Annotated Bibliography

Name: Yile Xu

Selected Topic: Infrastructure and social service–Analysis of health issues and public health system development in Kenya.

Word Count: 1688

Annotation 1

Wesolowski, Amya, et al. “Quantifying the Impact of Accessibility on Preventive Healthcare in Sub-Saharan Africa Using Mobile Phone Data.” Epidemiology, Vol.26, 2015, pp. 223-228.

The development problem this article is seeking to address is the lack of preventive care and access to public health facilities for Kenyans. In Kenya, a large number of child and maternal mortalities are caused by several common factors including poor prenatal care and infectious diseases like diarrhea, malaria, and tuberculosis. These deaths are preventable by public health services like childhood immunization and antenatal care for pregnant women. Poor physical access to health facilities has been identified as an important contributor to this outcome and will be most critical in low-income settings. The main focus of this article is to quantify the relation between physical access to healthcare, actual travel behavior of a community, and the degree of healthcare received, which is the key to identify the role of physical accessibility to health care and alleviate health disparities in local community.

Using anonymized mobile phone data from 2008 to 2009, researchers developed a measure called the radius of gyrations for travel patterns or mobility of 14,816,521 subscribers across Kenya based on two spatial scales: the county level and the individual cell tower level. Travel time to the nearest health facility was calculated using a cost–distance algorithm which computed the “cost” of traveling on a regular raster grid based on information about the transportation network. To understand the outcome for various uptake of preventive healthcare interventions, survey data about basic health conditions were also collected in western Kenya and two outcome variables (completed childhood immunizations and antenatal care for pregnant women) were used. The mobility data were modeled with the estimated travel times to health facilities and data on the uptake of two preventive healthcare interventions. They reached a conclusion that long travel times to health facilities is strongly correlated with increased mobility in geographically isolated areas, and in areas with equal physical access to healthcare, high mobility predict which regions are lacking preventive care.

This article relates to Amartya Sen’s definition of human development in the perspective of overall freedom as it manages to alleviate the social unfreedom faced by Kenyans because of their lack of access to public health care. The technique of using collected mobile phone data to mapping the uptake of preventive healthcare described in this article can be an important approach for understanding the outcome of poor physical access and the development of preventive healthcare system in Kenya. Having access to public health is an important social freedom for everyone.

Annotation 2

Andrew J. Tatem, et al. “Defining approaches to settlement mapping for public health management in Kenya using medium spatial resolution satellite imagery.” Remote Sensing of Environment, Vol. 93, 2004, pp 42–52.

This article is trying to address the health consequences of urbanization in some Sub-Saharan Africa countries, especially Kenya from the public health planning perspective. In Africa, about half of the populations live in cities. Although these urban dwellers have significantly better access to health care facilities than their rural counterparts, they are at higher risk from directly transmitted diseases such as tuberculosis and HIV, as well as certain vector-borne diseases such as dengue fever due to the closer distance between settlements and higher population density. The mapping of settlements’ location, size, and distribution are crucial to the planning and management for health consequences of urbanization in the countries of SSA. However, the special resolution of previous settlement and population mappings for Kenya are proved too coarse to develop accurate population estimates at an administrative level fine enough to facilitate effective public. Therefore, the main purpose of this research is to create settlement maps across four contrasting Kenyan districts at a spatial scale fine enough to facilitate applications in public health management like malaria risk mapping, disease burden estimation, and corresponding solution development.

Both orthorectified Landsat Thematic Mapper (TM) 30 m spatial resolution imagery and Japanese Earth Resources Satellite 1 (JERS-1) Synthetic Aperture Radar (SAR) 12.5 m spatial resolution imagery were collected for four Kenyan districts: Bondo, Kisii/Gucha, Makueni, and Kwale. Supervised neural network per-pixel classifications were carried out on these satellite images to produce settlement mappings, and additional data texture layers including human population census data, land cover, and locations of medical facilities were also used for training site identification and validation. Sufficient ground data obtained allowed researchers to assess the efficacy of settlement mapping. The result proved that the combination of satellite imageries and derived texture layers is effective in identifying and delineating settlements in four districts of Kenya, and using neural network is an accurate approach to derive these maps. Therefore, this approach can be applied to the settlement and population mapping of the whole Kenya for public health applications.

This article is also related to Amartya Sen’s definition of human development in the perspective of social freedom because it manages to develop and validate a settlement mapping technique accurate enough which can be applied to the public health management of urban area in response to the health consequences of urbanization. It turns out that this method can be applied more broadly to the whole Kenya or even Sub-Saharan Africa (SSA) countries for public health applications. Therefore, the development goal of improving people’s social freedom by developing public health services mentioned by Amartya Sen is also the fundamental purpose of this article.

Annotation 3

Iyer HS, et al. “Geospatial evaluation of trade-offs between equity in physical access to healthcare and health systems efficiency.” BMJ Global Health 2020; 5:e003493.

