PhaseV Validates GLP-1 Adverse Event Model on Patient Data

PhaseV said in a press release that it validated a causal model for predicting weight changes, adverse events and treatment dropout among patients taking GLP-1 and related incretin medications. The company adapted the model using licensed MarketScan claims data and Veradigm electronic health record data.

PhaseV tested the model against patient trajectories that were excluded from its training data. Its predictions tracked observed changes in weight, the timing of a first adverse event and the timing of treatment dropout, according to the company.

The model links dose changes with weight, adverse events and dropout over time. It can estimate how different dose escalation schedules could affect a patient's risk of side effects or discontinuation, using treatment patterns observed in routine care.

The validation was carried out and reported by PhaseV itself; the release does not describe a peer reviewed publication or an independent assessment.