Table 2

Main results: link between SIA exposure and maternal health service uptake

Dependent variables: maternal healthcare access	Delivery‡			Antenatal care§
Dependent variables: maternal healthcare access	At home	At private facility	At public facility	No of antenatal care visits	No of tetanus toxoid injections
Model no	(1)	(2)	(3)	(4)	(5)
EXP_PREG	0.007	−0.065†	0.010	−0.025*	0.001
EXP_PREG	(−0.013 to 0.027)	(−0.093 to –0.036)	(−0.009 to 0.029)	(−0.054 to 0.003)	(−0.008 to 0.009)
Level 1 observations (child)	34 713	34 713	34 713	35 019	36 585
Level 2 observations (LGA)	686	686	686	686	687
Akaike information criterion	26 422.845	17 601.384	30 318.818	193 640.288	111 227.001
Prob.>χ²	<0.001	<0.001	<0.001	<0.001	<0.001

Notes. Two-level regressions with LGA random effect (see notes a and b for model specification). Sample include pregnancies for children born between October 2000 and December 2017 (the available data for the covariates means that the sample is limited to live births). Dependent variables as defined in table header. Main results only, omitting maternal and household determinants of healthcare access, survey round dummy, constant and multilevel variance parameters. Interaction terms between SIA exposure and survey round dummy were not statistically significant and the associated models have been omitted from the main results. Coefficients reported, 95% CIs in brackets.
*P<0.10.
†P<0.01.
‡Two-level logistic regression models with binary outcome indicators of facility access for delivery.
§Two-level linear regression models of number of antenatal care visits and tetanus toxoid injections prior to delivery. Two-level Poisson regression models yielded similar results and have been omitted from reporting.
LGA, local government area; SIA, supplementary Immunisation activity.