A P-VALUE IS ONLY AS GOOD AS THE DATA: Challenges when Endpoints are Based on Subjective Assessments
When endpoints are based on subjective assessments rather than objective data, a process of centralized endpoint adjudication can improve the quality of a clinical trial.
Central reading was most frequently used to improve determination of efficacy endpoints or in a combination of efficacy and safety endpoints. This data suggests that there is growing adoption of endpoint adjudication as a tool in drug development.ENDPOINT ADJUDICATION REGULATORY REFERENCES
There are no dedicated Health Authorities’ guidelines setting standards about endpoint adjudication. However, reference to processes and committees involved in the structured assessment of endpoints can be found in a variety of FDA and EMA guidance documents, for example:
- Establishment and operation of clinical trial data monitoring committees / FDA Guidance for Industry;
- Guideline on data monitoring committees / EMA Committee for Medicinal products for human use (CHMP);
- Clinical trial endpoints for the approval of cancer drugs and biologics / FDA Guidance for Industry;
- Diabetes mellitus – evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes - FDA Guidance for Industry.
KEY ELEMENTS OF ENDPOINT ADJUDICATION
There is also no established and accepted academic or industrial standard on how to implement and manage an endpoint adjudication process and different approaches are currently implemented. Based on their practical experience with adjudication programs, the authors have identified the following definitions and procedures as key elements in ensuring consistency and quality in an adjudication process:
A precise identification of the endpoints
A clear definition of the endpoints – possibly based on objective criteria or measurements, and supported by data standards that have been accepted by the scientific community – allows clear endpoint characterization and thus lead to uniform reporting of events.
An Adjudication Charter as the adjudication roadmap
The ground rules for the adjudication process are captured in a Charter that describes in detail all operational and technical aspects of the adjudication process. It must include all the definitions about endpoints and the source documents needed for the adjudication process, roles and responsibilities of the Adjudication Committee, and the assessment forms and workflows. It must include precise instructions for all the people involved in the process and it is essential that this document is approved before assessments are made.
An Adjudication Committee composed of expert clinicians, independent and well-trained
The panel of reviewers must have experience in the related therapeutic area and should be trained also on the technical and methodological aspects of the adjudication process. Reviewers should not have conflict of interests and conduct the assessments fully blinded to avoid observers’ bias.
A prospective operations plan to ensure process excellence
The endpoint adjudication process is generally quite complex as it involves several stakeholders and steps. The plan should cover the following:
- How the Central Committee is notified when assessments are due
- Timelines and procedures for the collection, redaction, translation, and anonymization of source documents
- Assembling data and source documents to be included in the Event Dossier and submitted to the Adjudication Committee members
- The adjudication workflow, i.e.,
• the process of querying the sites for incomplete event dossiers;
• how inter-rater disagreements are managed through consensus meetings;
• the event re-submission for changes in the supporting documentation;
• the audit trail and history of changes in adjudication.
A data management plan
The data management plan describes the data and documents for assembling the Event Dossier, the forms to collect the reviewers’ assessments, and the datasets and their formats for submission to the Authorities. The application of standards like CDISC / STDM further improves the quality of the process.
A Quality Assurance program
SOPs, training curricula, and a properly designed systems’ validation plan and report are key elements of compliance with applicable standards such as 21 CFR Part11.
A Quality Control program
Quality control checks and performance monitoring are required throughout the endpoint adjudication process to ensure a timely detection of errors and bias. Metrics implemented at critical process steps – for instance, a metric about the collection of source documents, or intra- and inter-rater variability – support the users and drive quality.
CONCLUSION
When using endpoints that are not based on objective measurements but require an expert assessment, bias can be introduced into a clinical trial. This bias can relate to efficacy or to safety endpoints. Bias or intra- and inter-rater variability can be reduced if a centralized adjudication process that follows a robust process and procedure is applied.
