Table 3 Marginal effects on likelihood of attending Health Check from single-level and multi-level logistic regression
Predictor variables | Marginal effects | |
|---|---|---|
Single-level | Multi-level | |
Invitation method (ref. standard letter) | ||
 Telephone | .162*** | .180*** |
(.0190) | (.0374) | |
 Personalised letter | .0118 | .0400** |
(.0173) | (.0182) | |
Age (years) | .00660*** | .00611*** |
(.00144) | (.00151) | |
Gender | .0434*** | .0443*** |
(male = 0, female = 1) | (.0157) | (.0164) |
Deprivation quintile | .0441*** | .0202** |
(1 = most to 5 = least deprived) | (.00633) | (.00843) |
Area morphology (urban = 0, rural = 1) | −.0149 | −.0357* |
(.0184) | (.0195) | |
CVD risk (% 10-year score) | −.00854*** | −.00738*** |
(.00270) | (.00270) | |
Practice size | 7.30e-06*** | 3.17e-06 |
(2.69e-06) | (1.38e-05) | |
Health Checks outside working | −.0375** | −.0828 |
hours (no = 0, yes = 1) | (.0148) | (.0786) |
Telephone invitations outside working | −.108*** | −.0685 |
hours (no = 0, yes = 1) | (.0194) | (.0973) |