Veranex Risk Based Monitoring (RBM) Business Delivery Model With 5D Framework

Veranex has successfully designed and implemented a “5D Framework” to support its RBM service delivery model, in order to create a better customer-centric service that is value-driven, transparent and meets expectations in a cost-effective manner.

The overall RBM business delivery model starts with using both business acumen and analytic thought processes to develop structured problem solving with data.

• Define: Asking the right questions with an understanding of the “Key Risk Indicators (KRIs)” is a fundamental step in our ‘Define’ approach.
• Distill: Industry standard statistical models are developed to extract information in both structured and unstructured data.
• Discover: Data trends & patterns are identified using descriptive, inferential, predictive, Machine Learning and Time Series Forecasting — these have been a core strength to address many KRI related problems.
• Deliver: Findings are then linked to the business insights for RBM Recommendations — keeping the insight storytelling simple and straightforward.
• Drive: Recommendations provided are finally linked into actions and risk mitigation efforts to track progress for the KRI related issues.

While we’ve presented the “Overall Delivery Model” for RBM, it is also important to shape strategies at each KRI level. Veranex presents the ‘D2P Analytics Value Chain’ approach to focus on individual strategies. Together, the ‘5D strategy’ along with the ‘D2P analytics’ approach strengthens the overall life cycle of RBM KRIs.

WHAT HAS OUR EXPERIENCE SHOWN? 

Using our RBM approach, we have found that the results (detected anomalies/KRIs) are reliable as well as actionable allowing for faster, dependable, and accurate detection with sustainable solutions.

Some examples:

• Using Descriptive Analytics, we can look at enrolment trends and screen failures to detect anomalies.
• Using Diagnostic Analytics, we can review those screen failures to identify potential reasons across sites and prioritize sites for follow-up.
• Using Predictive Modelling, we can answer some enrolment and screen failure questions
  • What is the probability of completing targeted enrolments in time period (t)?
  • How many subjects (n) need to be screened to meet the targeted enrolment?
  • How many new sites may be needed to complete the targeted enrolment within time (t)?
• Using Prescriptive Approach and based on the findings above, what actions or changes to process are needed to get enrolments back on track (e.g., Change to inclusion/exclusion criteria? Staff training on enrolment and protocol specified procedures?  Data quality issue?)

In our next-in-series, we will give a deeper look into how this is accomplished.