Table 3. Secondary health outcomes at endline.
| Indicator | Cronbach Alpha | Intervention | Control | ||||
|---|---|---|---|---|---|---|---|
| Diarrhea | N | % | N | % | PR (95% CI) a | PD (95% CI) b | |
| During the last 2 days, index child had three or more loose stools per day | - | 730 | 5.9 | 720 | 4.7 | 1.62 (0.71, 2.22) | 0.01 (‒0.018, 0.042) |
| During the last 7 days, index child had three or more loose stools per day | - | 731 | 7.1 | 721 | 6.0 | 1.20 (0.74, 1.93) | 0.01 (‒0.019, 0.043) |
| Water and sanitation insecurity scores | N | mean (SE) | N | mean (SE) | - | difference (95% CI) c | |
| Water-HWISE Scale d | 0.96 | 565 | 1.71 (0.37) | 388 | 2.71 (0.89) | - | ‒1.29 (‒3.19, 0.61) |
| Sanitation-Potential harmsd | 0.85 | 365 | 0.46 (0.026) | 327 | 0.50 (0.033) | - | ‒0.05 (‒0.13, 0.03) |
| Sanitation-Social expectations resultant repercussions d | 0.79 | 366 | 0.28 (0.025) | 327 | 0.30 (0.022) | - | ‒0.03 (‒0.09, 0.03) |
| Sanitation-Physical exertion or strain d | 0.57 | 366 | 0.42 (0.046) | 328 | 0.40 (0.043) | - | 0.01 (‒0.11, 0.13) |
| Sanitation-Night concerns d | 0.56 | 366 | 0.32 (0.022) | 328 | 0.37 (0.027) | - | ‒0.05 (‒0.12, 0.02) |
| Sanitation-Social support d | 0.88 | 366 | 0.10 (0.021) | 328 | 0.20 (0.023) | - | ‒0.10 (‒0.16, ‒0.43) |
| Sanitation-Physical agility d | 0.56 | 366 | 0.14 (0.017) | 328 | 0.14 (0.020) | - | 0.00 (‒0.05, 0.05) |
| Sanitation-Defecation place d | 0.81 | 366 | 0.35 (0.038) | 327 | 0.32 (0.028) | - | 0.02 (‒0.06, 0.11) |
| Water insecurity prevalence | N | % | N | % | PR (95% CI) a | PD (95% CI) b | |
| Water insecure (HWISE score 12 or more) d | - | 565 | 5.7 | 388 | 8.8 | 0.50 (0.21, 1.23) | -0.05 (-0.12, 0.03) |
Notes.
a We used log-linear binomial regression models to compare the prevalence of the outcomes between the intervention and control arms. Models accounted the stratified design by including woreda indicator variables [58], and accounted for clustering within kebeles by using generalized estimating equations with robust standard errors.
b Prevalence differences (PD) were calculated using post-estimation commands to estimate the average marginal effects.
c We used similar linear regression models to estimate the difference in the outcomes comparing the intervention and control arms.
d We asked respondents to indicate how often they felt some form of sanitation insecurity (never, sometimes, often, always). These items were then summed with all other items in that factor and divided by the numbers of items to create a score. The factors were predesignated, and based on a validation that was done in another study [23]. A higher score represents higher sanitation insecurity. d We used similar linear regression models to estimate difference comparing the outcomes between the intervention and control arms. This used a 12-item scale with four response categories (never, rarely, sometimes, often/always), and a total summed score of those response categories ranging from 0–36. A higher score indicates greater household water insecurity. We considered water insecure as a score of 12 or more, as described elsewhere [54].