Efficient Source Data Verification Using Statistical Acceptance Sampling: A Simulation Study.

Area of Science:

• Clinical trial methodology
• Data management
• Regulatory science

Background:

• Traditional 100% source data verification (SDV) is inefficient.
• Acceptance sampling methodology offers a sampling-based alternative for SDV.
• Existing strategies rely on sampling standards, which may not be optimal.

Purpose of the Study:

• To propose an alternative acceptance sampling strategy for SDV.
• To evaluate sampling plans algorithmically based on prespecified criteria.
• To ensure selected sampling plans meet desired data quality and efficiency targets.

Main Methods:

• The strategy was designed using empirical clinical trial data.
• Extensive simulations based on empirical data were performed.
• Performance was assessed for workload reduction and post-SDV error proportion.

Main Results:

• Simulations across 13 scenarios were conducted.
• In the default scenario, error proportion reduced from 0.056 to 0.023.
• This was achieved by inspecting only 40.5% of case report form entries, with targeted full inspections.

Conclusions:

• The proposed strategy enables substantial workload reduction in clinical trial monitoring.
• Acceptable data quality levels are maintained with the new sampling approach.
• The strategy is conservative, indicating room for further optimization.