Table 2.
Multilevel logistic models for the effect of consultation year on suicide exploration
| Model 1 | Model 2 | |||||
|---|---|---|---|---|---|---|
| Predictors | Odds Ratios | CI | p | Odds Ratios | CI | p |
| (Intercept) | 2.33 | 1.48–3.66 | < 0.001 | 12.72 | 5.72–28.27 | < 0.001 |
| Consultation year: 2018 (Ref = 2017) | 1.48 | 1.01–2.16 | 0.045 | 1.33 | 0.90–1.97 | 0.158 |
| Patients’ gender: Female (Ref = Male) | 0.64 | 0.43–0.98 | 0.027 | |||
| Patients’ age | 0.97 | 0.96–0.98 | < 0.001 | |||
| Random Effects for general practice | ||||||
| Within-group (residual) variance1 | 3.29 | 3.29 | ||||
| Between-group variance | 1.06 | 1.13 | ||||
| ICC | 0.24 | 0.26 | ||||
| N | 35 | 35 | ||||
| Observations | 624 | 624 | ||||
| Marginal R2 / Conditional R2 | 0.009 / 0.250 | 0.070 / 0.308 | ||||
CI = Confidence Interval, ICC = Intraclass Correlation Coefficient
1 The within-group (residual) variance in logistic regression is equal to π2/3