Maternal Mortality in Sub-Saharan Africa

Introduction

The variable we will be focusing on is maternal mortality ratio (national estimate, per 100,000 live births), wich is the SDG indicator 3.1.1, world bank code « SH.STA.MMRT », in Sub-Saharan Africa. We explore average and extreme levels, the relationship between the variable and real per capita income, and a map with comments. For the purpose of this project, we will use data from the World Development Indicators database using wbstats. In fact, this variable seems relevant to study, given that every year, 287,000 women worldwide die from pregnancy-related causes. What’s more, it seems particularly interesting to analyze on the African continent, because despite recent progress, it is in Africa that we find the highest maternal mortality ratio, with 57% of all maternal deaths occurring on this continent, although the data obviously varies from country to country. By comparison, in the industrialized world, one woman in 4,700 runs the risk of dying from pregnancy-related complications, whereas for African women the figure is one in 39. This problem is mainly exacerbated by a lack of access to basic healthcare during pregnancy and childbirth, with only around half of the 123 million women who give birth each year receiving prenatal and childbirth care, as well as their newborn children. The problem is that in developing countries, there are too few properly equipped health centers, and those that do exist rarely provide all the care that women and their babies need: indeed, the highest number of maternal deaths occurs in countries where women are least likely to receive care during childbirth from a qualified professional, such as a midwife, doctor or representative of another medical profession.

Maternal mortality: average and extreme levels, the state of play in 2020

my_indicators <- c(
  my_var = "SH.STA.MMRT", 
  gdp_capita ="NY.GDP.PCAP.CD", 
  pop = "SP.POP.TOTL"
  )

countries <- wb_countries()

ind <- wb_data(my_indicators, start_date = 2000, end_date= 2022) |> 
  left_join( countries |> select(-country), by="iso3c")

world2020 <- ind %>% filter(date=="2020")

world2020 |> 
  group_by(region) |> 
  drop_na(pop) |>     
  summarize(Average=weighted.mean(my_var,pop,na.rm=TRUE),
            Countries=n(),
            Missing=sum(is.na(my_var)),
            `Pop (millions)`=sum(pop)/1e6
            ) |>     
  arrange(Average) |> 
  flextable() |>  
  colformat_double(digits = 2)

region	Average	Countries	Missing	Pop (millions)
Europe & Central Asia	10.90	58	10	922.35
North America	19.97	3	1	369.58
Middle East & North Africa	46.32	21	0	479.97
East Asia & Pacific	49.69	37	11	2,340.35
Latin America & Caribbean	79.98	42	10	650.53
South Asia	121.85	8	0	1,882.53
Sub-Saharan Africa	491.44	48	0	1,151.30

world2020 |> 
  arrange(-my_var) |>   
  select("5 top countries"=country,region,`maternal mortality per 100.000 life-birth`=my_var) |> 
  slice(1:5)  |>      
  qflextable()

5 top countries	region	maternal mortality per 100.000 life-birth
South Sudan	Sub-Saharan Africa	1,223
Chad	Sub-Saharan Africa	1,063
Nigeria	Sub-Saharan Africa	1,047
Central African Republic	Sub-Saharan Africa	835
Guinea-Bissau	Sub-Saharan Africa	725

To begin this analysis we decide first to compare the result between the different continents in 2020 in terms of maternal mortality ratio. The results were really clear by underlying the case of Sub Saharan Africa, which is the most affected region by this issue. Indeed according to the world bank figures, 760 women out of 100,000 die as a result of giving birth in Sub Saharan Africa, which is much above the average of North america (12 out of 100 000) or Europe (25 out of 100 000) or even South Asia which is the second continent most affected by this issue with 401 maternal death out of 100 000. Moreover, when we focus more on a national scale to see which countries are the most affected we can see that out of the five countries most affected by maternal mortality, four are in sub-Saharan Africa. These findings lead us to focus on the case of Sub-Saharan Africa, which appears to be the most relevant region for tackling the issue of maternal mortality.

Maternal mortality: a matter of GDP?

afri_count <- ind %>%
  filter(region == "Sub-Saharan Africa")

mort2020 <- afri_count %>% filter(date=="2020")

world2020 |>
  ggplot(aes(
      x = gdp_capita, 
      y = my_var
      )) +
  geom_point(aes(size=pop, color=region)) +
  geom_smooth(se = FALSE) +
  scale_x_log10(
    labels = scales::dollar_format(),
    breaks = scales::log_breaks(n = 4)
  ) +
  scale_size_continuous(
    labels = scales::number_format(scale = 1/1e6, suffix = "m"),
    breaks = seq(1e8,1e9, 2e8),
    range = c(1,20)
    ) +
  theme_minimal() +
  labs(
    title = "Maternal mortality according to GDP per capita in 2020",
    x = "GDP per Capita (log scale)",
    y = "Maternal mortality (per 100,000 live births)",
    size = "Population",
    color = NULL,
    caption = "Source: World Bank"
  )

Now that we have established the fact that maternal mortality is the highest in Sub-Saharan Africa, it seems relevant to compare this data with the level of GDP per capita, in order to establish a potential link between the two. Thus, as we can see on this graph on a world wide scale there the lower the GDP per capita the higher the maternal mortality rate. Indeed this graph perfectly shows that Europe (yellow) and North America (blue) are concentrated on the lower right corner, which means that their GDP per capita is the highest, while their maternal mortality rate is the highest. On the contrary, the Sub-Saharan Africa region (pink) is spread on the upper left corner, meaning that the GDP per capita is lower with higher maternal mortality rate. With this analysis on a global scale, we can confirm that there is indeed a link between GDP and mortality rates. It now seems appropriate to focus on a smaller scale in order to take this analysis further.

