The CoVax Project

Source

• All used data sets were created in a non-English-speaking environment and contained special characters, therefore UTF-8 encoding is used for reading and saving the data sets.

• Reading and saving UTF-8 data sets changes the name of the first column in the data set. There did not seem to be a workaround and therefore changed name of the first column is therefore reflected in the code and data set previews.

Immunisation data

• Czech Covid-19 statistics are published and updated daily by the Ministry of Health [3] on its National Dashboard of Active Diseases. The dashboard provides statistics, visualisations, and access to a number of more or less aggregated data files. Data can be accessed through API as well as downloaded directly (.csv).

• Data collection, management, and publishing is managed by the Institute of Health Information and Statistics of the Czech Republic [4] through the National Health Information System (NHIS) and its subsystems, which pool data shared by individuals and institutions managing case tracking and healthcare (i.e., healthcare providers and regional public health offices).

• The vaccine uptake data set is updated daily (except for weekends) and available from the dashboard:
  https://onemocneni-aktualne.mzcr.cz/api/v2/covid-19/ockovani-orp.csv

Population data

• Summary population statistics are published by the Czech Statistical Office on its website [5] and updated annually. They are compiled based on observed births, deaths, and within-country mobility. Additionally, the data is checked when the outcomes of the most recent census become available.
• The data set is available on the website, however it is formatted for user-friendly viewing rather than data processing and was therefore pre-processed and saved in the project data folder.
• The original data set is accessible through the following catalogue:
  https://www.czso.cz/csu/czso/population-of-municipalities-1-january-2021
  or directly through the following download link:
  https://www.czso.cz/documents/10180/142756350/1300722101.xlsx

Variable description

The Codebook provides information about data sets used in the Data Management and Visualisation project. Please note that where variables were dropped and not used further, their Czech name is referenced. Variables that have been manipulated (and therefore translated) are labelled in English.

Immunisation data

• The data set is in the long format (i.e., several rows per observation to display unique combinations of all recorded variables).

Used variables

• region_name: The official names of the administrative regions in Czech (see Notes).

  • Uncategorised entries were dropped.

• region_id: The statistical codes of the administrative regions within the European NUTS hierarchy (see Notes).

  • Uncategorised entries were dropped.

• finished_vax: The number of individuals who completed their Covid-19 immunisation.

Unused variables

• id: A pre-assigned ID, possibly for internal identification purposes.

• date: The date when individual observations were reported.

• orp_bydliste_nazev: The official names of the administrative districts in Czech (see Notes).

  • Contains both labelled and unlabelled uncategorised data.

• orp_bydliste_kod: The statistical codes of the administrative districts within the Statistical Meta-information System (see Notes).

• Contains unlabelled uncategorised data.

• vekova_kategorie: The age group immunisation recipients fall within.

• ockovaci_latka_nazev: The name of the European Medicines Agency-approved vaccine that was administered.

• ockovaci_latka_kod: A code assigned to the approved vaccines, possibly for internal use of the Ministry.

• prvni_davka: The number of individuals falling within observed categories who received the first dose of a Covid-19 vaccine.

• druha_davka: A number of individuals falling within observed categories who received the second dose of a Covid-19 vaccine.

• posilujici_davka: A number of individuals falling within observed categories who received the booster dose of a Covid-19 vaccine.

Population data

• The data set is not in a machine readable format. It is separated into several groups for which the data are provided: the Czech Republic, Administrative regions (used in the visualisation), Administrative districts.

Used variables

• pop_total: The number of permanent residents per region

• region_id: The statistical codes of the administrative regions within the European NUTS hierarchy (see Notes).

Unused variables

• Population - Males: The number of male residents per category.

• Population - Females: The number of female residents per category.

• Average age - Total: Total average age per category.

• Average age - Males: Average male age per category.

• Average age - Females: Average female age per category.

Computed variables

• vax_per_pop: The proportion of individuals with completed Covid-19 immunisation per total region population; computed by dividing finished_vax (Immunisation data) by pop_total (Population data).

Notes

• Administrative districts of municipalities with extended competence: the smallest administrative districts in the Czech Republic, centered around 205 municipalities with extended competence [6].
• Administrative regions: A total of 14 administrative regions of the Czech republic with develoved powers, roughly equivalent to UK counties or groups of unitary authorities [7].
• NUTS: Eurostat’s Nomenclature of Territorial Units for Statistics used for comparison of devolved administration units across the EU as well as funding purposes [8].

