Will AI And Agile Project Management Help Advance RBM?

By Sidharth Ananthanarayan

Since 2013, regulatory policy has encouraged clinical trial management professionals to shift away from their rigid time monitoring framework and toward risk-based monitoring (RBM). The FDA’s 2013 monitoring guidelines aimed to improve human subject protection and clinical trial data quality by focusing on the most important aspects of study conduct and reporting and overseeing them as part of sponsor oversight responsibilities. In fact, 90% of findings during on-site visits can be identified through central monitoring.¹ Furthermore, in 2023 and 2025, the FDA RBM and ICH E6 (R3) final guidelines incorporated risk-based monitoring into the broader GCP modernization agenda, and the FDA proposed a risk-based credibility assessment framework for using AI.^2.3,4

The undeniable force behind these guidances is the growing number of clinical trials, especially Phase 3 studies, and data, with data points having grown from 11% to more 300% compared to a decade ago.^5,6 The old ways of doing things are getting harder. Yet, more than a decade later, the uptake is lagging.

The challenges are many: Clinical research has complex protocols that need to be complied with, including data integrity challenges, blinding responsibilities, timelines, cost management, and documenting nearly every action and decision to confirm the activities were completed. However, the issue is not whether RBM is accepted; it’s whether clinical research can truly change the way it manages work.

RBM Uptake Has Been Slow

A study in 2024 indicates that many companies have implemented RBM in approximately 50% of their studies at the highest, but many are still in the lower levels of adoption.⁷ The study also confirmed that only approximately 20% of the sample size had extensive training in RBM. The common barriers causing slow RBM adoption included that lack of knowledge, as well as a lack of awareness of the new process, skills gaps, and limited availability of tools that can aid in implementation.⁷ Another study in 2025 noted that among roughly 4,900 clinical trials, 65% used at least one RBM element but none adopted centralized monitoring or statistical monitoring, especially since up to 70% of their data could be generated outside the database.⁸

RBM Requires A Shift From Waterfall To Agile Project Management

Clinical research has been following a Waterfall project management methodology, in which feedback is only received at the end of the project. In software development, the Agile methodology changed the way software companies accepted the unknowns, the dependencies, and the user requirements; the feedback mechanism incorporated in each sprint helped make small but visible progress with strong effectiveness. Teams identified the risks or issues in each sprint and addressed them with adjusted planning for the next sprint. However, Agile has been limited to building software tools and solutions in drug development and hasn’t made its way into the day-to-day activities. Yet the shift from Waterfall to Agile project management is necessary to incorporate RBM in clinical research, as it can help clinical teams move from schedule-driven oversight to evidence-driven oversight.

How AI Can Help Sponsors Implement RBM

Although Agile helped the software development discipline, even within the realm of clinical research, it might not be the right fit as-is for risk-based monitoring in clinical research. This is where AI-based tools can help.

AI tools will not replace clinical judgment exhibited by CRAs, data managers, medical monitors, or quality teams. However, AI tools combined with RBM will create a shift in how the activities are performed, critical data points are identified, data is used to understand trends, and focus is centered on the risks. This is similar to what Agile did in software - moving risk monitoring from periodic, document-heavy, front-loaded planning to continuous sensing, prioritization, and course correction.

In general practice, a study team reviews the key risk indicators (KRIs) and quality tolerance limits (QTLs), protocol deviations, query aging, enrollment trends, missing data, safety signals, and site performance in separate categories or instances. The cadence may be monthly, but understanding and connecting these data points depend on manual effort by experienced stakeholders. The use of AI-based tools doesn’t add a new layer to the process but is a connective link that helps forecast the risk signals earlier, broader, and more dynamically. The AI-based tools can also help flag unusual patterns across sites, detect correlations that are hard to see manually, summarize emerging risks for cross-functional review, and free up the resources to be used where needed. The value is not the recommendation or identification of the issues; the value is the shorter loop between evidence and action along with limited manual effort.

