The Advantages of Artificial Intelligence for Risk-Based Monitoring in Clinical Trials
The Advantages of Artificial Intelligence for Risk-Based Monitoring in Clinical Trials
AUTHOR: MOULIK SHAH, FOUNDER AND CEO, MAXISIT
Digital transformation is underway across nearly all aspects of clinical trials. Risk-based monitoring (RBM) is one area ready for evolution. Traditional statistical methods don’t paint a full picture of important and likely risks, while siloed and disconnected systems make comprehensive data analysis difficult, if not impossible.
Dedicated risk-based quality management (RBQM) platforms that incorporate artificial intelligence and machine learning (AI/ML) enable sponsors to take a truly proactive RBM approach to clinical trials. With an RBQM system powered by AI/ML, sponsors can assess and address a wide range of clinical and operational risks, available in real-time or near-real time.
Here, we explore how AI/ML transforms RBM through its ability to rapidly process vast amounts of data from EDC, CTMS, eCOA and other data sources. With that information, these tools can identify risks that need correction today, as well as potential risks that can be avoided.
The Benefits of Risk-Based Monitoring
The FDA recommends that clinical trial sponsors take a risk-based approach to clinical trial monitoring. This proactive approach involves continually monitoring for and managing patient safety and data integrity.
RBM is a technique used to assess, mitigate, and monitor trial risks. It is an aspect of RBQM, which is an approach that involves applying RBM at a systematic level to manage quality throughout a clinical trial.
By identifying and mitigating risks early on—ideally in the planning stage—sponsors can focus on the data and processes necessary for successful execution. They can also solve data-related issues before they snowball into trial-derailing problems.
The FDA recommends sponsors monitor important and likely risks identified during the initial risk assessment, as well as risks detected while monitoring the trial. Information learned from one study can be used to inform subsequent trials.
Why a Dedicated RBQM Platform is Essential for Efficient Clinical Trials
Now that the volume, variety, and veracity of clinical trial data has grown to petabyte proportions, the need for dedicated RBQM solutions is altogether apparent. A dedicated RBQM system brings together both structured and unstructured data from multiple disparate sources into a central location. From that location, the system analyzes data to highlight critical risks before study personnel can take action.
Effective risk-based monitoring (RBM) requires more than automation and rule-based algorithms. To proactively assess and highlight clinical and operational risks, RBQM systems need the advanced analytical power made possible by AI/ML. These tools go beyond flagging current possibilities, sifting instead through volumes of present and past data to recognize patterns and associations while assessing the severity of those found and predicting the likelihood of future risks.
AI/ML Application
What does AI/ML-powered RBQM look like in action
A traditional RBQM system identifies a suspected anomaly in data purportedly captured during a patient’s site visit. Upon investigation, the sponsor learns the anomaly was caused by a data entry error. Analyzing past and current study data using AI/ML, the sponsor discovers a history of data entry errors at that site. Retraining site staff on the EDC and other procedures resolves the problem for the current trial and other trials moving forward.
During trial planning, an AI/ML-powered RBQM system analyzes historical data, benchmarks, and trends to flag risks most likely to occur. The system recommends—and the study team agrees—staff perform close oversight of those areas. Staff outlines their process in the risk monitoring plan. As a result of their proactive planning, study teams intervene immediately as high-risk situations occur. The trial proceeds according to set timeline without issues that could skew results. Example
These are only two of many potential use cases that illustrate the impact of predictive analytics. This type of information helps clinical operations and data management teams, clinical monitors, and sites, as all work more seamlessly toward a common goal: conduct an efficient, effective clinical trial with optimal results with as little burden on patients as possible.
What Sponsors Can Achieve with AI/ML-Powered RBQM
The speed of AI, combined with its pattern-recognition power, allows sponsors to fully realize the benefits of a risk-based approach. A few notable ways AI/ML-based RBQM platforms streamline and enhance RBM include:
AI/ML-powered RBQM systems automate data collection, cleaning, and review. Automation not only reduces risk caused by human error or oversight, but it speeds up processes. Clinical research associates (CRAs) typically tasked with manual data cleaning and review are free to focus on high-value activities, which benefits job satisfaction. It also allows sponsors to apply resources more effectively, thus conserving labor costs.
More thorough site risk assessments
AI/ML’s predictive capabilities can analyze data from previous studies, from participants’ electronic health record (EHR) data, and from other relevant sources to identify high-risk sites and/or patients. For example, the system may flag sites likely to have recruitment, enrollment, or performance issues. It may also identify study participants at higher risk of serious or adverse events. In addition to identifying risks, the tool may also suggest actions to be taken by monitors to help mitigate those risks.
Improved protocol compliance
ML-based algorithms can compare incoming trial data to protocol requirements, identifying discrepancies or non-compliance automatically. Study teams can address these issues in the moment, preserving the validity of trial results.
Improve current and future trial performance
Using advanced analytics, AI/ML-powered RBQM systems can identify data trends related to patient outcomes, protocol adherence, adverse events, and other metrics. Sponsors of adaptive trial designs can then make adjustments throughout the study based on clear evidence. The system can also identify broader trends to help inform future trials. For example, analyzing study data across therapeutic areas, patient populations, or phases may help sponsors develop more targeted trial designs or risk management strategies. This data-driven risk-based approach could improve trial performance.
MaxisIT RBQM Solution—Powered by AI
MaxisIT's RBQM Solution, aligned with the TransCelerate Methodology, provides customizable risk indicators, thresholds, and mitigation strategies through the utilization of RACT (Risk Assessment & Categorization Tool) and the Integrated Quality and Risk Management Plan (IQRMP). This comprehensive approach empowers users to evaluate crucial data and processes continuously throughout the trial period, revolutionizing monitoring across diverse study formats, from conventional site-based trials to cutting-edge decentralized statistical monitoring.
Utilizing AI capabilities, MaxisIT's RBQM Solution ensures ongoing real-time monitoring by identifying data patterns, outliers, anomalies, and critical risks. Leveraging Natural Language Processing (NLP) tools, it analyzes both structured and unstructured data, facilitating predictive risk assessments that enhance decision-making with actionable recommendations. This proactive approach significantly improves patient safety, protocol compliance, trial oversight, and site performance.
About MaxisIT
MaxisIT’s purpose-fit and intelligent clinical data analytics platform helps improve clinical trial performance, mitigate risk, and optimize clinical outcomes. We provide a centralized and reliable source of truth for diverse data types from various sources, giving life sciences companies real-time insight to shorten cycle time and increase return on investment.
Incorporating an end-to-end clinical data pipeline from intake to visualization, MaxisIT's solutions are powered by AI/ML and metadata-centric approaches. Our impressive portfolio of over 3,300 clinical trials and an unparalleled 100% customer retention rate affirm the quality and reliability of our services.
Moulik Shah Founder & CEO, MaxisIT
Moulik Shah is a passionate healthcare technology entrepreneur and the visionary CEO of MaxisIT, where he has been at the forefront of leveraging technology to transform pharmaceutical and life sciences clinical trials.
His dedication to improving patient outcomes and his leadership in directing patient-centricity, patient diversity, interoperability, and real-world-data-led collaborations have been at the core of his vision of an integrated healthcare ecosystem based on effective use of data and analytics platforms.
He has been instrumental in driving innovation and progress in the industry. Under Moulik’s leadership, MaxisIT has become a leading provider of clinical data and analytics which is driving real-world impact in the pharmaceutical and life sciences clinical trials.