Maternal mortality and GDP per capita in Sub-Saharan Africa

Map of Africa only:

data(World)

World <- World %>% filter(continent=="Africa") %>% 
  filter(iso_a3 != "ATA") %>% # remove Antarctica
  left_join(mort2020 %>% mutate(iso_a3 = iso3c), by="iso_a3",suffix=c("",""))

map1 <- World %>% 
  ggplot(aes(fill = my_var)) +
  geom_sf() +
  scale_fill_viridis_c() +
  theme(legend.position="bottom") +
  labs(
    title = "Map of Maternal mortality incidence in 2020",
    fill = NULL,
    caption = paste("Source: World Bank") 
   
  )

map2 <- World %>% 
  ggplot(aes(fill = gdp_capita)) +
  geom_sf() +
  scale_fill_viridis_c() +
  theme(legend.position="bottom") +
  labs(
    title = "Map of GDP per capita in 2020",
    fill = NULL,
    caption = paste("Source: World Bank") 
   
  )

map1+map2

To better understand this phenomenon, we decided to compare the maternal mortality rate with the level of economic development of each sub-Saharan African country, based on their respective GDP. The aim is to understand to what extent the degree of correlation between these data is linked, on a more national scale. As we can see on these maps, again the lower the GDP, the higher the mortality rate, and vice versa. For instance, we can see that Tchad has a GDP of about less than 2000, while its maternal mortality incidence is around 100 000. In contrast, South Africa, with a GDP of around 5,000, has a maternal mortality rate of well under 100 per 100,000.

Country individual evolution

Try to get the evolution for each country:

afri_count <- ind %>%
  filter(region == "Sub-Saharan Africa")

ymin <- 500 
ymax <- 5000

afri_count %>%
  ggplot() +
  geom_line(aes(x = date, y = my_var, color = "My Variable")) +
  geom_line(aes(x = date, y = gdp_capita, color = "GDP Capita")) +
  facet_wrap(~country) +
  scale_color_manual(values = c("My Variable" = "blue", "GDP Capita" = "red")) +
  scale_y_continuous(trans = "log10", limits = c(ymin, ymax), oob = scales::oob_squish)

We can see that the trend is towards a fall in the maternal mortality rate, in parallel with an average GDP per capita. However, this correlation reaches its limits when, using this graph, we compare the evolution of these data over time. For example, if we focus on Nigeria, which has seen its GDP per capita double over the last twenty years, its maternal mortality rate has fallen only slightly. Rwanda, on the other hand, has not experienced any particular increase in its GDP per capita, but has seen its maternal mortality rate fall particularly sharply.

Conclusion

In our analysis, we have shown that there is a link between GDP per capita and maternal mortality rates. This correlation therefore underscores the intricate interplay between economic development and maternal health outcomes, because countries with higher economic resources often possess better healthcare infrastructure, greater access to medical facilities, and improved maternal healthcare services. Conversely, nations with lower economic resources may struggle to invest adequately in healthcare, resulting in poorer maternal health outcomes. While GDP alone does not determine maternal health outcomes, this comparative analysis sheds light on the importance of economic development in shaping healthcare access and maternal mortality rates, which explains why developed regions such as Europe have a lower mortality rate compared to the Sub-Saharan region. It underscores the need for targeted interventions and investments in healthcare infrastructure and services, particularly in low-income countries, to improve maternal health outcomes and reduce mortality rates. Moreover, we also observed that this correlation is confirmed at a more precise scale, because as we already explained, in this region the lower the GDP the highest the mortality rate. However, it has also shown us that it is important not to make generalisations, even if the region as a whole has a very high mortality rate, because some countries manage to maintain a very low mortality rate, but once again with a very high GDP. However, it seems necessary to qualify this analysis. A comparison of GDP per capita and maternal mortality rates shows that they do not necessarily evolve at the same rate, or even that there is sometimes no correlation between them. This can be explained by the fact that changes in GDP per capita can take time to translate into tangible improvements in maternal health. Even if a country is experiencing rapid economic growth, it can take years or even decades for investments in health infrastructure and social programmes to translate into a significant reduction in the maternal mortality rate. As a result, there may be a temporal divergence between changes in GDP per capita and changes in the maternal mortality rate. Moreover, the link between GDP per capita and the maternal mortality ratio is not the only factor to be considered. Other variables, such as access to healthcare, the quality of maternal health services, cultural practices and government policies, can also influence maternal health outcomes. For example, some countries may experience a decline in maternal mortality despite moderate economic growth, thanks to targeted investments in maternal healthcare, women’s education and improved access to health services.