Artificial intelligence (AI) is being adopted by clinical trial sponsors, clinical research organizations (CROs) and site management organizations (SMOs) to drive efficiency. This includes protocol development, site identification, patient recruitment, patient engagement, study monitoring, and study closeout/data review. AI-enhanced data analysis is also a catalyst to the emergence of virtual trials and biosimulation. AI enablement is not without its regulatory and legislative challenges. Access to patient data is highly regulated by privacy, security, and confidentiality rules. In addition, both state and federal legislators have taken interest in regulating AI. Herein we detail leading and emerging applications of AI in the clinical research space, and delineate regulatory and legislative trends shaping the industry.
To receive a copy of the full report, please contact us.
The Traditional Clinical Trial Model Faces Efficiency and Complexity Challenges
Clinical trials rely heavily on labor and expertise. Identification of trial subjects remains a largely manual process subject to bias. Continuous patient engagement is challenging from a resource perspective. Data collection is increasing in complexity. Trial site monitoring remains a labor-intensive and expensive quality control step. Study closeout, including clinical data review, remains resource intensive. Finally, the use of data for future study planning demands a higher rigor of structured data.
Error and delay significantly impact trial timing and cost. 80% of trials exceed enrollment timelines. 72% of studies run more than 1 month behind schedule. 53% of enrollments exceed planned enrollment timelines. 30% of clinical trials terminate because they’re unable to enroll the target number of patients. 50% of sites charged with enrollment, enroll one or no patients in studies. 85% of clinical trials fail to retain a necessary number of patients and 80% of all clinical trials fail to finish on time. 
Through artificial intelligence (AI)-powered tools, sponsors (ex., pharma), clinical research organizations (CROs) and site management organizations (SMOs) are increasingly enhancing their operational capabilities to overcome these limitations and improve profitability. AI is designed to operate with varying levels of autonomy to sort and interpret data and generate predictions, recommendations, or decisions. This helps to improve clinical trial processes beyond human resource limitations. Table 1 summarizes an array of expanding AI applications described below.
AI can improve patient recruitment by identifying and screening potential participants based on inclusion and exclusion criteria. AI directed toward medical records can help by extracting the pertinent information needed in the final eligibility screening from all available electronic health records (EHRs). As EHRs lack a standardized format, AI machine learning can help translate information across disparate sources to map the database structures used for specific patient eligibility reviews. This can help reduce the time, labor and bias associated with patient recruitment, which is a common clinical research issue.
From a patient perspective, recruiting SMOs are not always aware of all the clinical trials that may be relevant to specific patients, leaving patients to rely on searching for matching trials themselves through clinicaltrial.gov. AI machine learning can be programmed to recognize qualified candidates from electronic health records and prompt physicians to inform patients of trials that could benefit them. In addition, the inclusion/exclusion language used to qualify clinical trial candidates can be intimidating to typical patients. Here, natural language processing can help translate complex medical jargon as the patient initially determines their eligibility.
Enhancing Participation And Retention
AI advancements allow CROs to extend remote patient monitoring during clinical trials, allowing a broader demographic and geographic range of patients to participate in trials more easily. AI algorithms also open up greater possibilities in interpreting individual patient data and a better understanding of their behaviors and needs. Enhanced patient-centric interactions improve understanding and result in greater patient retention.
Interpreting A Greater Stream Of Data
AI solutions can be used with IoMT (internet of medical things) to extract pertinent information like vital signs, exercise, sleep and more. Even medication adherence can be monitored by recording video of patients taking medication, with AI facial recognition and medication sensing features confirming the visual record for reporting. Wearable devices and other sensors can collect real-time data on patients’ physiological parameters, which can be analyzed using AI to identify patterns indicating the onset of a potential adverse event or complication. In short, AI can help improve patient monitoring during clinical trials, leading to better patient safety and more efficient trial execution.
Enhancing Patient Safety
Patient monitoring can also be improved by incorporating the use of AI, which in turn can help improve patient safety and reduce the risk of adverse events. AI algorithms can detect and predict adverse events by analyzing various data types, such as vital signs and patient-reported outcomes. This can help researchers identify potential safety concerns more quickly and take appropriate action, such as modifying the study protocol or adjusting the dosage of the drug being tested.
Increasing The Speed Of Study Closeout And Clinical Data Review
The duration from the “last subject last visit” trial milestone for the last phase 3 trial to the submission of the data package for regulatory approval, has been largely unchanged over the past two decades and presents a huge opportunity for positive disruption. It involves cleaning and locking the trial database, generating the last phase 3 trial analysis results (frequently involving hundreds of summary tables, data listings, and figures), writing the clinical study report, completing the integrated summary of efficacy and safety, and finally creation of the data submission package. AI, coupled with natural language processing can be vital in automated data extraction and document processing.
Driving Biosimulation To Improve Future Clinical Trial Design
Biosimulation utilizes simulations to model biological processes and symptoms. Through AI and machine learning, the simulation creates predictions which researchers can use to detect patterns in past clinical trials, evaluate relationships between factors and design trial parameters accordingly. Input to these simulations include the data from prior clinical trials, wherein the use of AI within current trials to uncover and structure data expand the library of data available for AI-enabled sponsors to model future trial designs. Researchers can use these models to test aspects such as dosages, drug interactions and effectiveness across demographics toward more robust or population-specific future trials.
