README.Rmd 2.27 KB
Newer Older
mabrook's avatar
mabrook committed
1 2 3
---
title: "Package: WIHS Data"
author: "Sara Levintow and Alan Brookhart"
levintow's avatar
levintow committed
4
date: "5/22/2017"
mabrook's avatar
mabrook committed
5 6 7
output: md_document
---

mabrook's avatar
mabrook committed
8
## WIHS Data Set #
mabrook's avatar
mabrook committed
9

mabrook's avatar
mabrook committed
10
The Women’s Interagency HIV Study (WIHS) is an ongoing observational cohort study with semiannual visits at 10 sites in the US. For some details see (1). This WIHS data was used by Lau et al in a 2009 paper on competing risks, see (2), and is available online. 
mabrook's avatar
mabrook committed
11 12 13

### Installation

mabrook's avatar
mabrook committed
14
You can install the package from our public Gitlab site
mabrook's avatar
mabrook committed
15

mabrook's avatar
mabrook committed
16
```{r setup, include=FALSE, warning=FALSE}
mabrook's avatar
mabrook committed
17 18
knitr::opts_chunk$set(echo = TRUE)
  library("devtools")
mabrook's avatar
mabrook committed
19
  library("tidyverse")
mabrook's avatar
mabrook committed
20
devtools::install_git("https://gitlab.novisci.com/nsPublicData/wihs2009.git")
mabrook's avatar
mabrook committed
21 22 23 24
  library(WIHS2009)
```


mabrook's avatar
mabrook committed
25 26 27 28 29 30 31
```{r setup2, eval=FALSE}
  library("devtools")
  library("tidyverse")
devtools::install_git("https://gitlab.novisci.com/nsPublicData/wihs2009.git")
  library(WIHS2009)
```

mabrook's avatar
mabrook committed
32 33
### Codebook ###

34
* `id` = patient identifier
mabrook's avatar
mabrook committed
35
* `idu` = injection drug use history 
mabrook's avatar
mabrook committed
36
* `white` = White Race or Non-white Race
37 38 39
* `age` = years of age on 12/6/1995
* `cd4` = CD4 count in cells/mm^3 measured at last visit before 12/6/1995
* `art` = time from 12/6/1995 to ART initiation (missing is censored)
mabrook's avatar
mabrook committed
40 41
* `study_end` = time from 12/6/1995 to end of study (10 years for everyone)
* `dropout` = time until patient is lost to follow-up (missing is censored)
42 43 44 45
* `aids_death` = time from 12/6/1995 to first of AIDS or death (missing is censored)
* `aids_death_art` = time from 12/6/1995 to first of AIDS, death, or ART initiation (missing is censored)
* `j` = indicator that `aids_death_art` is time until `art`

mabrook's avatar
mabrook committed
46 47 48 49 50
### Example data

```{r}
wihs
```
mabrook's avatar
mabrook committed
51 52 53 54

### Notes ###

* There is one record per patient.
mabrook's avatar
mabrook committed
55 56 57
* `idu` can be used as an exposure
* `cd4`, `age`, and `white` can be used as confounders
* The `art` variable can be used as a competing event time for time til death and AIDS
mabrook's avatar
mabrook committed
58
* Times here are measured in decimals (or Colombian time), such that 365.25 days is 1.000 years.
mabrook's avatar
mabrook committed
59 60 61 62 63 64
* The raw data (in ASCII form) and codebook are provided in the repository in `.\data-raw`.

### References

1. Barkan SE, Melnick SL, Preston-Martin S, et al. The Women's Interagency HIV Study. WIHS Collaborative Study Group. Epidemiology 1998;9(2):117-25.
2. Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am J Epidemiol 2009;170(2):244-56.