Pass/fail criteria
• Use scenarios and tasks
• Results
• Assessments, conclusions

3.3 Requirements for the documents to be submitted

Depending on the use-risk classification (see Fig. 1), the NMPA expects different documents.

Evaluation Report on Use Errors

The NMPA expects this report for IVD and medical devices with low or moderate “use risks.” It specifies the following structure for this:

1. Basic information
2. Level of use risk including justification
3. Core elements (e.g., users, use scenarios, and user interfaces)
4. Analysis of post-market use problems of similar medical devices
5. Use risk management
6. Conclusion

Usability Engineering Research Report

Manufacturers are required to create a more detailed report for high use risk devices. The NMPA specifies the following structure:

1. Basic information
2. Level of use risk including justification
3. Core elements (e.g., users, use scenarios, and user interfaces)
4. Usability engineering process
5. User interface requirements specification
6. Use risk management
7. Verification and validation of the user interface (include materials described in section 3.2 here, depending on evaluation route) 
8. Traceability analysis of user interface
9. User training protocol
10. Conclusion

4. Summary and conclusion

The requirements of the NMPA Usability Guidance are mainly consistent with the requirements of the FDA and IEC 62366-1.

Nevertheless, there are some regulatory differences, and in particular, there is an obligation to conduct usability tests in China for devices with high use risk.

Whether the NMPA requirements are intended only to ensure patient safety or foreclose the market cannot be evaluated from the outside.

Once again, manufacturers are faced with the task of creating many market-specific documents (in particular reports), which are largely based on the same usability data. IT systems that generate these approval documents at the push of a button are helpful here.