How Does GDP Correlate With The People With HIV And Their Acess To Antiretroviral Therapy?

Introduction

During these past decades, humanity has been the witness to great health events, among which for example we can find the improvement of treatments and destigmatization of a disease such as HIV. It was not long ago that HIV and AIDS were considered the consequences of homosexuality, transsexuality, and everything that socially produced aversion or they wanted to hide – for example, it was also linked with drugs and sex work. Even though some of these still prevail today, we have not gotten rid of all the stigma (nowadays the perspective is not as close to the LGBTIQ+ community, but it is strongly linked with poverty).

The COVID-19 pandemic has shown the international community that not only changes can be made and that after a long virus, there are mechanisms that, through international cooperation, can turn life back to what it was before. The main problem with HIV is that for example, even though it exists in every part of the world, it has never been considered a priority since it affected either minorities or poor people.

As we mentioned before, antiretroviral therapy has gotten better and its coverage is constantly increasing within those less wealthy areas. Moreover, the internationalization of the disease and the increasing infection and detection of patients have produced that the virus has turned into an international concern, rather than something that people needed to hide for. Becoming an international issue, the situation has improved mainly in the matters of pricing for the treatment, since the year 2000 the prices have dropped 100-fold, making the mean price $100 per person annually, producing more than 8 million patients living in developing countries to be now receiving it.

For this reason, the topic of our project radicates from the GDP of each sample country, alongside the retroviral treatment data that we have on each correlated with the number of people with the virus. The main question we are trying to answer is if the GDP of each country and region affects whether or not the virus increases and if the control of the virus also varies depending on it (antiretroviral therapy access).

Results

Data set that will be used for the project

library (goalie)
library(tufte)
library(tidyverse)

## -- Attaching packages --------------------------------------- tidyverse 1.3.1 --

## v ggplot2 3.3.5     v purrr   0.3.4
## v tibble  3.1.6     v dplyr   1.0.8
## v tidyr   1.2.0     v stringr 1.4.0
## v readr   2.1.2     v forcats 0.5.1

## Warning: package 'dplyr' was built under R version 4.1.3

## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

library(wbstats)
library(flextable)

## 
## Attaching package: 'flextable'

## The following object is masked from 'package:purrr':
## 
##     compose

library(knitr)

## Warning: package 'knitr' was built under R version 4.1.3

First code: finding the data

indicators <- c(
  gdp_capita ="NY.GDP.PCAP.CD", 
  pop = "SP.POP.TOTL",
  art_coverage = "SH.HIV.ARTC.ZS" ,
  art_pregnant = "SH.HIV.PMTC.ZS" ,
  hiv_adults = "SH.DYN.AIDS.ZS"
)

wbind <- wb_data(indicators,  start_date=2020) 

countries <- wb_countries()

Second code: organizing the data

wbind <- wbind %>%
  filter(!is.na(hiv_adults)) %>% 
  filter(!is.na(art_coverage)) %>% 
  filter(!is.na(art_pregnant)) %>% 
  left_join(countries, by=c("iso2c","iso3c","country")) 

Overall comparison of HIV prevalence in a country and the ART coverage provided (lowest % of HIV cases per region)

In this first table, we can observe the countries with the lowest number of cases sorted by regions. As we can observe the country with the best performance is Algeria in the MENA region, with only 0.1% of its total population infected and 85% of the infected receiving antiretroviral treatment. Meanwhile, we are also witnessing how in areas like South Asia, the lowest infection coverage is found in Afghanistan and its art. coverage represents only 9% of the population. In this table, we can observe also how low the rate of HIV-infected people is, since it is comprehended between 0.1-0.2% of the population.

wbind %>%
  group_by(region) %>% 
  arrange(hiv_adults) %>% 
  slice(1) %>% 
  select(region,country,hiv_adults, art_coverage) %>%
  flextable() %>% 
  set_caption("ART coverage and HIV prevalence")

These are the two regions with the most proficency to have HIV classified in terms of ART coverage, and pregnant ART coverage

As we can observe, in the previous chart there were two regions that had a 0.2% of adults with HIV, which has caught our interest since the rest of the regions had a percentage of 0.1%. We investigated the sample countries in these areas thoroughly and found that in East Asia and the Pacific region and Latin America and the Caribbean region, the population, even though the contagion is high has a relatively high coverage in general. Moreover, we also found that the art. for pregnant people is positively high too. So, in this case, having a low quantity of cases does not necessarily mean that your coverage is going to be higher, as compared to the case of South Asia.

wbind %>% 
  filter(admin_region_iso3c=="EAP") %>% 
  select(country, hiv_adults : pop) %>%
  flextable() %>% 
  set_caption("East Asia and the Pacific")

wbind %>% 
  filter(admin_region_iso3c=="LAC") %>% 
  select(country, hiv_adults : pop) %>%
  flextable() %>% 
  set_caption("Latin America and the Caribbean")

Details about ART coverage compared with GDP (only the two countries with the highest/lowest ART coverage per region)

It is in these two tables that we can really observe how the art. coverage has no correlation with the GDP, since between regions we can see how countries with really low GDP have great coverage, while in the case of the countries with higher GDP they have lower coverage. Nevertheless, there are certain exceptions in this case of Portugal, but again, we cannot see correlation since the rest of countries have a similar GDP and the art coverage variates.

wbind %>%
  group_by(region) %>% 
  arrange(-art_coverage) %>% 
  slice (1:2) %>% 
  select(region,country, gdp_capita, art_coverage) %>%
  flextable() %>% 
  set_caption("Highest ART coverage per region")

wbind %>%
  group_by(region) %>% 
  arrange(+art_coverage) %>% 
  slice (1:2) %>% 
  select(region,country, gdp_capita, art_coverage) %>%
  flextable() %>% 
  set_caption("Lowest ART coverage per region")

