Table 3.
OLS regression of the change in careseeking and mortality among children 2 to 59 months in Malawi, predicted by implementation strength and contextual factors
| Outcome | Predictors | Estimate | SE | P |
|---|---|---|---|---|
| Change in careseeking between baseline and endline (% points) | Intercept* | 6.83 | 4.52 | 0.145 |
| District population (total population/100,000) | −0.27 | 0.56 | 0.627 | |
| Health facility density (per 10,000 total population) | −3.27 | 5.31 | 0.544 | |
| Facility worker density (per 10,000 total population) | −0.49 | 0.32 | 0.134 | |
| iCCM-ready HSA density (per 1,000 under-five children) | 1.17 | 2.18 | 0.596 | |
| Change in mortality rate between baseline and endline (deaths per 1,000 live births) | Intercept* | −37.03 | 18.30 | 0.055 |
| District population (total population/100,000) | 1.16 | 2.25 | 0.610 | |
| Health facility density (per 10,000 total population) | 13.06 | 21.48 | 0.550 | |
| Facility worker density (per 10,000 total population) | 1.19 | 1.28 | 0.363 | |
| iCCM-ready HSA density (per 1,000 under-five children) | −3.07 | 8.82 | 0.731 |
HSA = health surveillance assistants; iCCM = integrated Community Case Management; OLS = ordinary least square; SE = standard error.
*
Intercept interpretable as the expected change for average district population (426,300), with a facility and facility worker density of zero and an iCCM-ready HSA density of zero.