Table 3. Effect of the intervention on PMTCT indicators, stratified by treatment intensity, Tanzania, 2014–2015a.
| Outcome | Baseline (births in 2014) | Endline (births in 2015) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Txb intensity | Control | Tx intensity | Control | DiDh Unadjustedj | DiD Adjustedk | ||||||
| High | Low | High | Low | High | Low | High | Low | |||||
| (95% CIi) |
(95% CI) |
(95% CI) |
(95% CI) |
|||||||||
| Total | 1830 | 171 | 391 | 590 | 94 | 210 | 374 | _ | _ | _ | _ | |
| Women retained in care (%)b,c | 1830 | 35.0% | 35.6% | 34.6% | 48.6% | 42.6% | 47.3% | 0.9 (-15.5, 17.3) | -5.6 (-19.1, 7.8) | -6.8 (-19.9, 6.3) | -2.0 (-16.0, 11.9) | |
| Full Sample | ||||||||||||
| Women retained in care (%)b,c | 1348 | 51.5% | 45.3% | 46.4% | 70.2% | 61.9% | 59.0% | 6.1 (-9.1, 21.3) | 4.0 (-7.5, 15.5) | -3.6 (-16.4, 9.1) | 4.8 (-8.9, 18.4) | |
| Women with evidence of care | ||||||||||||
| Women initiating ARTd | 1544 | 40.6% | 50.9% | 55.1% | 45.9% | 44.5% | 46.0% | 14.4 (4.9, 23.8)*** | 2.7 (-10.0, 15.3) | 8.1 (-2.6, 18.8) | 3.8 (-6.1, 13.6) | |
| Timing of ART initiatione | 787 | 27.9 | 26.7 | 27.3 | 23.3 | 29.3 | 24.3 | -1.6 (-6.5, 3.2) | 5.6 (2.8, 8.5)*** | -1.5 (-6.1, 3.0) | 5.3 (2.4, 8.2)** | |
|
Women with MPR≥95%f,g |
820 | 18.8% | 14.3% | 21.2% | 34.8% | 24.1% | 22.8% | 14.3 (2.0, 26.6)** | 8.2 (-9.2, 25.5) | 13.6 (2.5, 24.6)*** | 4.8 (-14.9, 24.4) | |
| Women with MPR≥80%f,g | 820 | 33.2% | 24.1% | 34.9% | 52.5% | 41.8% | 41.6% | 12.6 (-2.4, 27.5)* | 11.0 (-6.8, 28.7) | 11.6 (-1.6, 24.9)* | 5.0 (-11.5, 21.4) | |
* significant at the α = 0.10 level
**significant at the α = 0.05 level
***significant at the α = 0.01 level
a) estimates were weighted for selection, site size and, at baseline, missing women (see methods)
b) Tx = treatment, grouped into a “high” category, scores above the median, or a “low” category, at or below the median (see methods)
c) Women retained in care calculated for n = 1830 all women in the sample, and n = 1348 women with evidence of HIV care
d) Number of women initiating ART calculated for the n = 1544 women who did not have evidence of beginning ART prior to pregnancy; women were considered to have initiated ART if they had any evidence of ART use between pregnancy and 90 days postpartum
e) Timing of ART initiation calculated for n = 787 women who had an ARV start date on record, and who began ART after pregnancy
f) MPR ≥95 (and MPR≥80) are the women that have 95% adherence (or 80% adherence) or greater according to the MPR calculation, defined as the number of pill days dispensed over the number of days elapsed from the infant’s birth to 90 days postpartum
g) MPR calculated for n = 820 women who had complete ARV dispensing information
h) DiD = difference-in-differences estimate, generated through the linear regression of the outcome with an interaction variable for time and treatment status (see methods)
i) CI = confidence interval
j) Unadjusted for baseline imbalance
k) Adjusted for average number of CHWs per site and percent of HEID testing