Real-world Data

1. The source of real-world data

　a.Real-world data refers to data from the real medical environment, outside of traditional clinical trials, and reflects the health status of patients and the process of medical services in actual diagnosis and treatment. Real-world research is a research that focuses on related scientific issues, comprehensively uses epidemiology, biostatistics, evidence-based medicine and other multidisciplinary methods and technologies.

　b.Common sources of real-world data include: registration databases, hospital electronic medical records, regional health and medical data, medical insurance data, health records, etc. In addition to the above-mentioned situations, real-world data applicable to medical devices can also include data generated during the production, sales, transportation, storage, installation, use, and other factors during the life cycle of medical devices.

2. Classification of real-world data

　a.The above data sources can be divided into two categories according to the relationship between the time when the data was formed and the time when the research was carried out:

　　i.The first category: existing data resources;

　　ii.The second category: the purpose of clinical evaluation of specific devices, the establishment of clear data standards and data collection models, data resources formed in the conventional health and medical environment.

3. The key points of evaluating real-world data

　a.When using existing data resources for clinical evaluation, the reviews focus on: whether the population is qualified for the study, key variables and follow-up time are covered; whether medical device identification information and usage information can be obtained; the accuracy and completeness of the variables.

　b.When using active data collection, the first point is to consider the scientificity and feasibility of the research design. Second, to ensure the authenticity, accuracy, and traceability of the data is important. It is necessary to fully identify possible research bias and confounding factors before the start of the study; in the data analysis stage, statistical methods such as stratified analysis, multi-factor analysis, and propensity scoring are used to control and correct the influence of these confounding factors.

4. Services provided by CRCS

　a.The CRCS' RA department and clinical department will follow the "Technical Guidelines for Clinical Evaluation of Medical Devices Using Real-World Data" to determine whether the client's data cover the population required for the study, key variables, and follow-up duration, as well as the clinical evaluation of medical devices. Moreover, we carefully study the real-world data provided by customers, and consider whether its relevance, accuracy and reliability can meet the requirements asked by Center for Medical Device Evaluation (CMDE) or not.

　b.Through communication with the CMDE reviewers, we will be informed about the loopholes in the real-world data and timely negotiate with customers for supplements. As more and more import products go public by this mean in China, it is reasonable to believe that the viability of using real-world data as a way of doing clinical evaluation will become stronger and stronger.