Derives links between net benefit, PPV, and subgroup calibration to explain when decision curves exceed treat-all or treat-none lines, and proposes PPV curves as a complement.
Pepe, Jing Fan, Ziding Feng, Thomas Gerds, and Jorgen Hilden
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.AP 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Why decision curves go above or below treat-all and treat-none: a PPV- and calibration-based guide for clinical prediction models
Derives links between net benefit, PPV, and subgroup calibration to explain when decision curves exceed treat-all or treat-none lines, and proposes PPV curves as a complement.