As Protocol Complexity Grows, AI Decreases Site Workload And Improves Patient Recruitment
A conversation with Cedar Health Research CEO Todd Albin
The clinical research industry continues to expand, as do the challenges that clinical research sites face when conducting studies. From workforce challenges and increasingly complex study protocols to new initiatives for greater diversity and inclusion, there’s no shortage of obstacles for research sites to overcome. Cedar Health Research CEO Todd Albin uses his decades of experience in patient recruitment to drive new AI/ML strategies that expand access to clinical research opportunities for physicians and patients alike.
In this Q&A, Albin chronicles the increasing protocol complexity and deepening patient recruitment challenges that come along with it, while also offering ways in which AI-assisted technologies can improve upon traditional recruitment methods for quicker, more accurate, and, thus, more successful patient recruitment and enrollment.
Clinical Leader: Todd, you’ve been in clinical research for a while and seen the industry from a variety of sides. Can you tell us a little bit about your background and experience in clinical research?
Todd Albin: With over 25 years of experience in the clinical research industry, I have dedicated my career to accelerating the availability and affordability of new therapies for patients. My primary focus has been on addressing one of the biggest challenges in the clinical development cycle: patient recruitment.
I have managed clinical research investigator sites across various settings, including practice-integrated, stand-alone, and national site networks. I have facilitated multi-therapeutic outpatient Phase 1-4 trials, constantly seeking innovative approaches to enhance patient recruitment and retention. I also spent six years in the CRO space, where I supported patient recruitment and retention solutions, as well as the expansion of trial access through the establishment of new investigator networks. I am now leading a site and patient network that is using an AI/ML software solution as its core patient recruitment infrastructure in partnership with community healthcare providers.
Patient recruitment has changed quite a bit since you started in the industry. What trends have you seen in your past roles that were once considered the standard for patient recruitment?
During my career, I witnessed the rapid growth of protocol complexity, outpacing the innovation in patient recruitment solutions. In the late 1990s and early 2000s, we could effectively recruit subjects utilizing local media (TV/cable, newspaper, and radio). If a site was affiliated with a practice, we could run medical billing reports by diagnosis codes, pull paper charts, and flag them for providers. We could also send direct mail to prospective participants from a patient list. By the mid-2000s, recruitment professionals began using electronic media for patient recruitment. Sites could run email campaigns and build websites for recruitment to supplement existing tactics.
How are these advances in technology received by CROs and sponsors? Did they increasingly allow them to reach new and more patients?
Sponsors and CROs have had mixed reception. We often find that large sponsors are fixated on funding outdated recruitment techniques such as manual chart review, placement of local advertisements, and central campaigns. Smaller sponsors and biotechs have been more open minded to approve funding for innovative solutions. CROs are similar to large sponsors. If the solution is not in the box of the preapproved recruitment tactics, it gets declined with little consideration. Innovation is moving faster than the outdated budget and contract processes sites have with sponsors and CROs.
From a site perspective, what challenges or changes have you seen around these evolving patient recruitment strategies?
At the same time e-recruitment became more sophisticated, protocol complexity grew right beside it. Soon, sponsors began using professional central patient recruitment providers to supplement site recruitment. The level of sophistication and scale needed to utilize these tactics was typically beyond a site’s capability and budget. As I saw the effectiveness of site-managed, traditional recruitment media of newspapers and TV decline, I understood that central recruitment companies were going to be the best at executing sophisticated online campaigns. The challenge with these companies was their lack of understanding of how clinical research sites operate and how to properly qualify a study candidate for a site to enroll them.
That’s because prescreening of research candidates is difficult using a structured online questionnaire or large-scale call center scripted operation. If a patient makes a mistake when self-reporting their health history (like having diabetes and then later providing medication for a high cholesterol), the screening process continues, and this unqualified patient is referred to a site. In a dynamic conversation with a trained healthcare professional with access to patient records, these issues can be addressed and properly assessed.
And so, I left the site world and joined a central recruitment company to help them understand how to work better with sites and optimize site enrollment of prequalified referrals. This model operated with great success on many trials that required identifying thousands of additional patients for sites. Unfortunately, protocol complexity continued to increase to the point where a site could not rely upon self-reported medical history of candidates through basic screening questions and call center scripts. That complexity includes growing numbers of inclusion/exclusion criteria and, most impactfully, any criteria that requires a site to validate a candidate's medical records to confirm information they are not able to accurately self-report. In addition, protocols today have more visits, blood draws, and patient-facing technology. These complexities make it harder to find study volunteers willing to consent to these demanding studies.
Given the increased complexity of study protocols alongside the growth of centralized patient recruitment companies, what did your involvement with a patient recruitment company teach you about the process?
For most protocols, actual medical records were needed to confirm patient eligibility for a study. This resulted in a large number of referrals being sent to sites — at a significant investment by sponsors — who did not qualify, creating a significant burden for site personnel to process. This exposed both the inability of central recruitment companies to be an effective solution for complex trial recruitment as well as the need to identify patients through medical records held by their treating community physicians. The key challenge is this: How can this be achieved at the speed and scale of a central recruitment campaign?
Your question asks it perfectly. So, how can patient enrollment for complex trials be achieved at the speed and scale of a central recruitment campaign without burdening the coordinators and research staff?
At Cedar Health Research, we integrate with physician practices, using a patient recruitment AI/ML platform to expand clinical trial access to diverse patient communities. In doing so, we establish a direct connection to every healthcare partner’s EMR. While the patient data is deidentified, our internal recruitment team is able to run complex inclusion/exclusion searches across every connected EMR in a single platform to generate more accurate trial match lists based on data found within the patient records. Subsequently, this has saved our recruitment team dozens (if not hundreds) of hours of manual chart review and screening unqualified candidates and gives us more medically qualified patients from the start.
Based on your experience utilizing a wide range of study recruitment tactics, what have you seen when comparing more technology-driven tools and techniques to what you consider to be more traditional methods of recruitment?
One of our sites was brought on as a rescue for a pediatric vaccine study that had inclusion/exclusion criteria based on specific time frames for previous vaccinations and other complex qualifications. With a very basic search via labs and ICD10 codes, we initially saw 45,000 potential matches for the study. Reviewing 45,000 patient charts manually would have been impossible, especially coming in late as a rescue site for study enrollment.
After refining the search with our technology, the more complex elements in the inclusion/exclusion criteria reduced the 45,000 potential matches to 1,894 eligible candidates within our database. By comparison, our participation in a centralized ad campaign generated only around 80 candidates, for whom we still needed medical records and to vet against the study criteria, lowering the number to 30 eligible candidates. Physician referrals, although stronger referrals than external ad campaigns, only amounted to about 30 valid candidates. While we had a multipronged approach to recruitment, the technology enabled us to successfully exceed the sponsor’s targets and be recognized as a top enroller for the study.
About The Expert:
Todd Albin is the chief executive officer of Cedar Health Research, an independent, technology-enabled site and patient network in the United States conducting Phase 1-4, multi-therapeutic clinical trials. His mission is to revolutionize the way clinical trials are conducted to bring life-changing treatments to patients faster and more efficiently. Albin’s efforts were recently rewarded by the Society for Clinical Research Sites by winning the 2023 Site Patient Recruitment Innovation Award.