# load packages
if(!require(here)) install.packages("here",
                                    repos = "http://cran.us.r-project.org")
if(!require(tidyverse)) install.packages("tidyverse",
                                         repos = "http://cran.us.r-project.org")
if(!require(write)) install.packages("write",
                                     repos = "http://cran.us.r-project.org")

# import data
url_vaccines <- "https://onemocneni-aktualne.mzcr.cz/api/v2/covid-19/ockovani-orp.csv"

df <- read.csv(url_vaccines,
               encoding = "UTF-8")

pop_df <- read.csv(here("data_raw", "2021-01-01_region_population.csv"),
                   encoding = "UTF-8")

# NA → 0 for subsequent computations
df[is.na(df)] <- 0

# drop redundant columns
df <- df %>% select(-c(
  X.U.FEFF.id,
  ockovaci_latka_kod,
  orp_bydliste_nazev,
  orp_bydliste_kod,
  vekova_kategorie,
  ockovaci_latka_nazev,
  prvni_davka,
  druha_davka,
  posilujici_davka
  )) %>% 

# translate and simplify column names
  rename(
    date = datum,
    region_name = kraj_bydliste_nazev,
    region_id = kraj_bydliste_kod,
    finished_vax = dokoncene_ockovani)

# sum daily figures into totals
df <- df %>% 
  group_by(region_name, region_id) %>% 
  summarise(finished_vax = sum(finished_vax))

# merge with population data
df <- merge(df, pop_df, by.x = "region_id", by.y = "X.U.FEFF.region_id") %>% 
  mutate(vax_per_pop = finished_vax / pop_total)

# save tidy data (in UTF-8)
readr::write_excel_csv(df, file = here("data_tidy", "cz_covax_regions_tidy.csv"))

For visualisation, the required packages are loaded as well as the data. New simple features (sf) objects which provide the geo-spatial coordinates for visualising the regions (kraje()) on the map of the Czech Republic (as well as the map itself, republika()) is supplied by the RCzechia package. The sf object was not introduced in the data preparation stages as combining sf objects with data frames is difficult and the resulting files are large. Redundant information is removed from the sf object:

# load packages 
if(!require(here)) install.packages("here",
                                    repos = "http://cran.us.r-project.org")
if(!require(tidyverse)) install.packages("tidyverse",
                                         repos = "http://cran.us.r-project.org")
if(!require(ggplot2)) install.packages("ggplot2",
                                       repos = "http://cran.us.r-project.org")
if(!require(RCzechia)) install.packages("RCzechia",
                                        repos = "http://cran.us.r-project.org")
if(!require(sf)) install.packages("sf",
                                  repos = "http://cran.us.r-project.org")

# load data
df <- read.csv(here("data_tidy","cz_covax_regions_tidy.csv"),
               encoding = "UTF-8")

# load sf objects
repub <- republika()

regions <- kraje(resolution = "high") %>% 
  select(-c(KOD_KRAJ, NAZ_CZNUTS3))

The tidy data set is then merged with the tidy sf object, and the new object is explicitly coerced into being an sf object (as not doing so has caused issues with the visualisation package recognising the sf data):

df <- merge(df, regions, by.x = "X.U.FEFF.region_id", by.y = "KOD_CZNUTS3")

# df → sf object for ggplot2
df <- st_as_sf(df)

The visualisation is initiated using the ggplot2 package; the df is specified as well as custom aesthetics. The visualisation is labelled and formatted (see below for notes on aesthetics):

# specify df for ggplot
ggmap <- ggplot(data = df) +
  
  # read regional geographic data
  geom_sf(aes(fill = vax_per_pop), colour = "black", lwd = 0.3) +
  
  # read geographic data of the country
  geom_sf(data = repub, color = "black", lwd = 0.6, fill = NA) +
  
  # set colour (suitable for visually impaired)
  scale_fill_viridis_c(labels = scales::comma, direction = -1) +
  
  # label plot elements
  labs(
    x = "Latitude",
    y = "Longitude",
    title = "Covid-19-immunised to total region population ratio in the Czech Republic",
       subtitle = "Ratio of fully immunised population (i.e., having received 2 doses of an approved Covid-19 vaccine) to total population per region",
       caption = "Source: IHIS CR & CZSO
    Note: Cases of immunisation of foregin nationals and individuals of unknown region residence excluded",
       fill = "Immunised-to-total region \nresidents ratio") +
  
  # suitable theme (reading coords)
  theme_bw() +
  
  # sub/title alignment
  theme(plot.title = element_text(hjust = 0),
        plot.subtitle = element_text(hjust = 0))

# view
ggmap