Table IV.
All-sites morbidity and mortality by intervention arm among all live-birth singletons*
| Intervention Arm |
Arm 1 vs Arm 3† |
Arm 2 vs Arm 3† |
Arm 1 vs Arm 2† |
||||||
|---|---|---|---|---|---|---|---|---|---|
| Outcomes | Arm 1 (N = 807)* | Arm 2 (N = 869) | Arm 3 (N = 839) | Relative risk (95% CI)* | P Value | Relative risk (95% CI)* | P Value | Relative risk (95% CI)* | P Value |
| Illness/health problem, n/N (%) | 319/787 (40.5) | 309/841 (36.7) | 286/824 (34.7) | 1.11 (0.99–1.24) | .07 | 1.05 (0.95–1.15) | .37 | 1.06 (0.96–1.17) | .23 |
| No illnesses: 0, n/N (%) | 468/787 (59.5) | 532/841 (63.3) | 538/824 (65.3) | ||||||
| Number of illnesses: 1, n/N (%) | 158/787 (20.1) | 159/841 (18.9) | 155/824 (18.8) | ||||||
| Number of illnesses: 2, n/N (%) | 83/787 (10.5) | 87/841 (10.3) | 80/824 (9.7) | ||||||
| Number of illnesses: ≥3, n/N (%) | 78/787 (9.9) | 63/841 (7.5) | 51/824 (6.2) | ||||||
| Hospitalization, n/N (%) | 39/787 (5.0) | 30/839 (3.6) | 23/824 (2.8) | 1.80 (1.09–2.97) | .02 | 1.30 (0.77–2.21) | .33 | 1.38 (0.87–2.19) | .17 |
| Respiratory illness, n/N (%) | 141/786 (17.9) | 151/836 (18.1) | 138/822 (16.8) | 1.09 (0.93–1.28) | .31 | 1.08 (0.90–1.29) | .43 | 1.01 (0.85–1.20) | .89 |
| Diarrhea, n/N (%) | 50/786 (6.4) | 36/836 (4.3) | 42/822 (5.1) | 1.19 (0.85–1.67) | .32 | 0.83 (0.53–1.32) | .44 | 1.42 (0.97–2.09) | .07 |
| Malaria, n/N (%) | 21/786 (2.7) | 18/836 (2.2) | 14/822 (1.7) | 1.50 (0.77–2.90) | .23 | 1.25 (0.63–2.48) | .52 | 1.20 (0.65–2.21) | .57 |
| Mortality | |||||||||
| Neonatal mortality (<28 d), n/N (rate/1000)‡ | 31/833 (37.2) | 37/883 (41.9) | 20/859 (23.3) | 1.56 (0.87–2.81) | .14 | 1.79 (1.08–2.97) | .03 | 0.87 (0.58–1.32) | .52 |
| Infant mortality (<6 mo), n/N (rate/1000) | 29/799 (36.3) | 39/863 (45.2) | 29/829 (35.0) | 0.99 (0.59–1.66) | .96 | 1.27 (0.87–1.87) | .22 | 0.78 (0.54–1.12) | .18 |
Since this was collected at the family level (vs by infant), data for live birth twins are unidentifiable and therefore excluded. The data presented in this table are for all singleton live births or twin births resulting in only 1 live birth.
Relative risks, 95% CIs, and P values comparing proportions for the pairwise comparisons of the intervention arms are obtained from generalized linear models with generalized estimating equations adjusting for country while controlling for cluster correlations with an independent working correlation structure and a robust sandwich estimator (empirical estimates). This specified model did not converge for hospitalization and malaria, and thus, a log-binomial model without adjust for cluster correlations was fitted. As these are exploratory analyses, no correction for multiple comparisons have been made.
The denominators for neonatal mortality rate at 28 days are all live births.