README.md

In npeters1322/hospEpi: Perform Basic Hospital Epidemiology Tasks

hospEpi

The goal of hospEpi is to provide useful functions for epidemiological analyses in a hospital setting. Functions so far deal with summarizing and plotting patient location history data and disease-exposure (or risk factor) data.

Installation

You can install the development version of hospEpi like so:

devtools::install_github("npeters1322/hospEpi", build_vignettes = TRUE)

Using build_vignettes = TRUE makes sure the package vignette is built when you install the package, which might be useful if you want to explore the vignette.

Example

After loading the package, say you have some disease-exposure data you want to plot and/or summarize. An example dataset comes with the package and gives a simple idea of what your data should look like (at least for the most part) before using any functions.

#use the example disease-exposure dataset from the package
de_data <- disease_expose_data
head(de_data)
#>   id disease exposure1 exposure2        exposure3
#> 1  1      No         0        No            child
#> 2  2      No         0       Yes    office worker
#> 3  3      No         1        No      shop worker
#> 4  4     Yes         0        No            child
#> 5  5      No         1       Yes education worker
#> 6  6     Yes         1        No    office worker

To work with my plot and summary functions, you’ll want a dataset that only contains 0s and 1s in each variable, so you should clean your data using another function in this package if your data is not like that.

#clean the disease-exposure data
cleaned_de_data <- clean_disease_expose(data = de_data, disease = "disease", noDisease = "No", exposures = c(3,4,5))
head(cleaned_de_data)
#>   disease_Yes exposure1 exposure2_No exposure2_Yes exposure3_child
#> 1           0         0            1             0               1
#> 2           0         0            0             1               0
#> 3           0         1            1             0               0
#> 4           1         0            1             0               1
#> 5           0         1            0             1               0
#> 6           1         1            1             0               0
#>   exposure3_education worker exposure3_office worker exposure3_shop worker
#> 1                          0                       0                     0
#> 2                          0                       1                     0
#> 3                          0                       0                     1
#> 4                          0                       0                     0
#> 5                          1                       0                     0
#> 6                          0                       1                     0

As you can see above, all columns are now binary. If you look back at the example data before it was cleaned, you’ll see that for the first person, exposure3 was “child.” Now, in the cleaned dataset, there is a binary column for exposure3_child, and the first row is marked as 1 because that person was marked as a child in the original dataset.

With cleaned data, we can now make an object of class disease_expose. There are specific summary and plot functions in my package to work with objects of that class, so if you want to use those, you must create an object like so:

de_object <- disease_expose(cleaned_de_data)
class(de_object)

#> [1] "disease_expose" "data.frame"

Calling the disease_expose function will pull up a Shiny gadget, which will allow you to choose your disease column and any exposure columns you want to bring along from your data, and then it will create an object of class disease_expose if you receive no errors. If a Shiny gadget is not your thing, you can also manually do everything (choose your disease column and exposure columns by typing them into the function) by using the new_disease_expose function.

Now that you have a disease_expose object, you can plot it by simply calling the plot function with the necessary arguments. The only required argument is the disease_expose object, but you can also add additional arguments that will be passed to the geom_bar function. For example:

plot(de_object, position = 'dodge')

Now, you can see grouped bar charts of your disease-exposure combinations to see if there is anything interesting in the data.

You can also summarize your data like so:

de_summ <- summary(de_object)
de_summ
#>                            Incidence in Exposed Incidence in Unexposed
#> exposure1                             0.4701195              0.5140562
#> exposure2_No                          0.5221239              0.4671533
#> exposure2_Yes                         0.4671533              0.5221239
#> exposure3_child                       0.5333333              0.4789474
#> exposure3_education worker            0.4887218              0.4931880
#> exposure3_office worker               0.5344828              0.4791667
#> exposure3_shop worker                 0.4198473              0.5176152
#>                            Risk Ratio (RR) RR Lower 95% CI RR Upper 95% CI
#> exposure1                        0.9145294            0.77            1.09
#> exposure2_No                     1.1176715            0.94            1.33
#> exposure2_Yes                    0.8947173            0.75            1.07
#> exposure3_child                  1.1135531            0.91            1.36
#> exposure3_education worker       0.9909442            0.81            1.21
#> exposure3_office worker          1.1154423            0.91            1.36
#> exposure3_shop worker            0.8111187            0.65            1.01
#>                            Odds Ratio (OR) OR Lower 95% CI OR Upper 95% CI
#> exposure1                        0.8386983            0.59            1.19
#> exposure2_No                     1.2462384            0.88            1.77
#> exposure2_Yes                    0.8024147            0.56            1.14
#> exposure3_child                  1.2433281            0.82            1.88
#> exposure3_education worker       0.9822879            0.66            1.46
#> exposure3_office worker          1.2479871            0.82            1.89
#> exposure3_shop worker            0.6744282            0.45            1.01
#>                            Fisher P-Value Chi-Square P-Value
#> exposure1                      0.37104591         0.37179421
#> exposure2_No                   0.24293634         0.25688248
#> exposure2_Yes                  0.24293634         0.25688248
#> exposure3_child                0.34595536         0.35021941
#> exposure3_education worker     1.00000000         1.00000000
#> exposure3_office worker        0.34033555         0.34804377
#> exposure3_shop worker          0.06689589         0.06858653

As you can see, the function outputs several different statistics for each disease-exposure combination and saves them in a data.frame, which would allow for easy extraction if wanted.

There are also multiple functions for patient location history data, and you can find out how to use those by reading the package vignette or the help documentation for the functions.

npeters1322/hospEpi documentation built on April 30, 2022, 6:12 p.m.