The NMPA issued the Guideline for Medical Device Clinical Trial Data Submission Requirements (draft) on September 15, 2021 for feedback. The feedback needs to be submitted by October 10.
Significantly referring to the Clinical Data Interchange Standards Consortium (CDISC) standards, the draft guideline details requirements on the original database, analysis database, explanatory documents and program codes.
Basic Principles for NMPA Clinical Trial Data Submission
When submitting clinical trial data, the three principles of “true”, “traceable” and “easy to read” should be followed.
- “True” means that the data should be consistent with that recorded in the Case Report Form, CRF. During the data entry process, data should not be changed without authorization. If there were abnormal values or clerical errors in the CRF table or the original record, the researcher should be notified in the form of a record (such as the query procedure), and the data can be updated only after the original record has been signed and confirmed by the researcher.
- “Traceability” means that the statistical results in the analysis database and clinical trial report can be reproduced from the original database according to the data and codes submitted by the registration applicant.
- “Easy to read” means that reviewer is familiar with the data variables and their structural relationships, and it’s easy to retest of statistical analysis.
Highlights from the Guidelines
Original Database
- A single original data set should collect variables under the same subject. It is not recommended to collect variables from different subjects into one original data set
- Each data set needs to include the subject’s unique identification variable to achieve the correlation between observations of the same subject in different data sets
- The naming of data sets and variables should follow the principle of “easy to read”. It is recommended to refer to the English or pinyin of the data sets or variables when naming them, and the actual meaning of the naming can be associated with the name.
Analysis Database
- The analysis data set can be named based on the corresponding statistical results
- Variable naming and variable length are the same as the original database requirements.
Explanatory Documents
- It is recommended to use an Excel file to list the data sets, variables, labels, assignments and their corresponding relationships in the original database and the analysis database in the form of a table.
- The analysis database description file needs to specify the generation rules of the derived variables, and the variables and calculation methods involved.
Program Codes
- The codes to be submitted mainly include: codes for generating randomization codes, codes for generating analysis databases from the original database, codes for generating statistical results from the analysis databases, etc. The relevant codes for adjusting the format or generating forms do not need to be submitted. The submitted code should conform to the usual programming format and programming specification, with a clear structure and easy to read.
The NMPA gives examples for each point listed above. For the full document please email info@ChinaMedDevice.com.
For a webinar on 2021 New Updates: China NMPA Medical Device Clinical Pathways, please click HERE
For clinical trial Q&A, please click HERE