You are viewing the site in preview mode
Skip to main content
| Â | PHQ-9 score | Depression status |
|---|
Joint significance tests | F (df); p | F (df); p |
|---|
Time | 37.0 (3, 998); < 0.0001 | 7.6 (3, 998); < 0.0001 |
Time X intervention | 20.4 (4, 998); <0.0001 | 15.7 (4, 998); < 0.0001 |
Interaction effects | Beta (SE); p | OR (95% CI) |
Intervention X Month 2 | -4.03 (0.49); <0.0001 | 0.21 (0.13, 0.36) |
Intervention X Month 6 | -3.28 (0.50); <0.0001 | 0.13 (0.07, 0.25) |
Intervention X Month 12 | -2.89 (0.50); <0.0001 | 0.23 (0.12, 0.43) |
Intervention X Month 18 | -3.40 (0.50); <0.0001 | 0.16 (0.08, 0.33) |
- SE = standard error; OR = odds ratio; CI = confidence interval; df = degrees of freedom; PHQ-9 = Patient Health Questionnaire (9-item version)
- All models included these baseline covariates: age, relationship status (in a committed relationship; yes/no), any secondary schooling (yes/no), newly diagnosed with HIV (yes/no), PHQ-9 score (or depression status (yes/no) for that model) at eligibility screening, undetectable HIV viral load (yes/no). The model specification included time as a categorical variable with indicators for each follow-up assessment (month 2, month 6, month 12, month 18). The joint tests of overall effect of time and intervention are Type 3 tests of fixed effects. The coefficients of the main effects of time are not shown here. The key coefficients of interest are the interaction terms between Intervention and time, which are the (covariate-adjusted) difference between the Intervention (intervention = 1) group and the control group (Intervention = 0); a p-value < 0.05 indicates a statistically significant difference between the means of the two study arms. Further, note that to estimate the mean in the intervention group, one must add the main effect of time to the interaction between intervention and time
