Table 1 Summary of interpretation of coefficients and intervention effects in segmented regression for interrupted time series analysis using parametrizations of Bernal et al. and Wagner et al
From: Interpretation of coefficients in segmented regression for interrupted time series analyses
|  | Bernal’s parametrization | Wagner’s parametrization |
|---|---|---|
Model equation | \({y}_{t}={\beta }_{0}+{\beta }_{1}T+{\beta }_{2}^{B}{X}_{t}+{\beta }_{3}{X}_{t}T\) | \({y}_{t}={\beta }_{0}+{\beta }_{1}T+{\beta }_{2}^{W}{X}_{t}+{\beta }_{3}{X}_{t}(T-\delta )\) |
Interpretations | Coefficients | |
Baseline level | \({\beta }_{0}\) | |
Pre-intervention trend | \({\beta }_{1}\) | |
Difference in intercepts | \({\beta }_{2}^{B}\) | \({\beta }_{2}^{W}-{\beta }_{3}\delta\) |
Immediate effect (change in levels at intervention onset) | \({\beta }_{2}^{B}+{\beta }_{3}\delta\) | \({\beta }_{2}^{W}\) |
Gradual effect (change in slopes after intervention) | \({\beta }_{3}\) | |
Post-intervention trend | \({\beta }_{1}+{\beta }_{3}\) | |