Women's Access to Health Services in Ghana

'Leave no one behind' is one of the  UN SDG Framework's  six guiding principles. To assess progress of many of the  SDG goals , we need to know where people are. Some SDGs focus on issues facing specific cohorts of people based on age and sex. To help meet these needs,  WorldPop  has published annual gridded population estimate datasets for the years 2000 through 2020, and partnered with  Esri  to publish them in the  ArcGIS Living Atlas of the World  making them particulaly easy to access and integrate with your own data.

For Ghana, we used these WorldPop data to produce the map above, featuring the capital, Accra, and the surrounding region. The map shows the change in the population of women between the ages of 35 and 40 years. Notice a general trend for higher density in the outlying regions, but for some years this trend reverses.

In this case study we use population cohort estimates to explore accessibility of health services for women of child bearing age on the example of Ghana. Women of child bearing age are classified to be between the ages of 15 and 50 years, from now on we will call them WOCBA. Specifically, we want to know the proportion of WOCBA who can access a hospital within one hour. Women require access to reproductive health services from adolescence to the end of their reproductive years, independent of whether they give birth or not. However, especially the women giving birth need essential care to protect their own as well as their newborns' health. Inadequate health systems or facilities which are too far away therefore form serious danger to women's health ( Singh et al., 2014 ).

The population cohorts allow us to select and combine the relevant sex and age groups. Below you can see the distribution of WOCBA in Ghana in 2020. Most WOCBA live in the large cities in the South of Ghana, up to nearly 3000 WOCBA per square kilometer in the most densly populated areas. The North instead is dominated by large sparsly populated areas, following the general distribution of Ghana's total population.

Next, we need to know where hospitals are located and how accessible they are. Fortunatly, there are open access data available for this, too. The Living Atlas has the right dataset at hand, showing  health facilities for Sub-Saharan Africa , which we can filter to only show hospitals. Even more useful is a  gridded travel time surface  stating how many minutes each raster cell is away from a hospital. It is worth knowing that the travel time surface was generated with a mixture of motorised and non-motorised travel in mind:

• Motorised travel was assumed to follow the road network assigning 80 km/h on primary roads, 60 km/h on secondary roads, and 10 km/h on tertiary roads
• Other grid cells were assigned a walking speed of up to 5 km/h, depending on land cover and topography

Therefore, where motorised travel is not available to WOCBA or road conditions do not allow the travel speeds mentioned above this surface is likely to underestimate the time it takes to reach a hospital.

The map below shows the areas from which you can reach a hospital within one hour (green, yellow and orange area), from red areas it takes one to two hours. There are even a handful of dark red grid cells indicating travel times of over two hours.

Let's find out how many WOCBA live outside a one hour catchment area of hospitals and might struggle to receive the health care they need in time. This is easy, we simply sum up all WOCBA where they overlap with a travel time greater than 60 minutes.

The result shows that 32,046 WOCBA are unable to reach potentially life saving health care within one hour. Use the swiper from left to right to explore the spatial distribution of the percentage of WOCBA without quick access to health care on both the regional and district level.

We can see that on a regional level, the accessibility to health services is very high. Only two regions have small proportions of WOCBA that do not meet the criteria, Northern with 3.5% and Upper West with 5.1%. You can click on the region in question to see the percentage value.

This positive picture changes when looking at the district level. Now we can see seven districts with 10-20% of WOCBA outside the 1 hour catchment and one district (North Gonja) with an astonishing 53%. The majority of women here are potentially at risk due to hospitals not being quick enough to reach. Most of the affected districts are located in the North where the population and hospital density is low, however, even in the South deficiencies are now appearing. Keeping the limitations of the travel time surface in mind, it is likely that the actual number of affected districts is even higher.

This case study falls within  SDG goal 3  'Good Health and Well-Being' and shows the issue of scale. On a national and even a regional level Ghana's access to health care for WOCBA is satisfying, however, on the district level imbalances appear.

A recent publication focussing on this particular issue across sub-Saharan Africa came to a similar conclusion:

"While many of the countries met the targets at the national level, we found large within-country variation. Monitoring [of SDGs] under the current guidelines, using national averages, can mask these areas of need, with potential consequences for vulnerable women and children. It is imperative therefore that indicators for monitoring the availability and geographical accessibility of health care reflect this need, if targets for universal health coverage are to be met by 2030." ( Wigley et al., 2020 )

Would you like to perform a similar analysis, for example for a different country or a different year? WorldPop's population estimates in 1km and 100m resolution are available  here. 

For more ideas and best practices when working with WorldPop's multidimensional population data in ArcGIS Pro, have a look at this  new series of blog posts .