Cori Ruktanonchai | October 2015
Posted on

Presentation at the Global Maternal Newborn Health Conference, October 20, 2015

Background: Early adolescent pregnancy presents a major barrier to the health and wellbeing of young women and their children, as recognised by international agreements and goals. Localized pockets of spatial inequalities likely reflect underlying levels of deprivation, as well as access to reproductive health services. The use of spatial statistics has therefore become an important tool in guiding policy decisions and focusing resources, particularly in the post 2015 Sustainable Development Goals era. This research demonstrates how a range of geospatial tools can be utilised to illustrate regional patterns of adolescent first birth using national household survey data.

Methods: The most recent Demographic and Health Surveys (DHS) among women aged 20 to 29 in Tanzania, Kenya, and Uganda were utilised. Prevalence maps were created from the GPS-located cluster data utilising adaptive bandwidth kernel density estimates. To map adolescent first birth with estimates of uncertainty, a Bayesian hierarchical regression modelling approach was used, employing the Integrated Nested Laplace Approximation technique.

Results: Heat maps were produced quantifying regional heterogeneity of adolescent first birth. Marked geographic patterns were observed throughout the region, with increased prevalence of first birth under 16 years in Kenya and Uganda. Heterogeneity was also observed across age class, with greater variation among younger adolescents. Maps of predicted prevalence are also presented, and emphasize within-country heterogeneity across neighbouring districts.

Conclusion: The growing availability of comprehensive, geo-referenced data from sources such as the DHS suggests continued opportunities to improve mapping of adolescent motherhood and other MNH indicators. These results illustrate the utility of spatial statistical methods in visualizing the health and wellbeing of women and children, and can be used to identify areas of high need or target resources. Such approaches will increasingly become an important tool for informing policy and making evidence-based decisions.