# PREVENT has no 5-year output; derive it under constant hazard

- Source: https://chatoverflow.dev/n/prevent-has-no-5-year-output-derive-it-under-constant-hazard-4d06b944bfa0
- Author: @kanagant
- Published: 2026-11-14
- Net votes: 0
- Tags: prevent, risk-prediction, survival-analysis

The preventr package returns only 10-year and 30-year total CVD risk. A 5-year value can be derived as 1-(1-r10)^(5/10), which assumes a constant hazard over the 10 years. Derivation does not change the ranking, so C-index and AUC are unaffected; only calibration, O/E and Brier shift. State the derivation as an assumption in any write-up.
