Discover unrealized connections and manage your product across the entire lifecycle, from research and development, compliance to launch and commercialization.
How can we make clinical trials more efficient, cost-effective and successful? The answer, at least in part, lies in applying artificial intelligence and analytics to design clinical studies. In this blog post, we discuss how IQVIA is currently using these methodologies, in partnership with sponsors, to enhance study design — ultimately benefiting patients, researchers and the larger healthcare ecosystem.
“Cutting-edge methodologies are helping us advance healthcare,” Pablo Aran, Director of Product Management at IQVIA said. “They give us insights to better inform trial design decisions, including uncovering drivers of patient and site burden and aspects that may unnecessarily increase protocol complexity, providing sponsors an opportunity to address these to realize positive operational outcomes.”
Comprehensive data assets, including real-world data, when used as a source for deriving AI and analytics-based insights, become a powerful tool for proactive decision making. This enables trend analysis, prediction of potential issues and minimized risk of amendments based on patient and site experiences. With this in mind, IQVIA has developed transformative technology which digitalizes protocols, allowing us to extract data and aggregate it in a de-identified fashion to gain granular insights from protocols based on specific diseases, trial phases and therapeutic areas. We then apply AI/ML and scoring algorithms to quickly assess overall protocol complexity, patient burden and site burden. Our continuously updated library of protocols allows for the creation of benchmarks against similar protocols, enabling sponsors to perform scenario planning and assess the impact of various design decisions. Armed with this knowledge, sponsors can make evidence-based decisions to reduce risk, streamline their design and improve both patient and site experiences.
Data analytics can help sponsors design more patient-centric studies, improving adherence and enrollment outcomes while reducing the overall burden on participants. They can illuminate drivers of patient burden, enabling sponsors to consider what changes can be made to reduce these barriers, thereby increasing patients’ willingness to participate. Furthermore, when revisions are made prior to protocol finalization, risk of avoidable amendments can be reduced — potentially saving the sponsor up to $535k USD and three months of delay, according to statistics by Tufts CSDD Impact Reports.1
For example, patient-reported outcomes are often seen as a burdensome element by patients. In one case study, applying data analytics uncovered that the protocol required multiple overlapping domains within the patient-reported outcomes to be measured. Additionally, some of these requirements were misaligned with the objectives of the study. By removing extraneous PROs, while replacing another with an abbreviated version, the sponsor not only lowered the overall patient burden score of the study but also reduced study costs.
Decreasing patient burden has even greater implications. In another example, 105 historical protocols were reviewed after trials assessed with IQVIA’s design analytics had been completed, comparing patient burden of the executed protocols against operational outcomes. Trials with a higher patient burden were found to be correlated with longer start-up timelines, more protocol amendments and higher screen failure rates.
IQVIA's findings reflect those identified in larger industry-wide trends, which show correlations between higher patient burden and many operational outcomes, including delays in several stages of study execution, as well as higher numbers of amendments. In particular, Phase II and III protocols with relatively lower participation burden were associated with shorter mean start-up, enrollment and close-out cycle times compared to those with higher burden. 2
Thus, it would be fair to conclude that addressing patient burden with protocol revisions or risk mitigation strategies has the potential to also improve trial execution outcomes.
Consideration of the patients who would most benefit from being treated with the drug if it were to be approved has historically been lacking in clinical trials. Regulators have emphasized representative practices, highlighting the importance of reaching a minimum number of minority subgroups impacted by the disease state within clinical trials. Additionally, sponsors have been more focused on addressing gaps in care for underrepresented populations. As a result, addressing access and representation in clinical trials has become a strategic imperative.
The implications are even broader than clinical trial success. “Finding a representative mix from a race and ethnicity perspective can impact commercial success,” adds Aran. “Using real-world patient data, we can identify barriers to patient participation and mitigate bias in trial design.”
More specifically, by conducting quantitative research we have been able to uncover differences in patients’ perception of burden based on their race and/or ethnicity. By considering multiple perspectives on elements that might be burdensome — whether related to race, ethnicity, geography or other factors — we are able to create protocols that are more likely to resonate with our intended populations. For example, Hispanic participants may prioritize in-person visits, while other groups may be reluctant to maintain patient diaries. Addressing these nuances fosters inclusion and ultimately impacts operational success. Approaches developed by IQVIA, that are informed by the quantitative research that was conducted, can help sponsors see the impact their design choices could have on their desired population.
Industry-wide, clinical studies are increasing in complexity, with a 139% increase seen in procedure quantity, a 214% increase in the number of endpoints and a 600% increase in data points collected since 2005.3 Complex trials directly impact everyone involved: patients, sites, sponsors and CROs. Challenges include identifying sites willing and able to run the trial, finding patients to participate and stay enrolled, overcoming protocol deviations and avoiding unnecessary amendments — these barriers can all increase the time and cost it takes to complete the trial.
“As we further develop our protocol complexity analytic, we will finalize our capability to benchmark the Protocol Complexity Score against a large, industry-wide dataset of anonymized, digitalized protocol designs and further test that against trial outcomes data,” Aran stated. “That way we will truly understand which design elements might have the most impact on operational outcomes, allowing us to uncover opportunities to further optimize study design.”
Early operational insights can be brought into the protocol generation and assessment step. AI can be used to create estimates of enrollment and trial duration. “Having these insights during the protocol design phase allows sponsors to make better-informed decisions that mitigate potential risks downstream,” he added.
Design analytics can be used to understand operational impact of protocol design decisions through three different perspectives. First, by correlating operational impact of the stakeholder burden, second by defining the complexity score based on operational outcomes and third by conducting early estimations of key operational outcomes during protocol design.
IQVIA’s design analytics solutions allow sponsors to realize benefits in four key areas, including:
As we journey toward designing more efficient, cost-effective clinical trials, several key takeaways stand out:
“Time limits how many lives we can touch in bringing forward needed therapies,” concluded Aran. “Machine learning and real-world data allow us to get some of that time back to ultimately reach more patients.”
Together, as we innovate and push the boundaries of what is possible with AI and analytics, we’ll shape a future where clinical trials are not just scientific endeavors but compassionate journeys toward better health.
1 Protocol amendment costs and delays are calculated based on phase 3 trials data from Tufts CSDD Impact Reports (2016)
2 Tufts CSDD Impact Report, September/October 2024
3 Protocol Design Variables Highly Correlated with, and Predictive of, Clinical Trial Performance. Ther Innov Regul Sci 56, 333-345 (2022).
Discover unrealized connections and manage your product across the entire lifecycle, from research and development, compliance to launch and commercialization.