HIPAA Considerations In Clinical Trial AI
Researchers and healthcare organizations must comply with Health Insurance Portability and Accountability Act (HIPAA) regulations as well as other privacy laws to ensure patients’ public health information (PHI) is protected. Utilizing AI technology in patient recruitment can lead to challenges in ensuring compliance with these regulations. For example, AI algorithms may need to access a large amount of patient data to identify appropriate clinical trials. However, this data must be de-identified and protected to ensure patient privacy and prevent unauthorized access to PHI. Similarly, third-party AI-driven patient recruitment solutions need to be mindful of compliance with privacy regulations.
FDA Action In The Clinical Trial AI Space
The FDA is taking action to ensure the harmonization of Good Machine Learning Practices (GMLP), which will focus on removing bias from AI algorithms and enforce a standardized system for scouring patient health records. To maintain transparency with patients, the FDA hosts public workshops with the Patient Engagement Advisory Committee on how device labeling can foster transparency between manufacturers and users while also building trust in AI medical devices.
Given the growing role of AI in virtual trials, the device space is also relevant. Most recently, the FDA has issued an updated revision of its guidance document for the “Predetermined Change Control Plan for AI/[Machine Learning]-Enabled Devices,” which highlights and defines for device manufacturers “what” changes are being made to the device using machine learning and “how” the changes in algorithms will be redeveloped to still maintain safety and efficacy.
State Legislative Impacts On AI
At least 11 states have considered legislation dealing with the use of artificial intelligence in healthcare, according to the National Conference of State Legislatures’ AI legislation tracker. Currently, AI technologies are governed by a patchwork of laws and regulations. There are various efforts underway to address the regulatory gap, including the European Union’s Artificial Intelligence Act (which the European Parliament recently approved) and the Blueprint for an AI Bill of Rights published by the Biden Administration in October 2022. More recently, the Biden administration issued an executive order establishing new standards for AI safety and security, including privacy considerations.
Clinical Trial AI Investment Examples
Stakeholders across the clinical trial continuum are investing in AI. The global AI-based clinical trials solution provider market alone is valued at ~2B (2022) with an expected CAGR of over 20%.
- Developed with AI platform company Aily Labs, “plai” is being utilized by Sanofi to find clinical trial sites that will allow for more participation among underrepresented communities.
- AstraZeneca and Bayer are working with Altis Labs to develop “digital twins” which simulate real-world scenarios, to predict how patients and treatments may respond.
- Syneos Health (SYNH) is collaborating with Microsoft research to increase AI usage across its clinical and commercial activities as well as leverage developments from OpenAI. It had previously entered into a strategic partnership with Haystack Health, a Roivant Health portfolio company developing advanced AI and natural language processing solutions, to enhance the identification and enrollment of patients for clinical trials.
- ICON is partnering with Saphetor, which specializes in large-scale identification and interpretation of human genetic variants, to develop new approaches for clinical trials focused on precision medicine and rare diseases. They will leverage Saphetor’s VarSome.com platform related to the epidemiology, prevalence and location trends of genetic mutations to support patient recruitment and trial strategies.
Clinical Trial AI Resource Providers
- AICure, which has raised $52.8 M, offers AiCure Patient Connect™ to improve patient engagement, improve the relationship between the site and the patient, and achieve a deeper understanding of individual and population-wide disease symptomology for improved health and trial outcomes.
- ConcertAI, which has raised a total of $300M, offers Concerto. The platform provides real-world evidence services for precision oncology, getting access to electronic medical records, results of NGS diagnostics, and patient-reported outcomes to generate evidence for new therapeutic approaches.
- Lantern Pharma Inc. (LTRN) integrates artificial intelligence (AI) with oncology to enhance the efficiency of clinical trials via biomarker-led clinical trial design. Their RADR® platform employs a combination of AI and machine learning techniques to analyze over 25 billion oncology-specific data points.
- Owkin, which has raised $84.9 M, combines AI, Digital Twins, and novel statistical methods to enable smaller, more efficient trials. The company’s activities included work on Alzheimer’s Disease and Multiple Sclerosis
Target analysis in this environment requires an understanding of the clinical research space as well as federal and state regulatory and legislative policies guiding the use of AI and digital enablement. It also requires market access knowledge of pharma dynamics and strategic analysis on the AI landscape in clinical trial applications. With extensive experience in the core clinical development landscape including CROs and SMOs, as well as FDA regulatory considerations, Marwood’s services connect policy with market dynamics and strategy.
 J Med Internet Res. 2020 Nov; 22(11): e22179.
 https://www.appliedclinicaltrialsonline.com/view/streamline-and-improve-study-start (accessed Jan 2024)
 https://www.appliedclinicaltrialsonline.com/view/enrollment-performance-weighing-facts (accessed Jan 2024)
 https://www.pharmacytimes.com/view/accelerating-patient-recruitment-may-help-address-clinical-trial-delays-in-fda-approval-process (accessed Jan 2024)
 https://www.clinicalleader.com/doc/considerations-for-improving-patient-0001 (accessed Jan 2024)
 https://www.fda.gov/medical-devices/software-medical-device-samd/good-machine-learning-practice-medical-device-development-guiding-principles (accessed Jan 2024)
 https://www.fda.gov/medical-devices/software-medical-device-samd/predetermined-change-control-plans-machine-learning-enabled-medical-devices-guiding-principles (accessed Jan 2024)
 Grand View Research. AI-based Clinical Trials Solution Provider Market Size, Share & Trends Analysis (accessed Jan 2024)
About the Author: Mark Slomiany PhD MBA MPA is a Director of Advisory at The Marwood Group and a former faculty member of the Department of Cardiothoracic Surgery at New York University Langone Health, as well as a former research associate at the Mossavar-Rahmani Center for Business and Government at the Harvard Kennedy School of Government.