Assessing the effects of maternal HIV infection on pregnancy outcomes using cross-sectional data in Malawi

BMC Public Health

Table 8 Confounder identification

Variable	Outcome	Exposure
	Birth weight (p-value)	Perinatal death(p-value)	HIV status (p-value)
Age of a mother	<0.001	<0.001	<0.001
Region	NA	0.006	<0.001
Residence	0.016	0.946	<0.001
Education	0.011	0.43	0.012
Size of child at birth	<0.001	0.06	0.31
Ever had terminated pregnancy	0.845	<0.001	0.007
Anaemia	<0.001	0.90	<0.001
Marital status	NA	NA	<0.001
place of delivery	NA	0.249	0.97
Wealth	0.97	0.286	0.008
(a) Confounders that influence HIV and the outcomes for the 2010 MDHS
	Birth weight (p-value)	Perinatal death(p-value)	HIV status (p-value)
Age of a mother	<0.001	0.226	<0.001
Region	0.133	0.35	<0.001
Residence	0.49	0.126	<0.001
Education	0.096	0.716	<0.001
BMI	0.017	NA	0.577
Size of child at birth	<0.001	<0.001	0.304
Ever had terminated pregnancy	0.730	<0.001	0.001
Anaemia	0.21	0.50	<0.001
Marital status	NA	NA	<0.001
Place of delivery	NA	0.26	0.41
Wealth	0.001	0.287	0.017
(b) Confounders that influence HIV and the outcomes for the 2015-16 MDHS

Confounder Identification A verification of the variables was done by fitting a linear regression model for the effect of each confounder on child birth weight and did a crosstabulation using Pearsons Chi-square test for the effect of the confounders on perinatal mortality. We also checked the association between the identified variables and mothers HIV-infection status. Variables that were not known to influence either HIV status or birth weight or perinatal death were labelled as “not applicable”

ISSN: 1471-2458