Building Cohorts with STROBE Diagrams

Introduction

The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement provides guidelines for reporting observational studies. A key component is the flow diagram that shows how the study population was selected. We encourage use of the companion strobe package, which provides tools to create STROBE-compliant flow diagrams while building your cohort.

This vignette demonstrates how to use the strobe functions via the standalone strobe package to:

1. Initialize a cohort with inclusion criteria
   
2. Apply sequential filters with exclusion tracking
   
3. Generate and customize STROBE flow diagrams
   
4. Review the complete filtering log

For more details, visit the strobe package website.

Installation

You can install the development version of the sRtr package from GitHub using devtools or pak:

# Install devtools if needed
install.packages("devtools")

# Install sRtr (and strobe, a suggested package)
devtools::install_github("VagishHemmige/sRtr", dependencies = TRUE)

Getting Started

library(sRtr)
library(dplyr)
library(strobe)

Basic STROBE Workflow

The STROBE workflow in strobe follows these steps:

1. Initialize the cohort with strobe_initialize()
   
2. Filter sequentially with strobe_filter()
   
3. Plot the diagram with plot_strobe_diagram()
   
4. Review the log with get_strobe_log()

Example: Kidney Transplant Cohort

Let’s build a cohort of kidney transplant recipients with specific criteria:

TX_KI %>%
  strobe::strobe_initialize(inclusion_label = "All kidney transplants") %>%
  strobe::strobe_filter(
    condition = "DON_AGE > 18",
    inclusion_label = "Donor over 18 years old",
    exclusion_reason = "Excluded: Donor age ≤ 18"
  ) %>%
  strobe::strobe_filter(
    condition = "DON_ABO == 'B'",
    inclusion_label = "Donor blood type B",
    exclusion_reason = "Excluded: Donor blood type is not B"
  ) %>%
  strobe::strobe_filter(
    condition = "REC_AGE >= 18 & REC_AGE <= 65",
    inclusion_label = "Recipient age 18-65",
    exclusion_reason = "Excluded: Recipient age outside 18-65 range"
  )

Reviewing the Filtering Log

# View the complete filtering log
strobe::get_strobe_log()

The log shows: - Each filtering step - Number of records included/excluded at each step - Cumulative counts - Exclusion reasons

Creating STROBE Diagrams

strobe::plot_strobe_diagram(
  incl_width_min = 3,
  incl_height = 1,
  excl_width_min = 2.5,
  excl_height = 1,
  lock_width_min = TRUE,
  lock_height = TRUE,
  incl_fontsize = 16,
  excl_fontsize = 14

)

For more details, see the separate documentation of the strobe package.