Our Clinical Research 101 series takes an in-depth look at key steps and tips for navigating the clinical research process. The seventh installment in our Clinical Research 101 series is by Karin Bezuidenhout, Data Management Operations Lead at CHÉOS. The following is an overview of clinical data management, including why it is important and what it entails.
Ms. Bezuidenhout is a Data Management professional with more than 12 years’ experience in clinical data management, working on multiple global Phase II, III, and IV studies including various EDC and paper-based systems. As Data Management Operations Lead she is responsible for the organization and execution of CHÉOS and Canadian HIV Trials Network data processes and strategies. If you would like to inquire about our services, please submit your request here.
What is Clinical Data Management?
Clinical Data Management is the process of collecting, cleaning and managing participant data collected for a clinical trial while adhering to regulatory standards and good clinical data management practices. Clinical Data Managers are involved in all stages of the clinical trial, from protocol development until completion of the trial with the main focus to assure the overall accuracy, integrity and quality of the clinical trial data.
Responsibilities of Clinical Data Management includes:
Study Start-Up
(e)CRF design and development
The Data Manager will design case report forms used to collect participant data based on protocol requirements. The forms can be in paper or electronic format.
Database build and validation
The Data Manager will set up the database entry screens as per the data points specified in the protocol.
The Data Manager will document the database entry screen validation activities by testing the database against valid and invalid test cases to ensure the user requirements are met.
Edit check programming and validation
Edit checks is the auditable process of assessing the contents of a data field against its expected logic, format, range or other properties and intended to reduce error. Queries will fire when discrepant data is entered and submitted in the database.
Validation includes the creation of valid and invalid test cases for each check to ensure that the check will fire as expected.
The Data Manager will document all required edit checks in the study specific edit check plan and will also document the applicable validation activities.
User acceptance testing
UAT is performed by the end-user and covers the testing of the database to ensure that the database meets the study specific requirements by imitating normal use conditions. UAT occurs before the database is released to production.
Database user training
The Data Manager will provide database training to the end-users at study start-up as well as providing ongoing training whenever post-production changes are made to the clinical study database. Training will focus on the study specific database design and entry requirements.
Study Conduct
Data entry and verification (paper)
For paper based trials, the paper Case Report Forms will be entered into the study database by a Data Entry Coordinator. All data entered will also be verified to ensure the accuracy of the entered data.
With EDC (electronic data capture) databases, the data will be entered directly into the EDC database by the site coordinator. Data entered into EDC databases will be source-verified by a clinical monitor.
Discrepancy management
Query management is the cleaning of data by addressing the queries that were generated by the edit checks in the clinical database. Queries may be generated electronically (programmed edit checks) or may be raised based on discrepancies identified through manual review of the data (manual queries).
For each query/discrepancy the Data Reviewer must consider the cause of the discrepancy, assess the error and determine the appropriate action (closing the query if the response is satisfactory, issuing a re-query if further clarification or action is required).
Data coding
Adverse Events, Concomitant Medications and Medical History are coded to categorize the reported terms appropriately so that it can be analyzed.
Dictionaries used include MedDRA, a clinically validated medical terminology dictionary and thesaurus used for adverse event classification, and WHODrug, an international dictionary of medicines that provides information about a drug’s active ingredients and therapeutic uses.
Data review
The Data Manager will review the data collected in the database to verify the completeness, accuracy, and consistency of the data points.
SAE reconciliation
Reconciliation to ensure all SAEs (serious adverse events) were reported as expected.
Data transfers
With the data transfer process files are transferred to and from the clinical study database, while taking the necessary steps to preserve the integrity of the transmitted data and the study database.
The Data Manager will ensure that the file transfers meet the applicable regulatory, ICH GCP, and sponsor requirements, including maintaining study blinding, ensuring data security, and transferring the data to the requester.
Post-production changes
Post-production changes to the clinical database are sometimes required after it has been released into production and includes changes to forms, fields, structure, attributes, and edit checks. Changes to a production database require a thorough justification, an impact assessment, and validation. Depending on the database, there may also be system limitations to consider.
The Data Manager will follow a robust change control procedure to ensure the integrity of the clinical study database and the production data. The process will include extensive review and approval of the database design before release of the changes to the production environment.
Study Close-Out
Quality control
QC activities will be undertaken based on the quality control measures established at the beginning of the study. The purpose of QC is to verify that the quality requirements for the project have been fulfilled.
For paper based trials this also includes QC of entered data against the paper Case Report Forms. Usually critical data will undergo 100% QC while only a random sample of non-critical data will undergo QC.
Database Lock
At the end of the trial, once all study data has been entered, all data review and cleaning activities have been completed and the team has agreed that the data is ready for analysis, the clinical study database will be locked to prevent further changes and modifications to the database and data. This also means that all write/edit access rights have been revoked.
After a database has been locked Data Management will provide the final datasets for analysis.
Data Archive
After the database has been locked, the database and data will be stored on a separate storage device for long-term retention as per the study specific requirements.
Why is Clinical Data Management important?
The Clinical Data Manager ensures that the data collected from the trial participants and entered into the clinical trial management system is complete, accurate and consistent, thereby supporting the integrity and quality of the statistical analysis and results.
Clinical Data Management provides the data and the database in a usable (fit-for-purpose) format in a timely manner and is essential to the overall clinical research project as its key deliverable is to provide complete, accurate and consistent data to support the research question.
Additional resources:
- CTN Research Toolbox
- CHÉOS Clinical Research Navigation Tool
- Data Management in Clinical Research: An Overview
If you have any questions on the details of Data Management, please contact researchsupport@cheos.ubc.ca and we can assist you.
