README.md

In mustafaascha/nsqipr: Using ACS-NSQIP in R (unofficially)

ACS-NSQIP analysis for a study of the "July Effect"

Purpose

This package is meant to document and increase the reproducibility of a paper examining the "July Effect" among procedures recorded in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP).

Instructions

After extracting the ACS-NSQIP data, place the .txt data files into the project folder "data/txt". By default, the report will use the first six files in the "txt" data folder. This is because the "July Effect" analysis requires resident PGY data, which was available for only a select number of years.

This work depends on R, pandoc, and several R packages: tidyverse, pander, knitr, rmarkdown, mice, devtools, zeallot, Epi, and broom. R users who are familiar with packrat may take advantage of the included dependency source code.

Assuming the years in the ACS-NSQIP "data/txt/" folder correspond to those reported in our manuscript, "report.Rmd" may be knit to reproduce results from manuscript. Alternatively, Mac or Linux users may navigate to the working directory and reproduce these results using make.

Project structure

Before adding the data to the correct folder, the project structure looks like this:

nsqipr ├── data/txt │   └── PUT-DATA-HERE ├── data-raw │   └── paper_things.R ... ├── README.md └── report.Rmd

After adding the data, the project directory should have (about) the following structure:

nsqipr ├── data/txt │   ├── ACS_NSQIP_PUF_05_06_vr1.txt.gz │   ├── ACS_NSQIP_PUF07_TXT.txt.gz │   ├── ACS_NSQIP_PUF08_TXT.txt.gz │   ├── ACS_NSQIP_PUF09_TXT.txt.gz │   ├── ACS_NSQIP_PUF10_TXT.txt.gz │   ├── ACS_NSQIP_PUF11_TXT.txt.gz │   ├── acs_nsqip_puf12.txt.gz │   ├── acs_nsqip_puf13.txt.gz │   ├── acs_nsqip_puf14.txt.gz │   ├── acs_nsqip_puf15_v2.txt.gz │   └── PUT-DATA-HERE ├── data-raw │   └── paper_things.R ... ├── README.md └── report.Rmd

mustafaascha/nsqipr documentation built on May 15, 2019, 3:33 p.m.