Introduction to Clinical Data Management
Q: What is Clinical Data Management?
A: Clinical Data Management is the process of handling data from clinical trials. The inherent goal of any clinical data management system is to produce and maintain quality data.
Q: What are the steps in the process of Clinical Data Management?
A: The overall flow of clinical data handling is:
- Source data are generated. Common examples of source data are clinical site medical records, laboratory results, and patient diaries.
- If paper Case Report Forms (CRFs) are being used, the clinical site records are transcribed onto the CRFs.
- Data from the CRFs, as well as other source data, are entered into the clinical trial database. Electronic CRFs (eCRFs) allow data to be entered directly into the database from source documents. Data from paper CRFs are often entered twice and and reconciled in order to reduce the error rate.
- The data are checked for accuracy, quality, and completeness, and problems are resolved. This often involves queries to the clinical site. See more about data validation.
- The database is locked when the data are considered final.
- The data are reformatted for reporting and analysis. Tables, listings, and figures are generated.
- The data are analyzed, and the analysis results are reported. When significant results are found, this step may result in the generation of additional tables, listings, or figures.
- The results are integrated into high-level documentation such as Investigator’s Brochures (IBs) and Clinical Study Reports (CSRs).
- The database and other study data are archived.
These steps are not strictly ordered. For example, it is common in longer studies to generate intermediate discrepancies and listings periodically to identify problems that need correction before study completion.
Q: What are the elements of data quality?
A: The FDA Guidance for Industry: Computerized Systems Used in Clinical Investigations defines the fundamental elements of data quality for both paper and electronic records by “ALCOA”:
- Attributable: The source of the data is known.
- Legible: The data are readable and comprehensible to humans.
- Contemporaneous: The data are recorded when they are generated.
- Original: The data are the first recording from the primary source.
- Accurate: The data are correct.
Practically, quality data also requires at least three other aspects:
- Data are readily available, transmissible, and storable.
- Data are complete and unbiased.
- Data are in a format that is internally consistent and compliant with or readily transformable to accepted standards.
Q: How is 21 CFR 11 related to Clinical Data Management?
A: 21 CFR 11 details the predicate rules that are required to insure that electronic records are “trustworthy [and] reliable”. Proper implementation of 21 CFR 11 helps ensure that the Attributable, Legible, Original, and Accurate aspects of the ALCOA standard are met.
Q: How is Validation related to Clinical Data Management?
A: Validation of clinical data management programs and procedures is required to document that clinical data management standards are met. See more about database validation.
Questions? Contact Ofni Systems at (919) 844 2494 or at firstname.lastname@example.org