Short answer: Dr. Chadi Nabhan says enrolled patients often don't reflect US demographics, so Ryght links site twins to US Census data. It uses clinicaltrials.gov history, such as prior EGFR-mutated tumor studies, to rank sites on biomarker capability.
Watch this answer from Episode 34 (starts at 9:22) · Episode 34 notes and transcript
From Episode 34: AI Is Rebuilding Clinical Trials, starting at 9:22.
Sam Nadler [9:22]: Incredibly interesting. Tell me a little bit about, we talked about it on our call, but how does diversity and the biomarker complexity kind of influence the results and or just the trials themselves?
Dr. Chadi Nabhan [9:37]: It's a really important question. When you look at existing clinical trials in The US and, cancer takes the lion's share of these trials, but even cardiovascular, metabolic, neuro neuroscience and so on, most patients that are enrolled in these clinical trials do not represent the demographics of The US. So you see often ninety to ninety nine percent of patients enrolled are Caucasian, while twenty five percent of patients in The US are considered non Caucasian. You see a lot of folks who are enrolled who have good socioeconomic status, they're able to travel, they have good psychosocial support and so on and that's really not what you see in real life. So what we do because we have the digital twins globally, at least in The US, we have the ability to link that against The US census and I think you have that, the census overlay, which is really interesting. So you're able to actually tell in the using The US census and API into The US Census, what are the demographics of the patient population each particular site serves. So if you have a particular study that requires 20 or 25% underrepresented minorities, then that becomes actually high on the list when you are deciding to tier these sites and you're able to recommend sites that serve this patient population so the denominator is higher. The biomarker overlay is really critical and as you know at least in cancer all of the trials now have biomarkers or mutations or things that you have to actually look at. The way we do that is several ways, but as you know, we have ability to look at every clinical trial that each site has done in the past because we are linked to clinicaltrials.gov where you have the repository of all clinical trials that have been done at each particular site. And if you know that a specific site has conducted prior studies that required biomarker testing, then at least you know they have that capability.
Dr. Chadi Nabhan [11:41]: If you know more granular details that this particular site has conducted studies in the EGFR mutated tumors, then you know that at least they have that patient population so you actually rank them higher because that information is available through clinicaltrials.gov. So we basically look at past experience of these sites that allows us to predict also future success If you have a sponsor that have come back and forth to this specific site many times before for their particular studies, then that site must have been successful in enrollment and so on. So that census and biomarker is key, in the cancer world and I think it's going be beyond cancer as well. But that's how we do it for now and we're able to try to partner also, we're trying to form strategic partnerships with a lot of entities out there that we can layer out the genomic data on the digital twins. That's work in progress on the roadmap.
