Table 5.
Delivery location characteristics by group and period
| Characteristic | Pre trial period | Trial period | Crude Estimate of Change in Prevalencea | Crude Estimate of Difference in Change in Prevalenceb | Adjusted Estimate of Difference in Change in Prevalence (95 % CI), p-valuec | |||
|---|---|---|---|---|---|---|---|---|
| Intervention | Control | Intervention | Control | Intervention | Control | |||
| Deliveries, N | 2,041 | 2,305 | 3,766 | 3,960 | ||||
| Delivery location, (%) | -- | |||||||
| Hospital | 24.9 | 29.8 | 46.8 | 44.3 | 21.9 | 14.5 | 7.4 | |
| Clinic | 2.5 | 2.3 | 7.0 | 1.1 | 4.5 | −1.2 | 5.7 | |
| Home/Other | 72.6 | 67.9 | 46.2 | 54.6 | −26.4 | −13.3 | −13.1 | |
| Birth at facility level (hospital or clinic), (%) | 27.4 | 32.1 | 53.8 | 45.4 | 26.4 | 13.3 | 13.1 | 7.70 (−2.61, 18.01), p = 0.1434 |
| Birth at facility with C-section or any neonatal care capabilitiesd, (%) | 26.2 | 31.0 | 53.3 | 45.2 | 27.1 | 14.2 | 12.9 | 7.36 (−3.39, 18.12), p = 0.1797 |
| Birth at facility that has C-section capabilities, (%) | 26.1 | 31.0 | 47.5 | 40.5 | 21.4 | 9.5 | 11.9 | -- |
| Birth at facility that has bag and mask capabilities, (%) | 10.7 | 11.1 | 43.7 | 41.9 | 33.0 | 30.8 | 2.2 | -- |
| Birth at facility that has oxygen or mechanical ventilation capabilities, (%) | 20.8 | 26.6 | 47.4 | 42.6 | 26.6 | 16.0 | 10.6 | -- |
aCrude change in prevalence = (percentage during the trial period minus percentage during the pretrial period)
bCrude difference in change in prevalence = (percentage during the trial period minus percentage during the pretrial period for the treatment group) minus (percentage during the trial period minus percentage during the pretrial period for control group)
cTest to assess whether the change between the pretrial and trial periods differed by treatment group. The adjusted percent difference of the percent differences is the estimate of the difference in changes in probability from pretrial to trial period for the given characteristic in the treatment arm minus changes in probability from the pretrial to trial period for the given characteristic in the control arm. P-values were calculated from generalized linear models accounting for the cluster-level variance and adjusted for randomization strata
dNeonatal care capabilities include bag and mask, oxygen or mechanical ventilation. Facilities with any of the capabilities are tested for differences between the pretrial and trial periods by treatment group