Data Integrity in Clinical Trials
The integrity of clinical trial data is essential to:

  • Providing regulators, prescribers and payers information they can rely on to make good decisions for public health and patients
  • Retaining public trust in medicines and medical devices
  • Encouraging participation in clinical trials

As Sponsors increase virtualization, risks to data integrity grow, including how and where clinical data are stored.

The Move from Paper to Electronic
21 CFR Part 312, Investigational New Drug Application, defines the roles and responsibilities of Sponsors, Clinical Investigators, and Contract Research Organizations (CROs). FDA issued Part 312 when most clinical trial processes were paper-based, and nearly everything that a protocol required of a clinical trial subject was under the Clinical Investigator’s control or direct influence. Site staff documented all protocol-required data in pen on carbonless triplicate Case Report Forms (CRFs). The Sponsor’s monitor would verify these entries against paper medical records, document queries on query forms, and eventually send one of the carbonless copies back to the Sponsor for entry into the clinical data management system. Local labs would analyze a subject’s blood sample on the order of the Clinical Investigator. X-rays would be taken and read locally.

Sponsors looked to change these local, paper-based processes to get higher quality data in-house faster, with the aim of getting to market faster. To do this, Sponsors turned to IT solutions offered by CROs and specialty clinical service providers.

With electronic CRFs (eCRFs) in Electronic Data Capture (EDC) systems, Sponsors could save time and money by:

  • Programming queries about possible anomalous data at the point and time of entry by Sites
  • Streamlining the query process
  • Eliminating the need to collect and transport or ship paper from Sites to the Sponsor

Sponsors turned to central labs and companies that specialized in collecting images from Sites and reading them centrally. This helped reduce variability in data and provide a central, uniform source of data for the Sponsor’s Clinical Data Management and Biostatistics departments.

Virtualization of Clinical Trials
Today, nearly all Sponsors execute clinical trials in a highly virtualized way. Sponsors outsource data collection, management, and reporting both to CROs as defined in Part 312 and to clinical service providers providing specialized services.

It is not uncommon today for a Sponsor to contract with at least six different companies to execute a Phase 3 oncology study:
1.   Clinical Data Management CRO, for configuring a clinical trial database in an Electronic Data Capture (EDC) system and managing the resulting data
2.   Interactive Voice/Web Response System (IxRS) provider for managing drug supply, assigning subjects to treatment arms, and providing Sites with the ability to perform emergency blind breaks
3.   Central Clinical Laboratory for analyzing standard lab values
4.   Specialty Laboratory for analyzing biomarker data
5.   Central Image Reader for reading digital images
6.   Electronic Patient Reported Outcome (ePRO) provider for assessing subject quality of life parameters

Virtualization and Data Integrity Risks
Increased virtualization brings two common risks to the integrity of clinical trial data.

Data Storage
eCRF data are stored in EDC systems. These systems should have role-based user access and audit trails on the eCRF data. The EDC system is typically integrated with the IxRS system, which populates the EDC system with basic information about the subject (e.g., subject ID).

Lab data are usually stored in the lab’s Laboratory Management System. Results of the central readers’ interpretations are stored in the Central Reader’s database application. ePRO results are generally transferred wirelessly from the Subject’s device to the ePRO provider’s database application. In nearly all cases, these systems have role-based user access and audit trails.

All of these providers extract data from their applications to provide to the Sponsor or the Sponsor’s CRO. It is not uncommon for these data to be stored outside of the EDC system in folders on file shares.

Risk: Depending on the procedural controls around granting, reviewing, monitoring, and revoking user access, inappropriately authorized users may have inappropriate levels of access to clinical trial data. This presents risks to all 9 data integrity attributes.

Data Format
The format in which these data are often stored can also be problematic from a data integrity perspective.

Risk: When clinical trial data are stored as Excel, text, or .csv files, it is possible for data to be changed by users with little to no trace of the change. This again presents risks to all 9 data integrity attributes.

Combine both risks and we have a perfect storm for data integrity problems.

Validant provides GCP data integrity assessments and, if necessary, remediation execution services.