There is digital evolution occurring across all aspects of clinical research. Advanced analytics tools that use artificial intelligence and machine learning (AI/ML) give research teams the power to make sense of the vast data being collected through clinical research. These tools enable advanced statistical predictive modelling and process automation to enhance overall study quality while freeing staff to focus more on highlighted critical items.
This includes the adoption of Risk Based Monitoring (RBM) to manage the flood of new data coming into the clinical data environment from eSource, eConsent, eISF, EMR and other digital sources. This data offers a wealth of valuable information, but only if companies have the technology to comb the data coming in with a focus on driving quality and efficiency. To accommodate this shift, sponsors need more comprehensive dataflow management technologies that use AI/ML to streamline review and decision-making processes.
AI/ML-driven data management doesn’t just accelerate data capture or provide generalized insights. These platforms seamlessly integrate large volumes of data from a breadth of sources, and feature automation tools to streamline the review process.
The most sophisticated solutions, like IQVIA’s Central Monitoring (CM) platform, clean and house all data in a single location, and use custom machine learning algorithms to identify data errors, outliers, and false entries.
The analytical engine features natural language processing capabilities, which means it can identify site or study risks in both structured and unstructured data. The AI/ML features enable predictive analytics that can improve patient safety and site performance by detecting and mitigating potential issues.
When risks are found, the platform can provide alerts and produce dashboards summarizing RBM analyses and updates, all with little or no human interaction. This includes predicting protocol deviations, which can cause safety concerns for patients and delays in project delivery. In one recent project, the platform analyzed historical trend data to predict when protocol deviations were most likely to occur, allowing the monitoring team to proactively implement mitigation actions. The model showed over 70% prediction accuracy, and it was able to detect ~100% of protocol deviations within data reviews.
The impact of such predictive analytics can be substantial for monitors, sites, and the patients they treat.
It also brings greater efficiencies and job satisfaction. By automating the data collection, cleaning and review process, CRAs are free to focus on more value-added tasks, and sponsors and sites get faster access to information.
IQVIA has worked with several large organizations to help them enhance quality and efficiency in their projects. In one example, a large pharma company running a global trial came to IQVIA for help in accelerating their RBM process while reducing burden for staff.
The sponsor was already using central monitors to support CRAs in the field but they wanted to further streamline the data review. Prior to the deployment of the AI/ML system, alerts for site risks were managed through manual processes where human monitors reviewed all the data then decided whether a risk warranted an action item, or that no action was required.
This process required many hours from a limited staff who were becoming overburdened by the volume of data to review.
As a solution, they deployed IQVIA’s Alert Automation, which is part of IQVIA’s RBM platform. Machine learning algorithms were trained using data from past trials to analyze and predict risks and determine the appropriate response. When a risk is identified, the platform automatically assigns the appropriate actions and sends an action alert to the team, freeing the monitors and site staff to focus on solutions rather than data review tasks.
A review of the implementation found that the AI/ML technology matched decisions made by experts more than 90 percent of the time, while reducing the number of hours required to complete RBM tasks by 75 percent.
Monitors are continuously looking for ways to speed data review and the identification of potential safety concerns for patients. AI/ML-driven RBM platforms can deliver the insights they need.
Advanced analytics technologies are becoming a must for RBM, especially in high data-volume environments. When these tools can rapidly capture and analyze the deluge of data coming from multiple sources and identify potential risks it speeds decision making and can improve data quality across the entire clinical environment.
As these tools become the norm, it will soon pave the way for predictive and proactive signal detection, driving more robust safety insights and a more efficient environment.
Moving into the future such AI-driven innovation and flexibility will be paramount for clinical trial success. It does require some digital transformation disruption as sponsors transition to a RBM model and embrace automation, but the quality and cost benefits will make this transition a worthy journey to take.
Predict enrollment more accurately and increase the number of patients per site using data-driven patient recruitment solutions.
Change the model of clinical research by integrating data, expertise, analytics and technology from study design through execution to power better decisions.
Explore our end-to-end, full service clinical development capabilities including Therapeutics expertise, Development Planning, Phase I early clinical development, Phase IIb/III and Phase IV Trials, Regulatory Submission and Post-Launch Studies.