Average ART coverage by region

As we have observed before, the countries that are part of South Asia region (which are only 2 since the rest do not have available data) is the one with the least population-for obvious reasons-, and the one with the lowest coverage, since Afghanistan counteracts Nepal. Its opposite is Subsaharan Africa, which is one of the regions with the most data and consequently the one with the highest population. Thus, is one with the highest accumulated cases and one of the areas with the most international help and efforts to fight the virus. Consequently, its weighted ART coverage average is also higher. This is one of the reasons we stress that the implementation of the 2030 agenda is very important for this century; because in the case of the Indicator n.3 (Health Indicator), thanks to international cooperation and investigation, countries with almost a 30% of the population infected like Eswatini, have currently around a 98% of ART coverage amongst those living with HIV.

wbind %>% 
  group_by(region) %>% 
  drop_na(pop) %>%      
  summarize(`Weighted average`=weighted.mean(art_coverage,pop,na.rm=TRUE),
            Countries=n(),
            `Pop (millions)`=sum(pop)/1e6
  ) %>% 
  arrange(`Weighted average`) %>% 
  flextable() %>%  # Create nice table
  colformat_double(digits = 0) %>% 
  set_caption("Average life expectancy by region")

Percentage of pregnant women living with HIV treated with ART, compared to the rest of the population with HIV (Average by region)

In general, the statistics show that for pregnant women is harder to get antiretroviral therapy than it is for a non-pregnant individual. This leads to untreated women that are giving birth to infected children, which is already pretty bad for the overall health of the country, because these kids should receive future treatment, which could have been avoided if the adequate measures had been taken, and in case they are not, they turn into potential contagion points.

wbind %>%
  group_by(region) %>% 
  arrange(art_pregnant) %>% 
  slice (1) %>% 
  select(art_coverage, art_pregnant) %>%
  flextable() %>% 
  set_caption("% Pregnant HIV+ Women with ART coverage")

## Adding missing grouping variables: `region`

GDP per capita compared to ART coverage

As we can see in figure 1, we cannot see a continuitiy between GDP per capita and Art. coverage. Moreover, as we can see in the graph the quantity of general poulation does not really affect the art coverage.

wbind %>%
  ggplot(aes(
    x = gdp_capita, 
    y = art_coverage
  )) +
  geom_point(aes(size=pop, color=region)) +
  geom_smooth(se= FALSE) +
  scale_size_continuous(
    labels = scales::number_format(scale = 1/1e6, suffix = "m"),
    breaks = seq(1e8,1e9, 2e8),
    range = c(1,20)
  ) +
  scale_x_log10() +
  theme_bw() +
  labs(
    title = "GDP per capita compared to ART coverage",
    x = "GDP per capita",
    y = "Antiretroviral therapy coverage (% of people living with HIV)" ,
    size = "Population",
    color = NULL,
    caption = "Source: World Bank, through `wbstats`"
  ) 

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

## Warning: Removed 4 rows containing non-finite values (stat_smooth).

## Warning: Removed 4 rows containing missing values (geom_point).

library(tmap)
data(World)

my_map=World %>% 
  filter(iso_a3 != "ATA") %>% 
  left_join(wbind %>% rename(iso_a3=iso3c) %>%      select(iso_a3,gdp_capita, pop,art_coverage),
            by="iso_a3")

my_map %>% ggplot(aes(fill = art_coverage)) +
  geom_sf() +
  scale_fill_viridis_c() +
  theme(legend.position="bottom") +
  labs(
    title = "Map of ART Coverage 2020",
    fill = NULL,
    caption = paste("Source: World Bank") 
  ) +
  coord_sf()

library(tmap)
data(World)

my_map2=World %>% 
  filter(iso_a3 != "ATA") %>% 
  left_join(wbind %>% rename(iso_a3=iso3c) %>%      select(iso_a3,gdp_capita, pop,hiv_adults),
            by="iso_a3")

my_map2 %>% ggplot(aes(fill = hiv_adults)) +
  geom_sf() +
  scale_fill_viridis_c() +
  theme(legend.position="bottom") +
  labs(
    title = "Map of HIV Prevalence 2020",
    fill = NULL,
    caption = paste("Source: World Bank") 
  ) +
  coord_sf()

Conclusions

The conclusion we can extract from this research project on GDI number 3 regarding the correlation between HIV contagion and antiretroviral coverage and GDP of each country, we surprisingly found that there is no correlation between these two variables. Through this investigation, we found that in certain countries, for example, Eswatini has around a 30% of the population contaminated with the virus, and a GDP per capita of 3,424 USD, has better-compared coverage than Portugal has–Portugal has 7 times the GDP per capita of Eswatini. And the same happens comparing among countries in the same region, for example in the Caribbean and Lat. Am. region, Haiti has better coverage than Suriname, having ¼ of its GDP per capita and more rate of contaminated population (the difference, in this case, is equivalent to a 45% of the people living with HIV not being able to access art. in Suriname).

Consequently, we can determine that there is no correlation between the variables that we were analyzing, consequently, for future research, we suggest that instead of considering GDP may be an analysis of the international aid, like NGOs and IOs, since they are in many cases who provide for the citizens when the government does not cover certain areas like education or healthcare. The research of this project is really important because if we do not control this silent pandemic (which remains a taboo in many parts of our society) the consequences will affect countries in which healthcare is not a powerful source, and instead of creating this network of countries in which we all participate and help each other, the dynamics of the global north and south will remain the same if there is not an improvement in those countries with less medical resources to treat the damnified population.