Screening for Disparities

This package is used to screen for disparities in accordance with the recommendations by the QIC Ethics Subcommittee.

Setup

To install the package, run the following:

devtools::install_github("maduc/demographics", host = "github.research.chop.edu/api/v3")

How does this package work?

screen_demos()

The main function in this package is screen_demos(), which copies a SQL query that will add demographic data to your cohort and generate simple bar plots for an outcome metric stratified by each demographic field.

The demographics included for stratification are:

• Race (Non-Hispanic White, Non-Hispanic Black, Hispanic or Latino, Other)
• Primary Language (English, Non-English)
• Payer Type (Commercial, Government, Other)

The syntax of the function is as follows:

demographics::screen_demos(table = "TABLE_NAME", metric = "METRIC_FIELD",
                           qmr_con = CON_NAME, datamart = T/F)

Notes about the above:

• table can be either a datamart or an R dataframe. Whichever you choose to use, it must contain PAT_KEY, VISIT_KEY. If you are not using a datamart, only retain your primary key and the fields necessary to run the function. Doing otherwise may generate an error.
• metric must be a numeric field in your datamart or dataframe.
• qmr_con is the name of your QMR_DEV connection object.
• datamart refers whether or not table is a datamart. Defaults to TRUE.

get_demo_data()

get_demo_data() is a helper function to screen_demos(). It returns the dataset with demographic information and copies the SQL query used to generate the dataset to the clipboard. You can paste the query into Aginity after running get_demo_data() or screen_demos(). Use this function if you want to do further manipulations with your dataset. All of the arguments match screen_demos, however the datamart argument does not default to TRUE and must be specified.

demographics::get_demo_data(table = "TABLE_NAME", metric = "METRIC_FIELD",
                           qmr_con = CON_NAME, datamart = T/F)

If you have any questions or encounter any issues, contact Nonye Madu or Paul Wildenhain.