Our digital twin-generating AI models are trained on vast amounts of historical, individual-level data. By accounting for the information contained in these models, TwinRCTs are designed to provide more power with smaller control groups than other randomized trial designs.
The data collected from each patient at their first visit is input into the pre-trained AI model to generate their digital twins. These digital twins provide comprehensive, longitudinal predictions of every patient's potential outcomes if they were to receive placebo and/or standard of care.
Each patient in a TwinRCT is randomized to a treatment or control group. At the end of the trial, information from the patients' digital twins is used to compute a precise, unbiased estimate of the treatment effect using our regulatory-qualified PROCOVA™ framework.
AI-powered TwinRCTs hit enrollment targets faster because they require fewer patients to achieve the same power compared to traditional trial designs.
Many patients are wary of participating in clinical trials due to the prospect of receiving a placebo. TwinRCTs make control groups smaller, increasing a patient’s chance of receiving the experimental treatment.
Incorporating the comprehensive information from patients’ digital twins into trial analyses allows for more precise estimates of treatment effects at the population, subgroup, and even individual patient level.