Using AI to Match Patients With Clinical Trials in Minutes

Jun 28, 2023

Since 2020, the number of clinical trials worldwide has increased by more than 60%, from 325,000 to more than 450,000, according to ClinicalTrials.gov. The explosive growth in clinical trial activity should improve patient access to more investigational therapies in many disease areas and widen the range of options to consider. Despite the proliferation of clinical trials, participation is still lagging, and in an analysis of NCI Health Information National Trends Survey data, authors found that only 9% of people recalled ever being invited to partake in a clinical trial. This disconnect marks a significant opportunity for physicians and patients to improve dialogue about clinical trials as a potential treatment option.

While this disconnect is caused by a web of factors, it is important to consider the barriers to referring patients. A 2022 survey of primary care physicians, which explored their hesitancy to refer patients for clinical trials, found that the biggest obstacles were a lack of awareness of open trials and a lack of clinic-wide process for identifying appropriate trials. Lack of time was also significant among respondents, including lack of time to assess study protocols and discuss research with patients.

Physicians need better tools and support to quickly filter through large sets of data and identify viable trial options for their patients—a task that artificial intelligence (AI) is well designed to handle.

Addressing the Inefficiencies of Manual Clinical Trial Search Methods

Although manually matching a patient with a suitable clinical trial can be time-consuming, the process to determine eligibility is straightforward and well-defined. The first step in the process is to sift through clinical trial registries to assess a patient’s potential eligibility based on such demographics as age, sex, country, and medical condition. The next step is to pre-screen potential eligibility by comparing a patient’s medical profile with a trial’s inclusion and exclusion criteria. This process currently involves manually checking each clinical trial protocol, a time-intensive task even for practitioners familiar with study criteria.

Using AI to Expedite the Matching Process

AI-based technologies provide a tool adept at filtering vast amounts of eligibility data into a manageable set of options. TrialSearch AI, a tool developed by myTomorrows, leverages this potential and expedites the eligibility checking process. Physicians simply input de-identified details from a patient’s medical summary, and then the tool cross-references the data with clinical trial eligibility criteria sourced from public registries. The output is a list of open clinical trials and Expanded Access Programs (EAPs) for which a patient may be eligible. It then prompts the physician to complete manual checks and ask them to confirm whether a pre-approval option is suitable for the patient. Employing AI as a search tool thus reduces the physician’s review time commitment from hours to a matter of minutes.

A study assessing the performance of TrialSearch AI on ten fictitious randomized patient profiles suggests that use of this tool can reduce pre-screening time by up to 90%. Equally important, TrialSearch AI reliably identified all trials for which the patients were potentially eligible and ensured that no potential treatment option was overlooked. The results imply that physicians in the referenced scenarios could have used TrialSearch AI to obtain trusted data to inform their decisions and conversations with patients.

Facilitating Patient Access to Investigational Therapies

Research indicates that patients want to participate in clinical trials. An international survey by the Center for Information and Study on Clinical Research Participation (CISCRP) found that 71% of people worldwide who have never participated in a clinical trial would be willing to do so.

These data suggest there is interest to explore clinical trials as a pathway, when applicable. With AI-powered clinical trial matching tools, physicians have an easy and practical method to improve their own understanding and make informed, comprehensive decisions and recommendations to their patients.

Get more information about TrialSearch AI and how it can help physicians refer patients to eligible clinical trials. There, physicians also can sign up to beta test TrialSearch AI.

Author

Devyn Osborne

View all posts

REFERENCES & ADDITIONAL READING

Anderson A, Getz N, Benger J. Patient input into leveraging the healthcare professional’s role. Applied Clinical Trials Online. 2018;27(4). www.appliedclinicaltrialsonline.com/view/patient-input-leveraging-healthcare-professional-s-role
den Hamer D, Schoor P, Polak TB, Kapitan, D. Improving patient pre-screening for clinical trials: assisting physicians with large language models. arXiv. 2023. https://doi.org/10.48550/arXiv.2304.07396
Millar, MM, Taft, T, Weir, CR. Clinical trial recruitment in primary care: exploratory factor analysis of a questionnaire to measure barriers and facilitators to primary care providers’ involvement. BMC Prim Care.2022;23:311. https://doi.org/10.1186/s12875-022-01898-2
Occa, A, Leip, A, Merritt, AS, Stapleton, JL. Prevalence and correlates of invitation to participate in clinical trials among US adults. Prev Med Rep. 2022;26:101742. https://doi.org/10.1016/j.pmedr.2022.101742

Trends, Charts, and Maps. ClinicalTrials.gov Results Database. 2023. Accessed May 25, 2023. https://clinicaltrials.gov/ct2/resources/trends

ABOUT THE CONTRIBUTORS

Michel van Harten, MD
Michel van Harten is the CEO of myTomorrows. He has more than 15 years of experience in the life sciences industry as a practicing physician and healthcare executive. Dr. van Harten earned both his BSc in Economics and his Medical Degree from the University of Amsterdam. He worked as a physician and researcher at the Antoni van Leeuwenhoek Hospital and Memorial Sloan-Kettering Cancer Center.