It’s not a secret; artificial intelligence (AI) is booming. With its efficiency-increasing potential, it’s not even a surprise. AI holds great promise for optimizing and enhancing almost every industry — including clinical research. For example, researchers speculate that generative AI can streamline safety and pharmacovigilance by helping identify significant safety report information and write adverse event reports.1 Others look forward to using AI models to cleanse and augment datasets to make recommendations on ideal next steps for drug or device development.1 But perhaps the most important way AI can enhance clinical research is through patient enrollment.
AI and Patient Enrollment
Patient enrollment is one of clinical research’s most pressing issues. Not only is a low patient accrual rate the most common reason for trial termination, but 86% of trials also don’t meet their enrollment timelines.2,3
Among other reasons, patient enrollment is so challenging due to the lack of patient awareness of and access to clinical research. By automating the trial matching process and making research as care more accessible for a wider range of patients, AI has the potential to assuage this pressure point and streamline patient enrollment, ultimately accelerating the development of new medicines and therapies and benefitting everyone from sponsors and pharma companies to doctors and patients.
How would AI accomplish this?
Large language models (LLMs) can process huge amounts of paper and electronic patient medical records of varying formats and transform them into easily digestible and cohesive electronic health record (EHR) data. This EHR data can then be analyzed and searched to quickly match patients to clinical trials for which they meet the inclusion/exclusion (I/E) criteria. Some AI platforms can take patient enrollment optimization beyond matching to further enhance the clinical research process, such as myTrialsConnectSM from Elligo Health Research®.
AI-Enabled Patient Registries
myTrialsConnect is an AI-enabled patient registry that finds and engages qualified patients before, during, and after a study for the benefit of all research stakeholders.
For patients, myTrialsConnect promises empowerment and greater ownership of their healthcare journeys. When they sign up, patients get a copy of their full medical records as well as resources and customized messages on their unique conditions, concerns, and interests. The registry also matches patients to upcoming clinical trials based on automated reviews of their medical records and information gathered from chatbot-based surveys. myTrialsConnect offers providers a comprehensive view of their patients’ medical histories and makes it easier for them to recommend clinical trials and offers sponsors and researchers a virtual waiting room of pre-qualified patients who have not only been matched to the study but who have also already expressed their interest in participating.
This patient registry will make it easier for sponsors to identify, engage, and enroll patients in their clinical trials while giving providers and patients more control over healthcare and research as care.
The Downsides of AI-Powered Patient Enrollment
As with any innovation, there are some risks associated with using AI in clinical trial enrollment. Many have expressed concerns about output accuracy, as AI’s responses are dependent on the quality of its training data as well as the relevance between the model and the use case.1 Similarly, due to the newness of the technology and the open-source design of many models, AI also might pose an increased danger of security and data privacy concerns.1
To mitigate these risks and realize AI’s full efficiency-increasing potential, the research industry must first and foremost educate itself on AI. Most sponsors and researchers with a deep understanding of regulated research don’t have a deep understanding of AI technology, and vice versa, exacerbating existing risks and making it harder to remedy them. The industry must also be careful to only train AI platforms on current, accurate, and corroborated data; use best practices with prompt engineering, such as being aware of the quality of a query; and always leverage human oversight.
There is also the question of regulation. It’s easy for the research industry to use the myriad laws governing clinical research as excuses for not adopting new technology like AI. But they are just excuses. Yes, it will take time for the FDA and other governing bodies to catch up to innovations. But it is only a matter of time. AI regulations are coming and once they do, companies that have not yet adapted to the new technology will be left behind.
A Commitment to AI Quality
Elligo is working hard to remain at the forefront of AI innovation while mitigating risk and balancing compliance, efficacy, and most importantly, patient safety. We offer a proven method to ensure compliance and good clinical practice that includes:
- Stability, communication, and quality conduct supported by comprehensive infrastructure and reliable, continuous technology
- Precise oversight including PIs, coordinators, and account managers supervising every step of every study
- Extensive data security policies, including technology and training
With a 99% first submission success rate and no major or critical findings or 483s in FDA inspections to date, it’s clear our safeguards work.
What’s Next for AI in Clinical Trials?
Imagine what could happen when AI is properly integrated into clinical research. Perhaps the 86% of trials that traditionally don’t meet enrollment goals will meet them. Maybe data will be clearer and more actionable, decisions will be more precise, and patients will get the treatments they need faster. Is that future here yet? No — but if the research industry is careful, and if we take the right next steps, it will be soon.
In the meantime, contact Elligo to learn more about myTrialsConnect or our other AI-enabled services and what they can do to enable your research.
References
- Laws, L. How generative AI could change the life sciences landscape – an interview with Indegene. Outsourcing Pharma. Published 2023 June 7.
- Fultinavičiūtė, U., et al. Trial termination analysis unveils a silver lining for patient recruitment. Clinical Trials Arena. Published 2022 Oct. 23.
- Harrar, S., et al. Artificial Intelligence for Clinical Trial Design. Science Direct. Published 2019 Aug.