On average, a doctor sees about 20 patients a day.

There’s a constant flow of medical data passing between patients, doctors, and hospitals.

So far, the way we collect evidence to prove a device’s safety, performance, and effectiveness is so limited.

We rely mostly on clinical trial data, academic studies, and published literature.

If manufacturers could tap into larger pools of medical records, more data would be available.

This could speed up clinical evaluations and get safer, more effective devices into the hands of patients faster.

Would you share your own medical data in exchange for access to better treatment?

Today we take a closer look at the current status of real-world data in China and what are the use cases.

🍜 China’s Attempts with RWD

NMPA just released a YY standard draft to elucidate the “Requirements for the collection of the Real-World Data (RWD) for Medical Devices”. The link is at the end of the post.

This complements “Technical Guidance on Real-World Data for the Clinical Evaluation of Medical Devices” which was published in 2020.

The guidance lays out principles for using RWD throughout the entire device lifecycle.

Unlike traditional clinical trials, real-world studies take place in everyday healthcare settings.

They usually involve larger patient groups and place fewer restrictions on who can be included.

The results may be more representative of real clinical practice.

It’s now possible to capture data on rare diseases and adverse events in the long term!

Of course, there are challenges.

The very features that make RWD broad and inclusive also create problems.

Data quality can vary widely. Misclassified or unstructured records are a headache.

These issues increase the risk of bias and make causal inferences weaker than those drawn from randomized controlled trials.

Still, the potential benefits of RWD are too important to ignore.

🍜 Quality Evaluation of RWD

So where does real-world data actually come from?

Common sources include registry databases, electronic health records (EHRs), health insurance claims, routine surveillance data, patient-reported outcomes, and even (legally) device-generated data from wearables or health apps.

NMPA has set out six key criteria for evaluating RWD quality:

Representability

Does the data cover the right population for your study objectives?

Completeness

Is the dataset sufficient to answer your study questions, including variables, sample size, and follow-up time?

Accuracy

Does it correctly reflect the patient’s health status, device use, and interventions, with a standardized collection and governance process?

Traceability

Can each device be uniquely identified and tracked across its full lifecycle in the data?

Consistency

Was the data collected using the same definitions and processes, with controlled versions of case report forms or other tools?

Repeatability

Can outcome measures be reproduced? If not, the data collection and analytic methods are not reliable.

The draft also goes into study design principles, study types, and statistical methods for planning real-world studies.

Let’s connect if you'd like to dive deeper into the details.

 🍜 Regulatory Applications Scenarios

You can utilize RWD in both pre-market and post-market clinical evaluations. Out of 11 major use cases identified, here are the 6 most common scenarios:

Clinical evaluation with substantially equivalent devices

RWD can provide evidence to support clinical evaluation through the substantial equivalence pathway.

It helps demonstrate that differences between your device and the predicate do not affect safety or effectiveness.

Here you can read more details on the possible clinical strategies in China.

Supplementary evidence for NMPA registration

You can supplement existing clinical evidence to support product registration.

This can help manufacturers avoid the need to conduct additional trials in China when overseas clinical trial data already exists.

Urgently needed imported medical devices

For devices under the special approval pathway, RWD can be added to existing data.

This option is useful for foreign manufacturers that already have clinical evidence, but need to address ethnographic differences that could impact performance or safety.

Local RWD can help bridge that gap.

External controls of single-arm clinical studies

In single-arm trial designs, comparable cases and clinical data can be drawn from high-quality real-world databases to serve as external controls.

These databases must have robust quality systems, and both sponsors and regulators need to confirm the data’s relevance and reliability.

Adjusting the intended scope, indications and contraindications

Local RWD can support updates to a device’s intended scope after market launch.

For example, expanding therapeutic benefits or incorporating certain off-label uses that are legally recognized in other jurisdictions.

Updating IFU to reflect clinical values

RWD can also support updates to clinical claims in labeling.

For instance, new data show a stronger causal links between physiological parameters and functional outcomes that were not captured in previous studies.

Other scenarios include post-market studies of conditionally approved devices, setting single-group target values, and long-term evaluations of safety and effectiveness for high-risk implants.

Does one of these pathways fit your device strategy in China? Book a call below if you want me to check whether RWD could save you time and resources.

Someone forwarded this email to you? Don’t forget to subscribe.

Spoiler❗: In the next post, I’ll sprinkle real-world examples of medical devices that successfully used RWD to secure NMPA approvals and what you can learn from them.

🍜 Bonus Reading

• New Industrial Standard Draft: "Requirements for Real-World Data Collection of Medical Devices” (Chinese)

• Official Announcement: “Technical Guidelines for the Use of Real-World Data in the Clinical Evaluation of Medical Devices” (Chinese)

• New Industrial Standard Draft: Medical Devices Part 1: Application of Usability Engineering to Medical Devices (Chinese)

Disclaimer: This content is created solely in my personal interest and does not reflect the views, opinions, or policies of my employer or any other organizations I am affiliated with. Neither I nor my employer is liable for any use of this content. This newsletter does not compete with my employer’s interests. Any references to regulatory updates or guidance are based on publicly available information and are not intended to infringe upon proprietary or confidential content.