Why decision curves go above or below treat-all and treat-none: a PPV- and calibration-based guide for clinical prediction models

TL;DR

This paper clarifies how net benefit in clinical prediction models relates to calibration and positive predictive value, making decision-curve analysis more interpretable for clinicians.

Contribution

It introduces two new interpretations of net benefit based on calibration and PPV, and proposes positive predictive value curves as a practical tool.

Findings

Net benefit comparisons can be expressed through threshold-specific observed risk.

Net benefit relates to positive predictive value, aiding interpretation.

Positive predictive value curves complement decision curves for clinical utility.

Abstract

Net benefit is widely used and reported to evaluate the clinical utility of prediction models, yet its interpretation often remains difficult in practice. In this didactical note, we develop two complementary interpretations that make net benefit easier to understand for clinical audiences. We show that comparisons with treat-none and treat-all can be expressed through threshold-specific observed risk in patients above and below the decision threshold, linking decision-curve performance to calibration in clinically relevant subgroups. We also show how net benefit relates to positive predictive value, offering a more intuitive explanation of when acting on model predictions is justified. We derive and illustrate these results and propose positive predictive value curves as a practical complement to decision curves.