An Epidemiological Study of Women and Children’s Health in Yemen with Special Reference to Socioeconomic Determinants : Subnational household survey conducted in Yemen

IPPE
Maternal and child health are the least equitable health care in most developing countries. In Yemen, improving maternal and child health remains a major public health challenge. Inequalities in socioeconomic status (SES) are correlated with maternal and child health morbidity and mortality. SES is also a determinant of the use of maternal and child health (MCH) care services. Measurement of SES is complicated, perhaps because of its multidimensionality, representing components at the personal, household and community levels. Inadequate assessment of SES will undermine understanding of its impact on health. Constructing SES indices that reflect contextual realities is crucial in understanding SES’s role in maternal and child health, particularly in rural settings of low and middle countries. The main aim of this dissertation was to assess the role of SES as a determinant of MCH in rural Yemen. The dissertation constituted three sub-studies (Study I–III). Study I developed contextually-relevant SES indices to be used for investigations into MCH in Yemen. Study II examined the associations between the developed SES indices and indicators of utilization of MCH services. Study III investigated the associations between the socioeconomic status indices and women’s Female Genital Cutting (FGC) status and attitude toward FGC.

This dissertation was done on the basis of the subnational household cross-sectional survey performed in rural Yemen through 2008-2009, which was organized in six rural districts of four provinces. Altogether, 7341 households in the selected areas were contacted for interview and 7330 females of childbearing age from these households were invited for interview. All the necessary data on participants’ SES and demographic features, female’s reproductive features, maternal health care usage through the latest gestation, delivery, and maternal death, along with neonatal and infant death were accessible for 7295 (99%) women of childbearing age (15–49 years) in Study I, for 6907 (94%) women with a birth history in Study II, and for 7076 (96%) women of reproductive age in Study III. The SES indices were constructed with factor analysis using principal component analysis (PCA) in Study I. The relationships between socioeconomic status indices and respective outcomes were estimated with logistic regression in Studies I, II and III.

In the PCA, three SES indices (wealth, educational and housing quality) were pulled, which collectively explicated 54 % of the total discrepancy in SES. Factor scores were inferred and classified into tertiles.

In adjusted models that included several potential confounding variables, higher tertiles of wealth and educational indices were conversely related with spontaneous abortion (SA), stillbirth (SB), neonatal and infant death. No statistically significant associations were observed between any of the SES indices and maternal mortality.

Higher tertiles of wealth and educational indices but not housing quality index were correlated with usage of fair adequate antenatal care (adjusted odds ratio [AOR] 3.34, 95% CI 2.76-4.04, and AOR 3.68, 95% CI 3.15-4.31, respectively).
Lower tertiles in any of the SES indices were significantly associated with home delivery without the assistance of a trained practitioner vs. facility delivery (adjusted odds ratio [AOR] 0.66, 95% CI 0.52-0.86, AOR 0.43, 95% CI 0.34-0.55, and AOR 0.58, 95% CI 0.46-0.74, respectively).

Higher tertiles in any of the SES indices were associated with more likelihood of using any contraceptive methods (AOR 1.26 95% CI 1.10-1.45, AOR 1.44 95% CI 1.27-1.65, AOR 1.33 95% CI 1.16-1.52, respectively). At the same time, higher tertiles in each of the SES indices were associated with less likelihood of having unmet need for contraception (AOR 0.57, 95% CI 0.50-0.66, AOR 0.95, 95% CI 0.83-1.08, AOR 0.83, 95% CI 0.72-0.95, respectively)

The prevalence of women’s FGC was 48%, while daughters’ FGC was 34%. Almost 45.8% of the women surveyed believed that FGC practice should be discontinued. Older age, family marriage, and lower tertiles of wealth and education indices were associated with higher odds of having been circumcised and having had their daughters circumcised as well as positive attitude towards FGC. Early marriage was also associated with increased odds of FGC practice.

Mothers who had FGC were more likely to be in favour of their daughters having FGC, as well as in favour of the continuity of FGC practice (AOR 7.40 95% CI 6.01–9.13 and AOR 8.93 95% CI 7.30–10.91, respectively). Also, daughters were more likely to be subjected to FGC practice when the attitudes toward FGC of their mothers are supportive (AOR 2.26 95% CI 1.95–2.61)

Evidence from this study shows that using context-based household and neighbourhood factors, SES indices can be constructed and used to classify reproductive age women into different risk profiles concerning several indicators of MCH outcomes. However, no strong evidence was observed that the constructed SES indices had any influence on maternal mortality, an indication that maternal mortality may perhaps be influenced by standards of obstetric care, rather than SES of women. This perspective requires confirmation in further studies.

Further studies are required to assess the various ways through which housing quality, including the electricity, sanitation, and source of water could influence the occurrence of neonatal and infant death. The inconsistency in modern contraceptive use and fertility trends could be explained by the factors that may influence a couple’s decision to use contraceptive methods.

Studies in the area of maternal and child health inequality will attempt to get around the methodological challenges in making SES a precise and reliable measurement. Also, studies strive to relate it to maternal and child health outcomes as well as examine the variations and the specific outcome associations while adjusting for significant potential confounding covariates. Our findings have addressed that and can help in developing appropriate strategies that will reduce maternal and child health inequalities

Our study showed the importance of adapting appropriate policy that should capitalize on our findings and invest in girls’ education and women’s empowerment, which might enhance healthy behaviour and practice. It is important to engage religious and community leaders effectively to continuously improve the awareness of women’s rights and to support the change of attitudes and practice of FGC in the younger generation.