Addressing Challenges in Oncology Trials: The Integration of AI to Streamline Inclusion and Exclusion Criteria

Oncology trials present challenges in patient recruitment and eligibility assessment. Many trials struggle to meet their enrollment targets due to complex inclusion and exclusion criteria. This issue is particularly relevant in the United States, where around 80% of clinical trials fail to meet recruitment objectives on time because of these strict criteria. The integration of artificial intelligence (AI) into the management of oncology trials is being recognized as a tool to improve patient recruitment, reduce administrative tasks, and enhance trial effectiveness.

Understanding the Complexity of Oncology Trials

Oncology clinical trials have detailed inclusion and exclusion criteria. These trials can involve up to 70 different criteria. Each criterion necessitates a careful review of patient data, leading healthcare professionals to spend one to four hours per patient assessing eligibility. This lengthy process can delay trials and increase costs. Delays in clinical trials can cost sponsors between $600,000 and $8 million per day, putting pressure on clinical research organizations.

To tackle these challenges, AI solutions can provide potential benefits. Tools such as the CARAai™ Precision Trial Solutions, created by ConcertAI and Exigent Research, seek to streamline the patient recruitment process. This system uses advanced AI technologies to improve study feasibility, match patients to trials, and automate workflow tasks. This is especially important in oncology since timely access to clinical trials can impact patient treatment outcomes.

The Importance of Data in Patient Eligibility

A significant amount of data used in clinical trials exists in unstructured formats. This includes clinical notes and imaging reports. It is estimated that about 80% of healthcare data is unstructured, posing a challenge for conventional recruitment methods. Traditional approaches mostly depend on structured data sources, often missing valuable information in unstructured data.

Recent developments in Natural Language Processing (NLP) provide a way to address this data challenge. NLP can transform unstructured data into structured formats, which can improve patient profiles and recruitment efficiency. For instance, NLP applications have identified previously overlooked eligible patients in clinical notes. In a trial for multiple myeloma, over 40 patients were found who met the necessary criteria but were missed via standard recruitment methods.

Challenges in Recruitment and Patient Retention

A significant issue faced by clinical trials in the United States is patient retention. High dropout rates can compromise clinical research integrity, with studies indicating that 15% to 20% of trials do not recruit enough patients to finish. Predictive analysis can help mitigate this dropout rate by identifying potential patient dropouts based on risk factors. AI’s predictive abilities can enable researchers to engage participants at risk of dropping out, enhancing retention efforts.

Additionally, AI can streamline patient recruitment by using algorithms to identify and screen potential participants based on set inclusion and exclusion criteria. Tools like DQuest improve this process by analyzing patient eligibility data and categorizing trial options using AI. Enhancements in patient selection operations can lead to faster recruitment processes, aiding trials in reaching their targets more efficiently.

Improving Workflow Automations

The role of AI in oncology trials extends beyond recruitment and retention. It improves workflow automations. Hospitals and clinical research organizations increasingly use AI to ease the administrative tasks related to clinical trials.

AI algorithms can manage various tasks in trial management, from data entry and analysis to patient follow-ups and compliance monitoring. Automating these tasks allows healthcare professionals to focus more on patient care rather than repetitive administrative work.

It is important to recognize that these AI-driven workflow enhancements are becoming necessary. In a healthcare environment where efficiency and speed are vital, the FDA has started to update regulations to ensure that AI technologies prioritize safety and address biases in algorithms. This shift indicates a growing acceptance of advanced technologies while maintaining ethical standards.

The Role of Real-World Data in Enhancing Research

Real-world data (RWD) sources are increasingly relevant in oncology research, alongside traditional clinical trial data. RWD includes health information gathered outside of controlled clinical trials, such as data from electronic health records (EHRs), insurance claims, patient registries, and wearable device data. Combining RWD with AI can enhance data analytics and provide a broader view of patient populations.

RWD can lead to more accurate patient-to-trial matching by offering insights into the effectiveness of treatments in everyday clinical settings. Incorporating real-world evidence into trial designs can facilitate adaptive trial designs, adjusting study parameters based on ongoing results to better align with patient needs.

The European Health Data Space (EHDS) is an initiative aimed at standardizing health data across Europe to improve research outcomes through better data quality. Such efforts promote cross-border collaboration and enhance the applicability of research findings, creating a more comprehensive approach to oncology trials.

Challenges and Control Measures

While integrating AI in oncology trials shows promise, challenges remain. Data privacy, compliance with regulations like HIPAA, and algorithmic biases must be prioritized. The FDA has made these concerns a priority, indicating a commitment to establishing guidelines that ensure AI in medical devices meets safety and efficacy standards.

Building trust among the public regarding AI use in clinical trials is also essential. Ensuring transparency about how AI technology is used in recruitment and assessment can enhance patient confidence and willingness to participate in trials.

Collaborative Innovation in Oncology

Partnerships between organizations like ConcertAI and Exigent Research illustrate the potential for innovation in oncology trials. By combining AI technologies and community research networks, these collaborations can expand patient access and make oncology trials more representative of standard care populations.

Sibel Blay, CEO of Exigent, notes that this collaboration aims to increase clinical trial access, especially in community settings. Broadening clinical trial demographics to reflect the diversity of standard care populations ensures equity in research and may lead to findings that better serve larger patient groups.

In summary, as oncology research continues to change, the use of AI offers opportunities to address long-standing issues related to patient recruitment and trial management. The ability of these technologies to improve efficiencies, simplify processes, and enhance study outcomes is significant. The future of oncology trials in the United States is promising with these advancements, leading to better patient-centered results in cancer treatment.