DLNM Prediction & Interpretation

What problem does it solve?

This Skill helps statisticians and epidemiologists understand and extract meaningful insights from complex Distributed Lag Non-Linear Models (DLNMs) by interpreting predictions and model outputs.

Core Features & Use Cases

• Generate Predictions: Create predictions across the exposure-lag-response surface using crosspred().
• Extract Key Metrics: Obtain overall RRs, RR matrices, and cumulative RRs from model outputs.
• Interpret Centering: Understand and adjust the reference value (cen) for meaningful risk ratio calculations.
• Summarize Effects: Reduce complex models to interpretable one-dimensional curves using crossreduce().
• Use Case: After fitting a DLNM to air pollution and mortality data, use this Skill to generate and visualize the exposure-response curve and the lag-response curve at a specific pollution level, with clear interpretation of the centering value used.

Quick Start

Use the DLNM Prediction & Interpretation skill to generate predictions from the fitted model 'model' using the crossbasis object 'cb', centering at the median exposure.