redcap_user_lifecycle.Rmd

---
title: "REDCap User Lifecycle"
date: "`r Sys.Date()`"
output: pdf_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, include = TRUE, message = FALSE)
```

The report is an example of how you could charcaterize your REDCap users using data from the redcap_user_information. The data is easy to access using a few R packages that manage configuration data (dotenv), database connections (DBI and redcapcustodian), dates (lubridate), and data transformation (tidyverse)

```{r get_the_data, echo = TRUE}
library(dotenv)
library(DBI)
library(redcapcustodian)
library(lubridate)
library(tidyverse)

rc_conn <- connect_to_redcap_db()

# get the data
redcap_user_information <- tbl(rc_conn, "redcap_user_information")
```

We need to select the columns we care about and transform them to meet our needs:

```{r transform, echo = TRUE}
user_life_cycle <- redcap_user_information %>%
  select(username, user_creation, user_lastlogin, user_suspended_time) %>%
  collect() %>%
  filter(!username %in% c("site_admin", "master")) %>% # these accounts are junk
  mutate(
    lifespan = if_else(is.na(user_suspended_time),
                       today() - as.Date(user_creation),
                       user_suspended_time - user_creation),
    dormancy = today() - as.Date(user_lastlogin),
    creation = as.Date(user_creation),
    lastlogin = as.Date(user_lastlogin),
    suspension = as.Date(user_suspended_time)
  ) %>%
  select(
    lifespan,
    dormancy,
    creation,
    lastlogin,
    suspension
  )
```


```{r dormancy, echo = FALSE, warning = FALSE, fig.cap="This plot shows how long user accounts have remained dormant with no login activity. With the binwidth set to 30 days, each bar represents about one month."}
ggplot(user_life_cycle, aes(dormancy)) +
  geom_histogram(binwidth = 30)
```

 

```{r lifespan, echo = FALSE, warning = FALSE, fig.cap="This plot shows the lifespan of the redcap accounts on the system where lifespan is expressed in days."}
ggplot(user_life_cycle, aes(lifespan)) +
  geom_histogram(binwidth = 30)
```

\newpage

For the last plot we want to see three events in a REDCap account's lifecycle over time. To do that, we need to pivot the data, combining the dates of those events into a single column `event_date`.  We use the column names to populate the `event` column which will serve as a grouping variable and label for the dates in `event_date`.

```{r pivot_data, echo = TRUE}

lifecycle <- user_life_cycle %>%
  pivot_longer(
    cols = c("creation",
             "lastlogin",
             "suspension"
             ),
    names_to = "event",
    values_to = "event_date"
  )
```

Now the data can be easily plotted. 

```{r plot_events, echo = FALSE, warning = FALSE, fig.cap="The first graph shows account creation since system creation. The last login graph tells a similar story as the dormancy graph. The suspension graph shows when users were suspended if your system implements."}
ggplot(lifecycle, aes(event_date)) +
  geom_histogram(binwidth = 30) +
  facet_grid(event ~ .)
```