We Need a Next-Generation RBM Approach Today

Researching medicines and treatments has become increasingly complicated and decentralized. Sponsors and CROs use various electronic systems to support their efforts, often calling on third parties to help them do so. The GCP reported that, based on global inspections, we need to do a better job monitoring activities and managing data. The ICH GCP E6 R2 puts more onus on sponsors, investigators, and CROs and requires sponsors to adopt adequate processes for monitoring trial activity, communicating across sites, and maintaining data integrity. It also states that sponsors should use a risk-based approach to monitoring clinical trials.

These guidelines should be read in their entirety to best understand the recommendations. Sponsors and CROs are now expected to document a monitoring plan that describes methods, responsibilities, requirements, and strategy. Sponsors can choose on-site monitoring, a combination of on-site and centralized monitoring, or if justified, centralized monitoring alone. Their approach must be supported, documented, and reported on in enough detail to verify compliance with the monitoring plan. The guidelines also suggest that protocol design and implementation, monitoring plans, standard operating procedures and training all be used to help reduce risk throughout the trial’s entirety. Quality management activities need to be documented and communicated to all those affected to facilitate risk review and appropriate action.

Since comprehensive RBM is now a recommended approach, it is time to reevaluate the industry standard and to expect more from RBM solutions. Many of today’s offerings are too simplistic and inflexible to support the new requirements and do not afford a holistic view of the myriad datasets involved in trial management. A superior approach to RBM begins with integrating all critical data sources, regardless of format in near real-time. Giving site sponsors and CROs faster access to all of the datasets involved in clinical research processes- data that previously existed in silos-will pave the way for further enhancements, such as improved reporting and analytics functions, to drive gains in efficiency, quality, and oversight.

“Tearing Down” The Silos

Currently, sponsors and CROs struggle with the siloed nature of data collection. Operational data is separated from clinical data; integrating disparate sources is costly and time-consuming; and the clinical medical review process is error-prone and labor-intensive. Regulators have identified a number of data sources as essential, including source documents/data, including metadata, case report forms, central laboratory reports, documentation on the investigational medicinal product(s), documents/data from Interactive Response Technologies (IRT), protocol deviation reports, safety reports/drug safety systems, trial master files (eTMF), and quality metrics. The ability to access and manage all of these documents/data is flagged by regulators as a potential pitfall to RBM implementation.¹

We need to bridge the integrated data structure gap in Clinical Research area. Users could better manage risk if a connected set of data were available, as all the potential events could be correlated and any challenges would be easily discovered. Many of the existing RBM solutions are actually just simple analytic or reporting tools that do not aggregate data in near real time or perform interactive data discovery or RBM functionality. The tools offered by clinical informatics vendors are limited in functionality and unable to aggregate data. The custom solutions CROs build for sponsors are also inflexible and built as proprietary data warehouses that are not source-agnostic.

Integrated data collected for the purpose of clinical research is the first step in providing an effective risk management solution; it powers the reporting and analytic capabilities necessary for identifying and monitoring risk, allocating resources, and managing issues. The ability to have timely, quality connected data from the various sources collecting data for clinical research would allow users to better identify the correlated risk signals earlier on in a clinical study process. The technology needed to power this more integrated approach is already on the market and in use by some of the world’s most innovative companies. This type of system is built on big data technologies using the data lake concept, which allows for a loosely coupled data ingestion approach that is not hardwired with any of the source systems. The traditional approach uses an ETL methodology to ingest data into a predefined common structure. Data is extracted from a source, transformed, and then loaded (ETL) into a common structure that can only accommodate a certain level of variability. Data lake-based systems invert the process, extracting and loading data first. Then, end users can transform the data on a near real-time basis for analysis and signal/risk/issue detection.

This architecture is better suited for the complex nature of modern clinical research. Because RBM draws on such a vast array of data sources, it requires technology solutions that are flexible enough to load data from any application, contracts, financial, LAB, EDC, IRT, CTMS, database, line listing, file or report.
 
The ability to view data across sources, and sites, will facilitate improved remote monitoring, a step forward from focusing on reducing the amount of source document verification (SDV) and visits to sites where there might be a problem. This next-generation tool would allow users to better understand and categorize risk, update monitoring plans as needed, and create study-specific key risk indicators (KRIs), with thresholds designed to trigger action. It could also draw on machine learning technology and include a deep learning algorithm that would track user activity and predict risk and behavior based on past experience.

The solution could also integrate with other areas of the clinical research process. For example, the RBM solution could facilitate performance-linked payments to sites based on triggers, such as data being clean, which would incentivize sites and potentially drive more gains in efficiency and data quality.

Best practice now suggests that sponsors and CROs adopt risk-based monitoring that includes systematic, ongoing, and proactive risk assessment on an organizational and project or trial level and that is embedded in a quality-by-design and risk-based approach. This requires access to the myriad data sources cited as impactful by regulators. Technology can power ongoing, flexible data aggregation that allows users to proactively identify, even prevent, inadequate site behaviors. RBM becomes proactive instead of reactive -- a means of managing risk, not just monitoring it.

References

1. Schwartz, G. “Any Pitfalls in RBM Implementation? The Perspective of the Regulators” The Federal Institute for Drugs and Medical Devices, 3rd European Conference of Clinical Research, (2016)

About the Author

Sudeep Pattnaik is President & CEO of ThoughtSphere. Prior to founding ThoughtSphere, Sudeep was the global leader of products for Quintiles, the largest CRO Fortune 500 life science company in the world, creating and leading the strategy team behind a $60M integrated healthcare data hub. He also played a key role in defining the Risk-Based Modeling approach for optimizing the clinical development process and helped develop a best-of-breed RBM platform for the industry. He holds and MSc in Computer Science from Uktal University (India) and an MBA from Leeds School of Business at the University of Colorado, Boulder.

About ThoughtSphere
Founded by clinical information and technology industry experts with over 30 years of experience from the leading global CRO and eClinical providers, ThoughtSphere’s mission is to streamline and empower the clinical trials process by eliminating the two biggest challenges--integration of disparate clinical and operational data and making it accessible for use with existing tools for analytics and visualization. With this innovative platform, biopharma, medical device sponsors and CROs can reduce and optimize clinical development costs, aggregate operational and clinical data to enhance efficiency and effectiveness in the clinical trial processes and gain near real-time actionable insights. The product suite includes ClinDAP, the source-system agnostic next-generation data integration platform; ClinACT, the interactive visualization and analytics platform that enables RBM and CRO Quality Oversight; and SPACE, the integrated site budgeting, payments and contracting solution. At ThoughtSphere, we believe we can deliver on the promise of big data to drive health innovation. For more information, visit www.thoughtsphere.com.