Fig. 2

Measures of predictive performance under predictor measurement heterogeneity between validation and implementation setting. Results shown for random predictor measurement only (A), additive systematic predictor measurement only (B), and multiplicative systematic predictor measurement heterogeneity only (C). The vertical dashed line indicates predictor measurement homogeneity between validation and implementation setting. The x-axes show measurement heterogeneity parameters describing the predictor measurement at implementation relative to the predictor measurement at validation, where σ_ε denotes random deviations from the measurement at validation, ψ denotes an additive shift with respect to the measurement at validation, and θ denotes a systematic multiplicative association with the measurement at validation. Note that additional simulation scenarios were run to smooth the plots