Digital twins and virtual control arms
Using a model of what would have happened to a patient on standard treatment, so fewer people have to be randomised to it.
Prognostic models trained on historical trial and registry data generate an expected control trajectory for each enrolled patient, which can shrink the control arm or sharpen its estimate. Regulators have accepted external and synthetic control arms in narrow settings, mainly rare diseases and single-arm oncology submissions; in common cancers the concern is that treatment and supportive care drift faster than the model. Adoption is real but cautious.
How it works
A model predicts each participant's counterfactual outcome under control therapy; the prediction is used as a covariate or replaces part of the control arm.
- Fewer patients randomised to the inferior arm
- Faster, cheaper trials in rare subgroups
- Uses data already collected
- Bias when standard of care changes between eras
- Regulatory acceptance is case by case
- Model failure is invisible without a real control arm
Latest papers
topQuery for this technology: (TITLE:"Digital twins and virtual control arms" OR ABSTRACT:"Digital twins and virtual control arms") AND (cancer OR tumor OR tumour OR oncology OR carcinoma OR lymphoma OR leukemia OR leukaemia OR myeloma OR sarcoma OR melanoma OR glioma). Results are unfiltered search hits about Digital twins and virtual control arms, not a curated reading list.