Table 2.
Temporal trends of population viral suppression (multivariate analysis), ANRS 12249 TasP trial (2012 to 2016)
| Variable | Model 1 | Model 2 | ||
|---|---|---|---|---|
| Estimate [95% CI] | p‐value | Estimate [95% CI] | p‐value | |
| Calendar time (annual increase)a | 0.019 [0.00; 0.03] | 0.012 | 0.018 [0.00; 0.03] | 0.031 |
| Time since cluster opening (annual increase)a | 0.045 [0.03; 0.06] | <0.001 | 0.044 [0.02; 0.07] | <0.001 |
| Intervention arm (vs. control, at cluster opening) | 0.013 [−0.06; 0.03] | 0.554 | −5.031 [−0.07; 0.01] | 0.090 |
| Interaction of intervention arm on time since cluster openinga | 0.024 [−0.01; 0.06] | 0.131 | 0.026 [0.00; 0.05] | 0.021 |
| Proportion of male (within cluster)b | −0.150 [−0.43; 0.13] | 0.295 | ||
| Proportion of 16 to 29 years old (within cluster)b | −0.036 [−0.46; 0.39] | 0.868 | ||
| Proportion of 60 or more years old (within cluster)b | 1.332 [0.11; 2.56] | 0.030 | ||
| Proportion with at least secondary level of education (within cluster)b | −0.013 [−0.32; 0.30] | 0.930 | ||
| Proportion being employed (within cluster)b | 0.726 [−0.05; 1.50] | 0.065 | ||
| Proportion being student (within cluster)b | −0.171 [−0.67; 0.33] | 0.499 | ||
| Proportion being single (within cluster)b | 0.319 [−0.04; 0.68] | 0.106 | ||
| Proportion from poor households (within cluster)b | 0.089 [0.00; 0.18] | 0.056 | ||
| HIV prevalence (within cluster)b | −0.381 [−0.88; 0.12] | 0.142 | ||
Model 1 is adjusted on calendar time, time since cluster opening and trial arm. Model 2 is also adjusted on cluster‐level sociodemographic characteristics. Models are computed at cluster‐day level.
a
If the estimate is 0.044, it means that every year PVS increase by +4.4% (everything else being equal)
b
if the estimate is 1.332 and if the covariate increases by 0.1 (i.e. by 10%, for example from 20% to 30%), everything else being equal, PVS would increase by 0.1 × 1.332 = 0.1332, that is, by 13.3%.