A Trially.ai Perspective on the Lupus Rescue Study: Accelerating Recruitment with AI
The Quest Diagnostics case study, "Revitalizing Recruitment In A Lupus Rescue Study," details how a top pharmaceutical company’s lupus study struggled for 2 years. The clinical trial’s highly restrictive criteria—specifically requiring patients to be in an active flare—could not be met through traditional recruitment methods.
The study was ultimately rescued through a collaboration with PatientsLikeMe and Quest Diagnostics, which leveraged Quest's de-identified clinical database to pinpoint patients matching the narrow criteria. While this data-driven approach was an improvement, viewing it through the lens of Trially’s AI-powered platform reveals opportunities where the entire process could have been further streamlined and accelerated.
The core challenge in the lupus study was identifying patients who met the very specific inclusion criteria. Quest's solution was to use its database with broad criteria like ICD-10 codes, age, and location to find likely candidates. This is precisely the problem Trially Match is designed to solve, but with greater speed and precision.
Trially’s proprietary AI analyzes both structured and unstructured data from any source—including EHRs, lab results, and physician notes. Instead of relying on general codes, its LLM-powered platform can parse complex clinical details to confirm nuanced requirements like an "active flare" with approximately 95% accuracy. Had the study sites used Trially Match, they could have bypassed the initial 2-year struggle by generating a pre-qualified, stack-ranked list of candidates from their own patient populations from day one. This approach has been shown to decrease screen failure rates by as much as 73% by ensuring only the highest-quality candidates are identified from the outset.
Beyond initial identification, the rescue study relied on email outreach, web screeners, and phone screening—a multi-step process where patient drop-off is common. This is where Margo, Trially's AI agent, offers a transformative advantage by automating the time-intensive tasks of patient engagement. Margo functions as a "Super-Coordinator" to convert identified matches into enrolled participants.
Proactive Outreach: Once Trially Match identifies candidates, Margo initiates timely, multi-channel outreach via SMS, voice, and email, replacing the passive reliance on email open rates with proactive engagement.
Intelligent Pre-Screening: Margo conducts IRB-approved pre-screening to confirm both eligibility and interest. In the lupus study, 47% of patients who completed the web screener passed, implying a 53% failure rate. Margo’s intelligent screening, informed by deep EHR data, minimizes this by engaging only the most qualified leads and generating real-time summaries for the clinical team. This directly addresses the site staff burden, a key cause of trial delays.
Seamless Scheduling: Margo handles the final logistical steps by managing appointment scheduling and sending automated reminders to reduce no-shows. For candidates who are interested but not yet ready, Margo strategically re-engages them over time, building a trial-ready population for the future.
The Quest Diagnostics and PatientsLikeMe collaboration successfully rescued the lupus study and affirmed the research on the effectiveness of a data-driven, patient-centric approach. However, the Trially platform represents the next evolution of this strategy. By leveraging AI to enhance every step of the process, Trially not only accelerates recruitment but also optimizes site resources, ultimately helping get life-changing therapies to the patients who need them most at breakthrough speed..
