Introduction

The global pattern of death is evolving. In the last two decades, the world has shifted from infectious diseases to chronic diseases. The change is not equal among regions.

data <- read.csv("20222703 Causes Of Death Clean Output V2.0.csv")

data_clean <- data %>%
  rename(
    cause = Causes.name,
    deaths = Death.Numbers,
    country = Entity,
    year = Year
  ) %>%
  filter(year >= 2000, deaths > 0) %>%
  mutate(region = countrycode(
    country,
    "country.name",
    "continent",
    custom_match = c("America" = "Americas", "Kosovo" = "Europe")
  )) %>%
  filter(!is.na(region))

# Group causes
data_clean <- data_clean %>%
  mutate(cause_group = case_when(
    cause %in% c("Cardiovascular diseases","Neoplasms","Chronic respiratory diseases",
                 "Diabetes mellitus","Chronic kidney disease","Digestive diseases") ~ "Non-communicable diseases",
    cause %in% c("Lower respiratory infections","Diarrheal diseases","Tuberculosis",
                 "HIV/AIDS","Malaria","Neonatal disorders","Maternal disorders") ~ "Communicable diseases",
    TRUE ~ "Other"
  ))

data_grouped <- data_clean %>%
  group_by(region, year, cause_group) %>%
  summarise(deaths = sum(deaths), .groups = "drop")

totals <- data_grouped %>%
  group_by(region, year) %>%
  summarise(total = sum(deaths), .groups = "drop")

data_percent <- data_grouped %>%
  left_join(totals, by = c("region","year")) %>%
  mutate(percent = deaths / total * 100)

global_percent <- data_percent %>%
  group_by(year, cause_group) %>%
  summarise(percent = mean(percent), .groups = "drop")

Global mortality has shifted

Non-communicable diseases are currently the leading causes of deaths around the world, while deaths caused by communicable diseases have been declining significantly over time. This is due to the significant improvement that has been witnessed in healthcare, hygiene, and disease prevention methods. The increased cases of chronic illnesses indicate the effects of aging and unhealthy lifestyles.

p1 <- ggplot(global_percent %>% filter(cause_group != "Other"),
             aes(year, percent, color = cause_group)) +
  geom_line(linewidth = 1.2) +
  labs(title = "Chronic diseases now dominate global mortality",
       y = "Percentage of deaths") +
  theme_minimal()


ggplotly(p1, config = list(displayModeBar = FALSE))

Regional differences exist in disease transitions

There is an unequal distribution in disease transitions from communicable to non-communicable diseases. While Europe and the Americas are mainly experiencing chronic diseases, other regions like Africa have many communicable diseases. This regional variation is due to the difference in economic development and access to healthcare facilities.

p2 <- ggplot(data_percent %>% filter(cause_group != "Other"),
             aes(year, percent, color = cause_group)) +
  geom_line() +
  facet_wrap(~region, ncol = 2) +   # <-- was default, now 2 columns (more space)
  labs(title = "Regions are transitioning at different speeds",
       x = "Year", y = "Percentage of deaths", color = "Cause type") +
  theme_minimal() +
  theme(
    strip.text = element_text(face = "bold", size = 11),
    axis.text.x = element_text(size = 9),
    panel.spacing = unit(1.2, "lines")   # more space between panels
  )

ggplotly(p2, config = list(displayModeBar = FALSE))

Major causes of deaths have changed

The importance of different types of causes of deaths has been changed for all areas. There has been an increase in the contribution of non-communicable diseases to the number of total deaths, whereas the proportion of deaths due to communicable diseases has seen a decrease.

selected <- c(
  "Cardiovascular diseases",
  "Neoplasms",
  "Lower respiratory infections",
  "Chronic respiratory diseases",
  "Digestive diseases"
)

start <- data_percent %>%
  filter(year == 2000 & cause_group != "Other") %>%
  group_by(region, cause_group) %>%
  summarise(start = mean(percent), .groups="drop")

end <- data_percent %>%
  filter(year == 2019 & cause_group != "Other") %>%
  group_by(region, cause_group) %>%
  summarise(end = mean(percent), .groups="drop")

plot_data <- left_join(start, end, by=c("region","cause_group"))

p3 <- ggplot(plot_data, 
             aes(y = factor(cause_group, levels = c(
               "Communicable diseases",
               "Non-communicable diseases"
             )))) +

  geom_segment(aes(x = start, xend = end, yend = cause_group),
               color = "gray30", linewidth = 1.2) +

  geom_point(aes(x = start, color = "2000"), size = 3) +
  geom_point(aes(x = end, color = "2019"), size = 3) +

  facet_wrap(~region, ncol = 2) +

  scale_color_manual(values = c("2000"="#E15759","2019"="#4E79A7")) +

  labs(
    title = "The relative importance of causes of death has shifted",
    subtitle = "Share of total deaths (%) in 2000 vs 2019",
    x = "Percentage of deaths",
    y = "Cause type",
    color = "Year"
  ) +

  theme_minimal() +
  theme(
    plot.title = element_text(face="bold"),
    strip.text = element_text(face="bold")
  )

ggplotly(p3, config = list(displayModeBar = FALSE))

Profiles of regional deaths today

Modern profiles of deaths by region exhibit distinct disparities in the structure of causative factors. Most regions have a large number of non-communicable disease deaths, especially in wealthier regions; on the other hand, developing regions have a considerable amount of communicable diseases.

p4 <- ggplot(data_percent %>% filter(year==2019 & cause_group!="Other"),
             aes(region, percent, fill=cause_group)) +
  geom_col(position="fill") +
  scale_y_continuous(labels = percent_format()) +
  labs(title="Regional death profiles differ significantly",
       y="Share of deaths") +
  theme_minimal()

ggplotly(p4, config = list(displayModeBar = FALSE))

Stages of health transition in the world

Various regions have varied positions in the health transition spectrum. Europe and America are typified by low communicable diseases but high levels of non-communicable diseases. On the other hand, Africa displays high communicable diseases and low non-communicable diseases. There are regions in between.

scatter <- data_percent %>%
  filter(year == 2019 & cause_group != "Other") %>%
  group_by(region, cause_group) %>%
  summarise(percent = mean(percent), .groups = "drop") %>%
  pivot_wider(names_from = cause_group, values_from = percent)

scatter[is.na(scatter)] <- 0

p5 <- ggplot(scatter,
             aes(x = `Communicable diseases`,
                 y = `Non-communicable diseases`,
                 label = region)) +
  geom_point(aes(color = region), size = 4) +
  geom_text(nudge_y = 1.5, nudge_x = 0.5, size = 4) +
  scale_color_brewer(palette = "Set2") +
  labs(
    title = "Regions occupy different stages of the epidemiological transition",
    x = "Communicable (%)",
    y = "Non-communicable (%)",
    color = "Region"
  ) +
  theme_minimal()

ggplotly(p5, config = list(displayModeBar = FALSE))

Conclusion

It is not that our world is getting healthier; it is undergoing a shift in how we die. As chronic diseases take center stage worldwide, communicable diseases remain a prevalent issue in some parts of the world. This underscores the development and healthcare gaps across regions.