Table 2. Estimated effect of GEHIP on under age 5 child mortality, Upper East Region, Ghana, 2005–2014.
| Unadjusted Model (1) | Full Model (2) | Final Model (3) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | p-value | HR | 95% CI | p-value | HR | 95% CI | p-value | |
| Treatment | |||||||||
| Comparison area | 1.00 | ref | 1.00 | ref | 1.00 | ref | |||
| Treatment Area (GEHIP Area) | 0.97 | (0.66–1.42) | 0.888 | 0.96 | (0.65–1.41) | 0.826 | 0.97 | (0.66–1.42) | 0.910 |
| Period | |||||||||
| Pre: All person months prior to July 2011 (time before GEHIP began in July 2011) | 1.00 | Ref | 1.00 | Ref | 1.00 | ref | |||
| Post: All person time July 2011 or later | 1.17 | (0.80–1.71) | 0.430 | 1.22 | (0.80–1.86) | 0.359 | 1.22 | (0.82–1.81) | 0.327 |
| GEHIP Incremental Effect (DiD) (Treatment * Post) | 0.55 | (0.29–1.03) | 0.064 | 0.53 | (0.28–1.02) | 0.056 | 0.52* | (0.28–0.98) | 0.045 |
| Post-neonate infant interaction terms (months 1–12) | |||||||||
| Treatment * Post-Neonate-Infant | 1.14 | (0.66–1.96) | 0.639 | 1.13 | (0.66–1.92) | 0.663 | 1.13 | (0.66–1.93) | 0.665 |
| Period * Post-Neonate-Infant | 0.74 | (0.39–1.40) | 0.352 | 0.73 | (0.39–1.38) | 0.329 | 0.72 | (0.38–1.37) | 0.319 |
| GEHIP DiD * Post-Neonate-Infant | 1.34 | (0.45–3.99) | 0.597 | 1.40 | (0.47–4.20) | 0.546 | 1.38 | (0.46–4.12) | 0.564 |
| Post-infant interaction terms (months 12–60) | |||||||||
| Treatment * Post-Infant | 1.18 | (0.77–1.80) | 0.453 | 1.15 | (0.76–1.73) | 0.513 | 1.16 | (0.77–1.77) | 0.474 |
| Period * Post-Infant | 0.38** | (0.20–0.71) | 0.003 | 0.41** | (0.21–0.78) | 0.007 | 0.37** | (0.20–0.70) | 0.002 |
| GEHIP DiD * Post-Infant | 1.91 | (0.83–4.40) | 0.126 | 1.94 | (0.84–4.47) | 0.118 | 1.96 | (0.87–4.45) | 0.107 |
| Gender | |||||||||
| Male | 1.00 | ref | |||||||
| Female | 0.78** | (0.64–0.94) | 0.010 | ||||||
| Multiple birth | |||||||||
| Singleton | 1.00 | ref | 1.00 | ref | |||||
| Multiple | 3.16*** | (2.05–4.88) | 0.000 | 4.08*** | (2.76–6.02) | 0.000 | |||
| Gestation a | |||||||||
| 9 months | 1.00 | ref | |||||||
| < 9 months | 4.52*** | (2.96–6.91) | 0.000 | ||||||
| Birth spacing | |||||||||
| ≥ 24 months | 1.00 | ref | |||||||
| < 24 months | 1.71*** | (1.32–2.21) | 0.000 | ||||||
| Maternal Age | |||||||||
| 15–20 | 1.00 | ref | |||||||
| 20–34 | 1.07 | (0.77–1.49) | 0.690 | ||||||
| 35–49 | 1.00 | (0.63–1.57) | 0.983 | ||||||
| Parity | |||||||||
| Nulliparous | 1.00 | ref | |||||||
| Primipara | 1.03 | (0.74–1.45) | 0.843 | ||||||
| Multipara | 0.74 | (0.55–1.01) | 0.056 | ||||||
| Grandpara | 0.77 | (0.52–1.15) | 0.205 | ||||||
| Marital status | |||||||||
| Unmarried | 1.00 | ref | |||||||
| Polygynous | 1.08 | (0.74–1.56) | 0.703 | ||||||
| Monogamous | 0.93 | (0.68–1.26) | 0.627 | ||||||
| Literacy | |||||||||
| No | 1.00 | ref | |||||||
| Yes | 0.98 | (0.77–1.24) | 0.838 | ||||||
| Household wealth | |||||||||
| Less poor (Q2-5) b | 1.00 | ref | |||||||
| Poorest (Q1) | 1.06 | (0.82–1.36) | 0.657 | ||||||
| Nearest clinical health facility to household (km) c | 1.02 | (0.99–1.06) | 0.178 | 1.02 | (0.99–1.06) | 0.273 | |||
| Summary statistics: | |||||||||
| Observations (Person-time) | 424,909 | 424,909 | 424,909 | ||||||
| Number of. Subjects | 14,454 | 14,454 | 14,454 | ||||||
| Number of Deaths | 445 | 445 | 445 | ||||||
| Number of Clusters | 66 | 66 | 66 | ||||||
| F statistic (d.f.) | 3.84 (9) | 10.75 (23) | 8.57 (11) | ||||||
* p<0.05
** p<0.01
*** p<0.001
a Self-reported duration of gestation
b Principal component quintile for relative household economic status = 1 for poorest and zero for other quintiles
c Time-varying covariate; distances tabulated for each household to nearest CHPS facility for every month from 2000–2014