The effectiveness of population-wide screening in reducing SARS-CoV-2 infection prevalence in Slovakia

This repository contains the data and code for our manuscript:

Pavelka S, Van-Zandvoort K, Abbott S, Sherratt K, Majdan M, CMMID COVID-19 working group, Jarčuška P, Krajčí M, Flasche S*, Funk S* (*: equal contribution), The effectiveness of population-wide screening in reducing SARS-CoV-2 infection prevalence in Slovakia. Available at https://cmmid.github.io/topics/covid19/Slovakia.html.

How to download or install

You can download the compendium as a zip from from this URL: https://github.com/sbfnk/covid19.slovakia.mass.testing/archive/master.zip.

Or you can install this compendium as an R package, covid19.slovakia.mass.testing, from GitHub with:

# install.packages("devtools")
remotes::install_github("sbfnk/covid19.slovakia.mass.testing")

Included data sets

The repository contains three data sets:

The testing data set ms.tst can be loaded with

data(ms.tst)

Incidence of cases confirmed by PCR per county PCR.inc can be accessed with

data(PCR.inc)

The Rt.county data set contains the estimated median reproduction number in each county on 22 October 2020.

data(Rt.county)

This data set can be re-created using the The EpiNow2 R package by running (noting that it can take a long time to run depending on the hardware available).

source(here::here("data-raw", "scripts", "rt.r"))
source(here::here("data-raw", "scripts", "convert_data.r"))

The EpiNow2 R package that is used to estimate the reproduction numbers uses generation times and delay distributions saved in data-raw/data. They can be re-generated by running.

source(here::here("data-raw", "scripts", "rt-distributions.r"))

The Google mobility data set for Slovakia mob.slo visualised in Supplementary Figure S4 can be accessed with

data(mob.slo)

Figures and tables

To regenerate Table 1, run

county_table(here::here("figures", "table1.pdf"))

To regenerate Fig. 1, run

p <- pcr_incidence()
ggsave(here::here("figures", "fig1.pdf"), p, width = 7, height = 3)

To regenerate Fig. 2, run

rr <- risk_ratios()

ggsave(here::here("figures", "fig2a.pdf"), rr$figures$a,
       width = 7.5, height = 7, device = cairo_pdf)
ggsave(here::here("figures", "fig2b.pdf"), rr$figures$b,
       width = 7, height = 4, device = cairo_pdf)
ggsave(here::here("figures", "fig2c.pdf"), rr$figures$c,
       width = 7, height = 4, device = cairo_pdf)
rr$tables

To generate Table S1 and estimate the adjusted prevalence ratio, run

r <- regression("tableS1.pdf")

To regenerate Fig. S1 and estimate minimum specificity, run

spec <- estimate_min_specificity()
ggsave(here::here("figures", "figS1.png"), spec$figure,
       width = 3, height = 4, dpi = 600)
spec$estimate
#> [1] 0.9984521

To regenerate Fig. S4, run

p <- mobility()
ggsave(here::here("figures", "figS4.png"), p, width = 10, height = 6.5,
       dpi = 600)

To regenerate Fig. S6, run

p <- bed_occupancy()
ggsave(here::here("figures", "figS6.png"), p, width = 4, height = 3,
       dpi = 600)

To regenerate Fig. S7, run

p <- prevalence()
ggsave(here::here("figures", "figS7.png"), p, width = 6, height = 10,
       dpi = 600)