Logistic regressions of uptake of an NHSHC estimating the effect of the intervention, with control letter as a baseline; adjusted model includes sex (reference category male), age (reference category 40–44 years), and ethnicity (reference category white) as covariates and GP practice as a random effect. P-values for random effects are log likelihood tests between models that include and omit the random effect
| Model 1a: (Pseudo R2= 0.006) | Model 2b: (Pseudo R2 =0.408) | Model 3c: Pseudo R2= 0.410) | ||||
|---|---|---|---|---|---|---|
| OR (CI) | P-value | OR (CI) | P-value | OR (CI) | P-value | |
| Intercept | 0.08 (0.07 to 0.09) | <0.001 | 0.10 (0.05 to 0.19) | <0.001 | 0.07 (0.03 to 0.15) | <0.001 |
| Intervention | 1.70 (1.46 to 1.99) | <0.001 | 1.62 (1.37 to 1.90) | <0.001 | 2.28 (1.46 to 3.55) | <0.001 |
| Female | — | — | 0.99 (0.84 to 1.16) | 0.86 | 1.32 (1.02 to 1.71) | 0.036 |
| 45–49 years | — | — | 0.70 (0.57 to 0.86) | <0.001 | 0.69 (0.56 to 0.85) | <0.001 |
| 50–54 years | — | — | 0.71 (0.57 to 0.90) | 0.0047 | 0.78 (0.62 to 1.00) | 0.04501 |
| 55–59 years | — | — | 0.71 (0.55 to 0.93) | 0.0132 | 0.77 (0.59 to 1.01) | 0.055 |
| 60–64 years | — | — | 0.55 (0.38 to 0.79) | 0.0011 | 0.55 (0.38 to 0.78) | <0.001 |
| 65–69 years | — | — | 0.64 (0.42 to 0.98) | 0.0395 | 0.64 (0.42 to 0.98) | 0.0418 |
| 70–74 years | — | — | 0.61 (0.37 to 1.03) | 0.06 | 0.65 (0.39 to 1.10) | 0.110 |
| Black | — | — | 0.93 (0.77 to 1.12) | 0.44 | 0.92 (0.76 to 1.12) | 0.405 |
| Asian | — | — | 0.75 (0.49 to 1.14) | 0.18 | 0.74 (0.49 to 1.13) | 0.161 |
| Mixed | — | — | 1.35 (0.90 to 2.03) | 0.14 | 1.37 (0.92 to 2.06) | 0.125 |
| Other | — | — | 0.64 (0.48 to 0.86) | 0.0031 | 0.64 (0.48 to 0.86) | 0.0032 |
| Unknown | — | — | 0.13 (0.08 to 0.22) | <0.001 | 0.13 (0.08 to 0.22) | <0.001 |
| Interventiona Female | — | — | — | — | 0.62 (0.44 to 0.86) | 0.004495 |
| Variation in intercept | — | — | 1.27 | <0.001 | 1.31 | <0.001 |
| between practices, SD | ||||||
| Variation in slopes for the | — | — | — | — | 0.56 | <0.001 |
| intervention between practices, SD | ||||||
↵a Model 1: an unadjusted model that contains only the intervention as a fixed effect.
↵b Model 2: adds practice as a random effect and the demographics as fixed effects.
↵c Model 3: an interaction between sex and the intervention is added, and as well allowing varying slopes for the different GP practices, to investigate differences in effectiveness of the intervention depending on sex and on the individual GP practice. CI = confidence interval. NHSHC = NHS Health Check. OR = odds ratio. SD = standard deviation.