Global Disease Outbreaks related to Water, Sanitation, and Hygiene (WASH)

Author

Ranya Mulchandani

Published

December 6, 2025

1.0 Introduction

The Global Pandemic and Epidemic-Prone Disease Outbreaks dataset includes information on over 3,000 pandemic- and epidemic-prone disease outbreaks associated with more than 90 different infectious diseases, occurring globally from January 1996 to the present.

The aim of this analysis is to assess the proportion of global outbreaks captured in this dataset from the Disease Outbreak News (DON) that were due to water-borne diseases (e.g. cholera, diarrhoeal diseases) i.e. ‘WASH-related’ and where in the world they typically occurred.

This dataset was downloaded from (Disease Outbreaks Data Project 2024)

Full details on its curation are published by (Torres Munguía et al. 2022)

2.0 Methods

The following section describes the steps taken to load and prepare data for analysis in R (R Core Team 2019).

As the data was obtained from a freely available online source, no specific details are provided here on the exact data curation. This can be read directly in the published article.

2.1 Load libraries

Code

library(tidyverse)
library(readxl) # part of tidyverse - reading excel file
library(here) # file paths
library(janitor)
library(fs) # to use dir_tree()
library(ggplot2)
library(scales)
library(gt)

2.2 Import data

Code

# load raw global outbreak data
# outbreaks_raw <- read_excel("/cloud/project/data/raw/disease_outbreaks_HDX.xlsx") 
raw_data <- read_excel(here::here("data/raw/disease_outbreaks_HDX.xlsx"))

# load transmission route list per icd103n (table generated by AI) 
transmission_route <- disease_data_tbl <- read_delim(
  here::here("data/raw/transmission_route.csv"),
  delim = ";",
  comment = "",
  show_col_types = FALSE
)

2.3 Clean data

Code

# prepare data for analysis 
pre_processed_data <- raw_data |> 
  # janitor package to clean col names
  janitor::clean_names() |> 
  # drop 1st row after titles
  slice(-1) |> 
  # remove out rows from coronavirus dashboard as only interested in DONs
  filter(do_ns != "Coronavirus dashboard") 

# clean up transmission route table to allow for joining to main dataset 
processed_transmission_route <- transmission_route |> 
  mutate(wash_related = str_detect(`WASH-Related?`, pattern = "Yes"))

# join onto main dataset 
processed_data <- pre_processed_data |> 
  left_join(processed_transmission_route, by = c("icd103n" = "Disease"))

# write out processed data to folder - careful not to overright previous data
write_csv(processed_data,
          here::here("data/processed/my-processed-data.csv"))

3.0 Results

3.1 Visualisations

Figure 1 shows the number of outbreaks over time, split by whether WASH-related

WASH-related outbreaks only made up a small proportion of overall disease outbreaks (as reported in WHO’s DONs).
A large peak in 2009 was largely due to pandemic influenza (H1N1 influenza virus).
Number of outbreaks significantly increased in 2022-2025 compared to earlier years.

Code

# look at how many outbreaks there were over time, split by WASH vs non-WASH
processed_data |> 
  mutate(wash_related = factor(wash_related, levels = c(TRUE, FALSE))) |> 
  group_by(year, wash_related) |> 
  summarise(count = n()) |> 
  ungroup() |> 
  filter(!is.na(wash_related)) |> 
  
  ggplot() + 
  geom_col(aes(x = year, y = count, fill = wash_related)) + 
  theme_classic() +
  theme(legend.position = "bottom") + # move legend to the bottom of graph 
  theme(
    # Rotate the x-axis text labels by 45 degrees
    axis.text.x = element_text(angle = 45, hjust = 1) 
    # hjust = 1 aligns the right edge of the text with the tick mark
  ) +
  labs(
    title = "Outbreaks over time, split by whether WASH-related",
    subtitle = "As reported in WHO's Disease Outbreak News (DONs)",
    x = "Year",
    y = "Number of outbreaks",
    fill = "WASH-related" # Renaming the legend title here
  )

Figure 1: Outbreaks over time, split by whether WASH-related

Figure 2 shows the number of WASH-related outbreaks over time, split by WHO region.

Annual number of WASH-related outbreaks reported were largely between 0 and 30.
Very few WASH-related outbreaks were reported in DONs in 2020 (and only in Eastern Mediterranean) which was likely due to COVID-19 pandemic.
Very few WASH-related outbreaks were reported in 2024, and only from the South-East Asia Region.

