Published in:
Open Access
01-12-2023 | Research
Protocol for developing a dashboard for interactive cohort analysis of oral health-related data
Authors:
Vinay Sharma, Oscar Cassetti, Lewis Winning, Michael O’Sullivan, Michael Crowe
Published in:
BMC Oral Health
|
Issue 1/2023
Login to get access
Abstract
Introduction
A working knowledge of data analytics is becoming increasingly important in the digital health era. Interactive dashboards are a useful, accessible format for presenting and disseminating health-related information to a wide audience. However, many oral health researchers receive minimal data visualisation and programming skills.
Objectives
The objective of this protocols paper is to demonstrate the development of an analytical, interactive dashboard, using oral health-related data from multiple national cohort surveys.
Methods
The flexdashboard
package was used within the R Studio framework to create the structure-elements of the dashboard and interactivity was added with the Shiny
package. Data sources derived from the national longitudinal study of children in Ireland and the national children’s food survey. Variables for input were selected based on their known associations with oral health. The data were aggregated using tidyverse packages such as dplyr
and summarised using ggplot2
and kableExtra
with specific functions created to generate bar-plots and tables.
Results
The dashboard layout is structured by the YAML (YAML Ain’t Markup Language) metadata in the R Markdown document and the syntax from Flexdashboard. Survey type, wave of survey and variable selector were set as filter options. Shiny’s render functions were used to change input to automatically render code and update output. The deployed dashboard is openly accessible at
https://dduh.shinyapps.io/dduh/. Examples of how to interact with the dashboard for selected oral health variables are illustrated.
Conclusion
Visualisation of national child cohort data in an interactive dashboard allows viewers to dynamically explore oral health data without requiring multiple plots and tables and sharing of extensive documentation. Dashboard development requires minimal non-standard R coding and can be quickly created with open-source software.