🏥 Project Overview

This repository contains a Shiny-based Flexdashboard designed to analyze Emergency Department (ED) primary diagnosis patterns among patients with Autosomal Dominant Polycystic Kidney Disease (ADPKD). By leveraging data from the MIMIC-IV and MIMIC-IV-ED databases, this tool identifies the most frequent reasons for emergency visits to help clinicians better understand the unique needs of this patient population.

📊 Dashboard Features

The dashboard is organized into two main sections:

1. Overview

• Introduction: Detailed background on ADPKD and the clinical importance of emergency care research for those patients.
• Dataset Glance: An interactive data explorer featuring:
  • A searchable and filterable preview of the applied dataset.
  • Key Value Boxes: Real-time counts of total patients, unique ED stays, and unique diagnoses.
  • Demographic Distribution: Interactive pie charts showing breakdowns by Gender, Race, Age Group, End Stage Renal Disease status, and Transplant status.

2. Plots

• General Leading Diagnosis: A customizable bar chart showing the "Top N" most frequent primary diagnoses.
• Stacked Bar Chart: Visualizes the distribution of leading diagnoses across different demographic segments.
• Slope Graph: A sophisticated rank-comparison tool that tracks how primary diagnosis priorities shift across binary variables (e.g., Male vs. Female or ESRD vs. Non-ESRD).

🌟 Impact & Importance

This dashboard identifies the most frequent clinical drivers for emergency visits in ADPKD patients, bridging a critical research gap in hereditary kidney disease management. By characterizing these presentation patterns, the tool enables hospital administrators and renal medicine specialists to optimize resource allocation and develop targeted preventative care strategies for this complex patient population.

🛠️ Technical Stack

The dashboard is built using R and the following packages: * flexdashboard: For the layout and dashboard structure. * shiny: For interactive inputs and reactive data processing. * tidyverse & dplyr: For data cleaning and complex transformations. * plotly: For high-fidelity, interactive visualizations. * DT: For interactive data tables. * ggplot2: For custom-rendered slope and bar charts.

📥 Getting Started

To run this dashboard locally: 1. Ensure you have the all_adpkd_1.csv file in your working directory. 2. Open DATA555-dashboard.Rmd in RStudio. 3. Click "Run Document".

To open the dashboard from the website: https://livzhai.shinyapps.io/final_project/