r4epi / surveys

Survey templates (WIP)

Sitrep

The goal of {sitrep} is provide report templates for common epidemiological surveys and outbreak reports. The package further contains helper function that standardize certain analyses.

While templates are primarily for MSF analyses - they have been setup to be as generic as possible for use by the general applied epidemiology community.

Detailed information about the project and the templates can be found at https://r4epis.netlify.com.
A reference website for the functions in {sitrep} can be found at https://r4epi.github.io/sitrep.

{sitrep} includes a number of other R packages which facilitate specific analysis:
{epitabulate}: Tables for epidemiological analysis
{epidict}: Epidemiology data dictionaries and random data generators
{epikit}: Miscellaneous helper tools for epidemiologists
{apyramid}: Age pyramid construction and plotting

Installation

The {sitrep} package, is currently stored in a GitHub repository. Therefore, the procedure to install these packages have one extra step required.

To install sitrep from GitHub you must first install the remotes package.

# install.packages("remotes")
remotes::install_github("r4epi/sitrep")

If you are getting errors, check the frequently asked questions.

Available templates

Sitrep has four outbreak templates and four survey templates available. These templates will generate the following:

1. A word document with the situation report
2. A plain text markdown document (for conversion to other formats such as HTML or PDF)
3. A directory with all of the figures produced

You can access the list of templates in R Studio by clicking (see example below): file > New file > R Markdown… > From Template

You can generate an example template by using the check_sitrep_templates() function:

library("sitrep")
output_dir <- file.path(tempdir(), "sitrep_example")
dir.create(output_dir)

# view the available templates, categorized by type
available_sitrep_templates(categorise = TRUE)
#> $outbreak
#> [1] "ajs_outbreak"        "cholera_outbreak"    "measles_outbreak"   
#> [4] "meningitis_outbreak"
#> 
#> $survey
#> [1] "mortality"         "nutrition"         "vaccination_long" 
#> [4] "vaccination_short"

# generate the measles outbreak template in the output directory
check_sitrep_templates("measles_outbreak", path = output_dir)
#> [1] "C:\\Users\\alexf\\AppData\\Local\\Temp\\Rtmpcv1H8d/sitrep_example"

# view the contents
list.files(output_dir, recursive = TRUE)
#> [1] "measles_outbreak.Rmd"

Please note that the ‘sitrep’ project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

surveys's Issues

Code for maps

It seems maps are useful in all surveys, however not always used. I suppose some general code snippets for map generation might be useful and should be part of the report templates.

Any thoughts?

Generating an interview schedule

Is is something we want to do? Together with sample size calculation and the sampling in R, maybe we can also generate other supporting material for operations? Just an idea

Population for surveys

Where do the population estimates for clusters / whoe area come from? Is this manually extracted from different source or are there external data sources you would often use?

Vaccine coverage sample size calculation

I see one study using the OpenEpiMenu software.

• Checkout the method of the software and see if/how it can be replicated in R

Nutrition survey sample size calculation

As in #1 the Emergency Nutrition Assessment (ENA) for Standardized Monitoring of Relief and Transition (SMART) software is used.

• Replicate method in R
• Check if survey package fits in here
• check other materials regarding sample size calculation from WHO

CFR estimates

For the CFR estimates once could use simple binomial proportion CIs or maybe use the survey package to get better CIs for the different strata/cluster. Any input here is welcome.

I also see negative CIs which suggets these are not binomial proportions (should be between 0 and 1).

Mortality sample size calculation

Sample size calculation should be done in R.
Task:

• figure out the method used in ENA for cluster sampling (http://www.nutrisurvey.de/ena/ena.html)).
• Investigate what the survey package can do for us.

Random sampling seems to just be random number generation.