The project

The client is an international medical device company. Its product is approved in other major markets and is now seeking NMPA registration. The client team drafted the clinical evaluation report (CER) and asked us to review it. Identifying details have been removed.

1. The equivalence argument had gaps

Several items in the equivalence comparison table were marked “Same” when they were not fully the same. The larger problem was the supporting evidence. Almost every equivalence claim cited only the IFU and “the Technical File”, with no report numbers or document titles.

“Technical File” is a CE-marking concept. In a Chinese submission it is too vague to verify. Each equivalence claim needs a CH directory number and the title of the specific report. This problem ran through almost the entire comparison table.

2. The subject device and the comparator were not clearly separated

The subject device added several functional modules to the comparator device. Throughout the report, descriptions of the two devices were mixed. Structure, technical parameters and intended use did not show which parts were the same as the comparator and which parts were new.

Under NMPA review logic, new features need their own safety and effectiveness argument. If the report cannot draw this line, the reviewer cannot judge whether the submission is complete.

3. Safety and effectiveness indicators did not match

The indicators listed in the executive summary did not correspond to the conclusions of the literature analysis in the body of the report. The chapters were written separately. The link from literature evidence to the safety and effectiveness conclusion was missing, so the reviewer would have had to build it. A CER should make that link for the reviewer.

4. Citation numbering was unreliable

Citation numbers differed between sections. It was not possible to confirm that the same number referred to the same article in different chapters, and the total reference count also needed checking. Inconsistent citations make a reviewer question the reliability of the whole document.

What the review showed

The client team had strong product knowledge. The problems did not come from a lack of expertise. A CER follows its own logic: it is an argument written for NMPA reviewers, whose job is to find gaps in the reasoning. The main quality test is therefore not whether all the content is present, but whether the reviewer can follow the argument from evidence to conclusion.