VIP Data Visualisation Shiny App

A comprehensive R Shiny application for exploring systematically reviewed evidence on COVID‑19, influenza, and RSV vaccine effectiveness and safety.

Table of Contents

• VIP Data Visualisation Shiny App
  • Table of Contents
  • Key Features
  • Repository Structure
  • Data Sources
  • Local Installation and Running
    • Prerequisites
    • Clone the repository
    • Run the app in RStudio
  • Required R Packages
  • Using the App
    • Studies Tab
    • Outcomes Tab
    • Lightweight View
    • Download Buttons
  • Deployment
    • ShinyApps.io
  • Embedding on a Website
  • Configuration
    • Population mapping
    • Table column order and hiding
  • Filtering Logic
  • Downloading Data
    • Main Downloads
    • Filtered Downloads
  • Customising the App
    • Changing filters
    • Changing table columns
    • Changing styles
    • Updating logos
  • Acknowledgements

The Vaccine Integrity Project (VIP) interactive data tool summarises newly published research evaluating the effectiveness and safety of vaccines and immunisation products for:

• COVID‑19
• Influenza
• Respiratory syncytial virus (RSV)

The tool is built in R Shiny and provides two main interactive tabs:

• Studies - a list of included studies with reference information, study characteristics, and risk‑of‑bias assessments.
• Outcomes - vaccine effectiveness/efficacy and safety estimates extracted from the included studies.

The underlying data come from systematic reviews of peer‑reviewed research conducted as part of the VIP project.

Key Features

• Interactive filtering by:
  • Virus
  • Population
  • Age group
  • Study period
  • Study design
  • Outcome domain
  • Risk of bias
  • Vaccine and strain
• Responsive layout with collapsible sidebar filters.
• Freeze header row and first column in data tables.
• Responsive/mobile‑friendly table behaviour.
• Lightweight / full‑view toggles.
• Download full source Excel workbooks.
• Download filtered views as Excel files.
• About page with project information and collapsible definitions.
• Dynamic row counts and filter helper messages.
• Support for embedded deployment via iframe.

Repository Structure

├── app.R # Main Shiny application
├── shiny_aux/
│ ├── config.R # Study configuration (columns, filters, population map)
│ ├── outcome_config.R # Outcome configuration
│ ├── filter_ui.R # Filter UI components
│ ├── filter_server.R # Filter server logic
│ ├── helpers.R # Helper functions
│ └── styles.R # Custom CSS
├── characteristics_tables/
│ └── All_Study_Characteristics.xlsx
├── outcome_tables/
│ └── All_Tables_split.xlsx
├── www/
│ ├── VIP_Logo_Horizontal.png
│ └── VIP_Logo_Vertical.png
└── README.md

Data Sources

The Shiny app reads from two primary Excel workbooks:

• characteristics_tables/All_Study_Characteristics.xlsx
  • Sheet Study Characteristics
  • Additional sheets: Footnotes, Outcomes_Long, N_Long, Population_Long, Virus_Long, RoB_Long
• outcome_tables/All_Tables_split.xlsx
  • Sheet All
  • Additional sheets: Footnotes, Population_Long, Virus_Long

These workbooks are generated from the consensus data‑extraction file using data processing scripts (available upon request).

Local Installation and Running

Prerequisites

• R (>= 4.0 recommended)
• RStudio (optional but recommended)
• Internet access for package installation

Clone the repository

git clone https://github.com/your‑org/vip-data-vis.git
cd vip-data-vis

Run the app in RStudio

Open app.R. Click Run App.

Required R Packages

The app depends on:

library(shiny)
library(readxl)
library(DT)
library(shinyjs)
library(dplyr)
library(writexl)
library(ggplot2)
library(rlang)
library(bslib)

Install with:

install.packages(c(
  "shiny", "readxl", "DT", "shinyjs",
  "dplyr", "writexl", "ggplot2",
  "rlang", "bslib"
))

Using the App

Studies Tab

Displays one row per included study with columns such as:

• Study
• Virus
• Population
• Age range
• Study period
• Total N
• Study design
• Risk of bias
• Funding source
• Journal / PMID / PMCID / DOI / Link

Filters available:

• Virus
• Type of Outcome
• Age Group
• Population Type
• Study Period
• Study Design
• Risk of Bias
• Advanced population filters

Outcomes Tab

Displays extracted outcome estimates with columns including:

• Study label
• Population
• Age range
• Outcome
• Estimate
• Estimate type
• Risk of bias
• Vaccine formulations and strains
• Sample sizes

Filters available:

• Virus
• Type of Outcome
• Age Group
• Population Type
• Study Period
• Study Design
• Type of Estimate
• Risk of Bias

Lightweight View

Both tabs include a lightweight view toggle that hides less critical columns to reduce horizontal scrolling.

Download Buttons

• Download full Excel - exports the original source workbook.
• Download filtered view - exports only the currently visible rows, including a Footnotes sheet.

Deployment

ShinyApps.io

1. Create an account at shinyapps.io.
2. Install rsconnect:

   install.packages("rsconnect")

3. Authenticate:

   rsconnect::setAccountInfo(
    name = "your‑account",
    token = "your‑token",
    secret = "your‑secret"
    )

4. Deploy:

   rsconnect::deployApp(appDir = "path/to/app")

Embedding on a Website

Use an iframe:

<iframe
  src="https://your-account.shinyapps.io/vip_data_vis/"
  width="100%"
  height="700"
  style="border:none;"
  allowfullscreen
>
</iframe>

Configuration

Population mapping

shiny_aux/config.R contains:

pop_code_to_full <- c(
  "OA" = "Older Adults",
  "A"  = "Adults",
  "I"  = "Infants",
  "C"  = "Children",
  "IC" = "Immunocompromised",
  "HR" = "Other co-occurring conditions",
  "P"  = "Pregnant",
  "H"  = "Healthcare personnel"
)

Table column order and hiding

shiny_aux/outcome_config.R and config.R define:

• always_hide - columns never displayed in the main table.
• article_cols - columns hidden in lightweight view.
• column_order - display order.
• desired_widths - column width preferences.

Filtering Logic

Each filter uses long‑format helper tables (Population_Long, Virus_Long, etc.) to map user selections to a set of row IDs. The final data set is the intersection of all active filter ID sets.

When no value is selected in a filter, the filter returns all IDs, so no rows are accidentally dropped.

Downloading Data

Main Downloads

• Download Study Characteristics - full source workbook, cleaned but unmodified columns.
• Download Outcomes - all outcome rows in the source workbook.

Filtered Downloads

Both tabs include Download filtered view, which exports only the currently visible rows.

Selected columns can be removed at export time via drop_columns.

Customising the App

Changing filters

Edit shiny_aux/filter_ui.R and shiny_aux/filter_server.R to add or modify filters.

Changing table columns

Update the relevant configuration files:

• config.R for Studies.
• outcome_config.R for Outcomes.

Changing styles

Edit shiny_aux/styles.R.

Updating logos

Replace the files in the www/ folder with your own images.

Acknowledgements

This work builds on the Vaccine Integrity Project (VIP) led by CIDRAP.
We thank all reviewers, extractors, and contributors who made this systematic review possible.