Preventing malaria in pregnancy through community-directed interventions: evidence from Akwa Ibom State, Nigeria

Malaria Journal

Table 5 Robustness checks

	At least one ANC visit	Slept under net during pregnancy	Slept under net before interview	Took malaria prevention drug	Took at least 2 SP doses
Panel A: Bordering LGAs only
CDI effect	-0.047	0.051	0.090	-0.002	0.328
	(-0.188 - 0.093)	(-0.007 - 0.110)	(0.049 - 0.130)	(-0.168 - 0.164)	(0.217 - 0.440)
Observations	1732	1732	1732	1732	1732
R-squared	0.24	0.08	0.09	0.15	0.32
Panel B: Pairwise Matching
CDI Effect	0.086	0.090	0.048	0.166	0.401
	(0.079 - 0.094)	(0.068 - 0.113)	(0.017 - 0.079)	(0.141 - 0.192)	(0.389 - 0.413)
Treatment LGA	Onna	Eket	Eket	Ikot Abasi	Ikot Abasi
Control LGA	Mbo	Mbo	Mbo	Onna	Onna
Observations	934	914	914	940	940
R-squared	0.184	0.048	0.041	0.122	0.319

Notes: 95% confidence intervals in parentheses. Results reflect coefficients from a heteroskedasticity-adjusted linear probability model. Numbers in parentheses are standard errors clustered at the local government area (LGA) level. Multivariate controls include age, ethnicity, religion, marital status, respondent's educational attainment, occupational status and household wealth as measured by a principal component-based asset index. Based on a sample of 2652 observations.

ISSN: 1475-2875