Using AI and Advanced Analytics to Transform Clinical Trials

An increasingly pressurized drug development market is placing high expectations on contract research organizations (CROs).

To address the challenges of modern clinical research, the industry needs to find better, smarter ways to detect risk, streamline data review, and automate oversight

Ken McFarlane, VP of Strategic Consulting, at CluePoints shares insight on how harnessing the power of artificial intelligence (AI) and advanced analytics-powered technology will be key to success for modern CROs.

Embedding Risk-Based Strategies

There are now 4,431 CROs in the U.S. alone¹. In this competitive market, ignoring the advance of AI is not an option. Instead, CROs need to take a strategic view of how new technologies can help transform their business models and enable them to move beyond the traditional revenue streams of monitoring and source data verification (SDV). CROs today need to leverage AI and advanced analytics to enhance study efficiency, protect data integrity and meet sponsor expectations. 

Global regulators and sponsors are raising the bar when it comes to risk-based strategies. Around 8 in 10 industry leaders believe RBQM will improve the overall quality of research and more than 6 in 10 trust it will enable efficiency and cost savings². ICH E6 (R3) pushes a unified risk-based model, expanding on the concept of centralized monitoring and its ability to identify systemic or site-specific issues. At the same time, sponsors are continuing to examine streamlining the integrated data review (IDR) process. 

To be successful in this environment, and ensure readiness for what comes next, CROs need to embed risk-based strategies into operations. Moving from outdated methods to data-driven oversight will be vital, as will streamlining and simplifying the oversight and reconciliation process across teams. By taking these steps now, CROs can build more sustainable, long-term partnerships and ensure their long-term future.

Of course, AI will be central to this success. However, CROs also need to consider how to go beyond technology and lead organizational transformation that aligns people, processes, and platforms for regulatory-ready, data-driven trials. 

An AI-Powered Risk Ecosystem

The percentage of clinical trials implementing RBQM elements is expected to increase to 81% in execution and monitoring and 79% in documentation and resolution by 2027³. With adoption on the increase, we need to stop thinking of RBQM as an alternative or optional way of operating within a trial and instead move to it as the standard operating model for trial oversight moving forward. This new era of integrated RBQM as a standard, will harness advanced technologies and incorporate central statistical monitoring (CSM) to transform how CROs detect, act on and learn from data. 

CSM – an approach backed by the FDA⁴ – applies statistical algorithms to identify data outliers and discrepancies. By integrating AI thoughtfully, CROs can gain even deeper insights and create efficiencies. Ultimately this is about identifying and detecting data and quality issues with more speed and accuracy, increasing speed to market and delivering better outcomes for patients and sponsors. 

The addition of AI-driven tools to Integrated Data Review is already unlocking deeper oversights, accelerating timelines and automating previously labor-intensive processes. CROs need to seize these opportunities as the market continues to standardize this approach.

Enhancing RBQM with AI Tools

Traditional site monitoring is one of the most expensive components of clinical trials with onsite monitoring alone accounting for up to 30% of total trial expenses⁵. CROs can improve their ability to evaluate the performance of clinical trial sites by adopting adaptive site monitoring. Adaptive site monitoring leverages advanced analytics and machine learning (ML) to move beyond traditional SDV methods, enabling more focused SDR, limited SDV and more efficient data locking strategies. This allows CROs to quickly to re-assess site monitoring priorities, simplify site visit planning, optimize site recruitment, evaluate clinical research associate (CRA) effectiveness across studies and mitigate inherent risks.

Another example of how RBQM can be enhanced with AI is medical coding. Auto coding tools typically successfully code just 50-60% of input terms. Use of a synonym library can increase this to around 70-80% but this is labor-intensive to build and maintain and still results in a significant proportion of codes requiring manual coding⁶. A deep learning (DL) model can offer precise, AI-generated coding suggestions for concomitant medications and adverse events at up to 99% accuracy. The model can automatically handle updates to the WHODrug and MedDRA dictionary and leverage semantics by using embeddings that encode the meaning of words. Combined, these advances can reduce the time taken to code medical data and adverse events by around 75%.

Medical review provides another practical example of the real-world applications of AI. Medical and safety reviewers have historically had to rely on manual processes to analyze patient data and safety outcomes. This can lead to errors and inefficiencies. An AI-powered process using specialized large language models (LLMs) can rapidly identify common outliers, enhance query management with clear status updates and integrate data review and query generation. This reduces manual workload, allowing data managers to improve oversight and empowering them to drive data quality and overall trial success.

Conclusion

CROs looking to succeed and differentiate in an increasingly pressurized and competitive environment need to think strategically about how to leverage AI and advanced analytics. 

Future success will rely on going beyond risk detection to operationalize analytics, make proactive decisions, and resolve issues before they escalate. By combining AI with human expertise and regulator-backed integrated data approaches, CROs can succeed faster, focus resources where they matter most and deliver the efficiencies that sponsors demand.

References

1. https://www.ibisworld.com/united-states/number-of-businesses/contract-research-organizations/5708/   
2. https://link.springer.com/article/10.1007/s43441-024-00618-5
3. https://pmc.ncbi.nlm.nih.gov/articles/PMC11043178/
4. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/risk-based-approach-monitoring-clinical-investigations-questions-and-answers
5. https://www.americanpharmaceuticalreview.com/Featured-Articles/185929-Ways-to-Lower-Costs-of-Clinical-Trials-and-How-CROs-Help/
6. https://cluepoints.com/a-new-era-of-automation-improving-efficiency-outcomes-with-intelligent-medical-coding/

Ken has spent that last 24 years in the Clinical Trial R&D space. 12 of those 24 years was spent working in various clinical operations roles within Sponsors & CROs, with a large focus in monitoring, project management and clinical trial oversight. The other 12 years Ken has spent working for clinical trial technology vendors, where he’s been working to innovate and streamline clinical trial processes to benefit sponsors, CROs, sites, and the patients they serve.