Random sampling model would boost uptake of risk-based monitoring in trials

The trials sector continues to favor traditional monitoring methods over risk-based approaches due to a lack of a straightforward methodology.

Researchers at the University of Utah reached this conclusion after analyzing the uptake – or lack of uptake – of risk based monitoring in drug studies.

“Clinical trial monitoring is evolving from labor-intensive to targeted approaches. The traditional 100% Source Data Monitoring (SDM) approach fails to prioritize data by significance, diverting attention from critical elements. Despite regulatory guidance on Risk-Based Monitoring (RBM), its widespread implementation has been slow,” they write.

In traditional monitoring– known as “100% source data monitoring” - a review compares all reported data elements to the original source documents. SDM does not differentiate data based on its significance or potential impact to participant safety or study results.

Also – the authors write – “The broad scope distracts monitors from concentrating efforts on critical elements. Additionally, 100% SDM results in intensive time and resource endeavors to rectify minor errors.”

Risk-based monitoring

In contrast, risk-based monitoring – which is supported by the US Food and Drug Administration (FDA), European Medicines Agency (EMA) and industry group TransCelerate - directs focus and activities to the evolving areas of greatest need which have the most potential to impact patient safety and data quality.

The potential benefits are clear, yet according to the authors, use of the approach is limited. In part they cite a lack of awareness and mixed perceptions of the value proposition of RBM as factors. They also note the “lack of a universally accepted, straightforward RBM methodology as a significant hurdle”

To address the latter issue, the authors propose a risk-based monitoring approach that includes random sampling of participants and variables for SDM and describe a method for centralizing documentation.

They write “In our proposed two-step SDM method, sampling occurs at the participant level followed by the variable level. If critical issues are found during monitoring, additional variables are monitored, which allows for an adaptive SDM approach.”

DepositPhotos/magann