Ссылка
click to show
click to show
QuantHealth preprint predicted VESALIUS-CV hazard ratio close to actual trial result
On 31 August 2025, QuantHealth released a preprint describing a computer simulation of the VESALIUS‑CV trial. The simulation, posted to the OSF archive since 11 August 2025, estimated a hazard ratio of 0.78 for the primary composite endpoint.
On 25 June 2026, ClinicalTrials.gov published the trial’s observed hazard ratio of 0.75. VESALIUS‑CV evaluated evolocumab, an LDL‑lowering antibody, with a median follow‑up of 55.2 months; the first major event occurred in 336 of 6 129 evolocumab‑treated participants versus 443 of 6 128 placebo recipients.
The OSF analysis defined the endpoint as cardiovascular death, myocardial infarction, or ischemic stroke, whereas the trial registry listed the first of coronary death, myocardial infarction, or ischemic stroke. These two overlapping composite measures correspond to the predicted 0.78 and observed 0.75 hazard ratios.
QuantHealth’s model was trained on de‑identified insurance claims and electronic health records, mapping the drug to associated proteins, diseases, and biological pathways. Using protocol criteria, it generated a virtual control group from existing records and estimated time‑to‑event under evolocumab and placebo scenarios. Prior to VESALIUS‑CV, the approach was validated against 22 completed cardiovascular trials, achieving an F1 score of 0.83 in 24 of those comparisons.
All authors were QuantHealth employees during the study, and the work was funded by QuantHealth.
🔗 Read original →
2 ·