A 2025 scoping review in NPJ Digital Medicine analyzed 142 studies from 2013 to 2024 and found AI applications in use across safety, efficacy, and operational risk prediction in clinical trials. These models have been used for adverse event prediction, treatment effect estimation, phase transition prediction, site selection, enrollment, protocol informativeness, and related operational risks.⁹ There are also reports of companies testing AI-based tools and applying simulations to understand how the tool can support RBM.¹⁰

The Future Of RBM

The question still remains whether the AI-based tools embedded into RBM are going to prompt an Agile project management moment for clinical research. The AI-based tools are not the answer, but they provide an opportunity to automate some of the monitoring activities, visualize the data points in combination, and make decisions based on that combined information. Agile changed the rhythm of software development, and that’s what the AI-based tools in combination with RBM could do for clinical research. It would shift the way the teams review the information, implement targeted intervention, evaluate the intervention, and take the feedback. That shift in the decision-making process will be the Agile-like moment that can offer flexibility and consistent control to the teams.

Although technology is advancing rapidly and the regulatory guidelines are supportive, it is important to continuously monitor the information and record the use of AI-based tools. AI does not magically make RBM mature; it only addresses the poor data quality, fragmented systems, unclear ownership, weak process design, and unvalidated models. This can help clinical teams move from schedule-driven oversight to evidence-driven oversight.

References:

1. FDA, "Oversight of Clinical Investigations: A Risk-Based Approach to Monitoring," August 2013. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/oversight-clinical-investigations-risk-based-approach-monitoring.
2. FDA, “A Risk-Based Approach to Monitoring of Clinical Investigations: Questions and Answers”, 2023, https://www.fda.gov/regulatory-information/search-fda-guidance-documents/risk-based-approach-monitoring-clinical-investigations-questions-and-answers.
3. FDA, “E6(R3) Good Clinical Practice”, https://www.fda.gov/regulatory-information/search-fda-guidance-documents/e6r3-good-clinical-practice-gcp, September 2025.
4. FDA, "Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products," draft guidance, January 2025. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological
5. Tufts CSDD, "Rising Protocol Design Complexity Is Driving Rapid Growth in Clinical Trial Data Volume," January 2021. https://www.globenewswire.com/news-release/2021/01/12/2157143/0/en/Rising-Protocol-Design-Complexity-Is-Driving-Rapid-Growth-in-Clinical-Trial-Data-Volume-According-to-Tufts-Center-for-the-Study-of-Drug-Development.html 
6. Getz et al., "Insights Informing Strategies for Optimizing the Collection of Clinical Trial Data," *Therapeutic Innovation & Regulatory Science*, 2025. https://link.springer.com/article/10.1007/s43441-025-00899-4
7. Dirks et al., “Comprehensive Assessment of Risk-Based Quality Management Adoption in Clinical Trials, https://link.springer.com/article/10.1007/s43441-024-00618-5, Therapeutic Innovation & Regulatory Science, 2024.
8. Stansbury et al., “Risk-Based Quality Management: A Case for Centralized Monitoring”https://link.springer.com/article/10.1007/s43441-024-00719-1, Therapeutic Innovation & Regulatory Science, 2024.
9. Teodoro et al., “A scoping review of artificial intelligence applications in clinical trial risk assessment”, https://www.nature.com/articles/s41746-025-01886-7, npj Digital Medicine, 2025.
10. Ménard et al., "Enabling Data-Driven Clinical Quality Assurance: Predicting Adverse Event Reporting in Clinical Trials Using Machine Learning," *Drug Safety*, 2019. https://link.springer.com/article/10.1007/s40264-019-00831-4

About the Expert

Sidharth Ananthanarayan is a clinical research and quality professional with a doctoral degree in Business Administration. He has contributed to multiple clinical development programs, with particular experience supporting cell therapy products for oncology, autoimmune, and neurological indications. His work has involved clinical operations, computer system validation, regulatory compliance, AI quality strategy, and cross-functional collaboration in regulated environments. Sidharth has also contributed to initiatives focused on improving operational processes and supporting quality-driven decision making throughout the clinical research lifecycle. In addition, he actively engages with the professional community through networking, knowledge sharing, and research-related activities.