Code

# look at how many WASH outbreaks there were over time, split by country region 
processed_data |> 
  filter(wash_related == TRUE) |>  # keep only WASH related outbreaks
  group_by(year, who_region) |> 
  summarise(count = n()) |> 
  ungroup() |> 
  filter(!is.na(who_region)) |> 
  
  ggplot() + 
  geom_col(aes(x = year, y = count, fill = who_region)) + 
  theme_classic() +
  theme(legend.position = "bottom") + # move legend to the bottom of graph 
  theme(
    # Rotate the x-axis text labels by 45 degrees
    axis.text.x = element_text(angle = 45, hjust = 1) 
    # hjust = 1 aligns the right edge of the text with the tick mark
  ) +
  labs(
    #title = "Outbreaks over time, split by WHO region",
    #subtitle = "As reported in WHO's Disease Outbreak News (DONs)",
    x = "Year",
    y = "Number of outbreaks",
    fill = "WHO " # Renaming the legend title here
  )

Figure 2: WASH-related outbreaks over time, split by WHO region

Figure 3 shows the number of WASH-related outbreaks by disease and WHO region.

WASH-related outbreaks disproportionately impacted the African Region.
Cholera was a large contributor to WASH-related outbreaks across all regions
Cholera especially impacted the African region.

Code

# look at what the most common WASH related outbreaks were - by region
processed_data %>%
  filter(wash_related == TRUE) |> 
  group_by(who_region, disease) |> 
  summarise(count = n(), .groups = 'drop') |> 
  filter(!is.na(who_region)) |> 
  
  ggplot() +
  geom_col(aes(x = who_region, y = count, fill = disease)) +
  
  scale_fill_discrete(
    # Apply the wrapping function directly to the labels of the scale
    labels = scales::label_wrap(30) 
  ) +
  
  theme_classic() +
  labs(
    #title = "WASH-related outbreaks*, split by WHO region",
    #subtitle = "*As reported in WHO's Disease Outbreak News (DONs)",
    x = "WHO Region",
    y = "Number of outbreaks",
    fill = "Disease" # Renaming the legend title here
  ) +
  
  guides(
    fill = guide_legend(
      ncol = 3,  
      byrow = TRUE,
      title.position = "top",
      direction = "horizontal"
    )
  ) +
  
  theme(
    legend.position = "bottom", # move legend to the bottom of graph 
    legend.text = element_text(size = 9),
  ) +
  
  scale_x_discrete(
    # Wrap text after 10 characters
    labels = scales::label_wrap(20) 
  )

Figure 3: WASH-related outbreaks split by disease and WHO region

3.2 Table summary

Table 1 provides a summary of the top 10 most common diseases responsible for the outbreaks reported in this dataset between 1996 and 2025 by WHO region, in order of number of outbreaks reported.

The single biggest contributor to the WASH-related outbreaks in this dataset is cholera, which disproportionately affected the African Region.
Supplementary analysis (not shown) found that the five most common countries to report cholera outbreaks in the African region were: Democratic Republic of the Congo (n=13), Mozambique (n=9), Burundi (n=8), Nigeria (n=8), and United Republic of Tanzania (n=8).

Code

processed_data |> 
  filter(wash_related == TRUE) |> 
  group_by(who_region, disease) |> 
  summarise(count = n(), .groups = 'drop') |> 
  filter(!is.na(who_region)) |> 
  arrange(desc(count)) |> 
  head(10) |> 
  gt()

Table 1: WASH-related outbreaks split by disease and WHO region

who_region	disease	count
African Region	Cholera	155
African Region	Acute poliomyelitis, unspecified	82
Eastern Mediterranean Region	Cholera	40
Region of the Americas	Cholera	23
Eastern Mediterranean Region	Acute poliomyelitis, unspecified	18
European Region	Acute hepatitis A	15
South-East Asia Region	Cholera	12
Western Pacific Region	Cholera	12
European Region	Infections due to other Salmonella	11
European Region	Acute poliomyelitis, unspecified	9

4.0 Conclusions

WASH-related outbreaks only make up a relatively small proportion of all global outbreaks.
WASH-related outbreaks disproportionately affect the African Region, largely due to cholera outbreaks.
A major limitation of these data is that this dataset only consists of outbreaks reported in the World Health Organization (WHO)’s Disease Outbreak News (DONs) which is known to underrepresent all global outbreaks.

5.0 References

Disease Outbreaks Data Project. 2024. “Disease Outbreaks Data Project.” Humanitarian Data Exchange (HDX). https://data.humdata.org/organization/disease-outbreaks-data-project.

R Core Team. 2019. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org.

Torres Munguía, M., F. Badarau, M. Díaz Pavez, I. Martínez-Zarzoso, and F. Wacker. 2022. “A Global Dataset of Pandemic- and Epidemic-Prone Disease Outbreaks.” Sci Data 9: 683. https://doi.org/10.1038/s41597-022-01797-2.