Skip to content
Branch: master
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.

Files

Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
general_code
inequality_over_time_app
lorenz_curve_app
ntile_relative_app
prescribing_over_time_app
top_centile_app
README.Rmd
README.md

README.md

Introduction

This is a list of companion applications for our paper, Opioid prescribing patterns among medical providers in the United States, 2003-17: retrospective, observational study. The full citation for this paper is:

Kiang MV, Humphreys K, Cullen MR, and Basu S. Opioid prescribing patterns among medical providers in the United States, 2003-17: retrospective, observational study. BMJ 2020;368:l6968. doi: 10.1136/bmj.l6968

For interested readers, we have created several interactive results viewers. These online applications allow readers to change the parameters of plots and tables in order to view results for different provider specialties, states, years, measures, drugs, or sensitivity analyses. A brief description of each app is provided below.

Update 1/11/20: Due to limitations on our data use agreement, we have removed tables from the public Shiny apps The tables are still generated in the backend but are not displayed in the frontend. If you are granted access to the restricted access data (through the process outlined in the main README) you can display these tables in the Shiny applications by uncommenting out the table section in each app’s ui.R file (usually after the plotting functions).

The code for both our paper and these apps is available online in the official paper respository.

Interactive results viewers

  1. Prescribing patterns within a single year: This application is an interactive version of Figure 1 in our paper. For any specified year, it shows the cumulative proportion of drugs dispensed as a function of the cumulative proportion of ranked providers. That is, what proportion of the drug is prescribed by the top proportion of providers. It allows users to specify different sets of drugs, measures, states or provider specialty, and level of trimming (i.e., sensitivity analyses). Importantly, due to limits of the shinyapps.io server, this app only runs a subset of the drugs in our paper. To run it for all drugs, you must run the app locally.

  1. Prescribing patterns over time: This application is an interactive version of Figure 2 in our paper. It shows the proportion of different drugs prescribed by the top 1%, 5%, 10%, 25%, or 50% of providers from 2003 to 2017. It allows users to specify different sets of drugs, measures, states or provider specialty, and level of trimming (i.e., sensitivity analyses).

  1. Closer look at the top centile: This application allows for looking at the provider specialties in the top centile of prescribers and across all providers at different levels of trimming and different drug types (Fig. S5). In addition, it shows the centile to centile transactions of prescribing patterns for different years, drugs, and levels of trimming (Fig. 3).

  1. Measures of global inequality in prescribing: This application is an interactive version of the Gini analysis in our paper (Supplemental Materials). We measured the disproportionate level of prescribing using several metrics of inequality. In our paper, we present the Gini coefficient and show that all measures of inequality are highly correlated. This app users to specify different sets of drugs, inequality metrics, measures, states or provider specialty, and level of trimming (i.e., sensitivity analyses).

  1. Prescribing of the top providers relative to the median providers: This application is an interactive version of Supplemental Materials Fig. S6 in our paper. As described in the paper, we divided providers into centile groups (i.e., 100 equally-sized groups) and compared the prescribing patterns of the top 25 centile groups relative to the median (50th) centile group. It allows users to specify different sets of drugs, measures, states or provider specialty, and level of trimming (i.e., sensitivity analyses).

You can’t perform that action at this time.