Regarding the improvement of health services in four sub-Saharan African countries: Kenya, Malawi, Rwanda and Tanzania, this article tries to alleviate the issue of health inequality through figuring out the optimal allocation of healthcare resources and promoting equitable physical access to health facilities. When making decisions on health facility distributions, policy-makers always need to balance the trade-offs between equity of geographical accessibility and health system efficiency regarding the economies of scale. To be more specific, allocating health facilities in urban areas with more concentrated population will improve the system’s efficiency because of the economies of scale. However, preferential deployment of services in urban area will exacerbate urban rural disparities and health inequality. In order to achieve more equitable geographical accessibility for the whole country, some efficiency has to be sacrificed. Therefore, the most important idea presented in this work is developing the Geographic-Population Services Access (Geo-PSA) model in order to examine the overtime trade-off between geographical accessibility and health system efficiency of all countries which assists policy-makers to optimize their health facility distributions.

For each of the four African countries, both descriptive characteristics data and geospatial data including the location of care facilities and geographical feature maps were collected from administrative government databases and satellite imagery. Geo-PSA combined a conceptual framework for interpreting the relationship between equity and efficiency with a geospatial analytical approach. Travel time to the nearest health facility was chosen as a measure of equity while population density became a measure of efficiency due to economies of scale. The researchers used the Access Mod 5 algorithm to estimate travel time based on topography information along the way and the corresponding speed. The model produced negative correlation between population density and travel time in all country and stronger correlation in Kenya and Malawi than the other two countries, which suggested that all countries favor efficiency over equality but some countries exhibit more equitable health facility allocations.

This article is related to Amartya Sen’s definition of human development in the way of expanding people’s social freedom, specially their access to public health services. The primary goal drives this research is to alleviate the issue of health inequality in some countries but also consider national health system’s efficiency at the same time. In the process of increasing the coverage of healthcare services, the developed Geo-PSA model can be constantly applied overtime to evaluate the trade-offs between equity and efficiency and monitor progress towards equitable access of services. Moreover, this analytical model also enables policy-makers to identify locations with inefficient resource allocation (e.g. with low population density but short travel times) and respond to these inefficiencies based on their resources.

Annotation 4

Burke, Marshall, et al. “Sources of variation in under-5 mortality across sub-Saharan Africa: a spatial analysis.” Lancet Glob Health 2016; 4: e936–45

This research tries to address the relative high child mortality and its substantial variations in sub-Saharan Africa. With only 24% decline, Sub-Saharan Africa region failed to meet the Millennium Development Goal of a two-thirds reduction in under-5 mortality rate between 1990 and 2015. Moreover, large differences in progress also exist within sub-Saharan Africa countries. In order to further reduce under-5 mortality rate and at the same time address the remarkable variations in progress, it is important to understand the main sources of variations in mortality. While some believe that country-level factors like national policies and conditions drive first-order changes in mortality, others believe that subnational variations of mortality and distribution of disease is much more substantial. Regarding this controversy, the main purpose of this article is to analyze and examine the underlying determinants of under-5 mortality by developing high-resolution spatial maps overtime across sub-Saharan Africa.

The main data sources for this analysis were from the Demographic and Health Surveys (DHS) on the location and timing of 3.24 million childbirths and 393685 deaths across 28 African countries. Under-5 mortality rates of three decades: 1980s, 1990s, and 2000s of each location-based clusters were calculated by researchers, and they interpolated cluster-specific under-5 mortality rates onto a 0.1 degree latitude×0.1 degree longitude grid to estimate mortality levels for each cell of the grid to create maps of mortality rate. Based on these maps, the importance of within-country versus between-country variation factors in each decade was quantified by least- squares regression of cell-level mortality estimates on various indicator variables. The result indicated that the role of subnational factors outweighs country-level ones when explaining mortality patterns, and mortality spatial variation analysis within countries shows that mortality rates are positively correlated with regional factors including temperature, malaria, and recent conflict.

This article is related to Amartya Sen’s definition of human development in the perspective of reducing under-5 mortality rate and improving population life expectancy. The sustainable development goal for all sub-Saharan African countries is to reach the projected under-5 mortality rate of 25 deaths per 1000 children in 2030. The result of this study suggests that local determinants are likely to have greater influences on future trends in mortality and policies and interventions that target areas with exceptionally high mortality might be more effective than national policies. Further analysis on impact of the regional environmental, economic, or political conditions identified by this study can also help to improve the targeting of interventions to the areas in needed.

Conclusion

My topic is about the improvement of Kenya’s public health services. It is learned that although the child mortality rate in Kenya decreased substantially in recent decades, its reduction still did not meet the millennium goal of two-thirds between 1990 and 2015. The overall child mortality rate is still much higher compared to other countries and there are significant variations within country. Moreover, infectious diseases like HIV, tuberculosis, and malaria are also serious issues in both urban and rural area. All these public health issues are largely caused by the poorly coverage of preventive care and health facilities in some regions of Kenya. Therefore, it is crucial to analyze the causing factors of health issues and improve people’s equitable access to health services through the optimal allocation of healthcare resources, which is the focus of my exploration. These four articles provide some geospatial data science methods to analyze health issues and improve health services distribution.