Hospital Readmission Analysis (SQL Project)

Project Overview

This project analyzes hospital readmission data to uncover patterns and insights related to patient demographics, medical specialties, procedures, and chronic conditions. The analysis focuses on identifying high-risk patients, understanding readmission trends, and providing actionable recommendations for improving healthcare outcomes.

Data Source

• Dataset: hospital_readmissions.csv
• Total Rows: 25,000 records
• Fields: Patient age, specialty, diagnosis, lab procedures, medications, and readmission status

Key Objectives

1. Analyze readmission rates by department, diagnosis, and age group
2. Identify high-risk patients (multiple readmissions)
3. Assess impact of diabetes management on readmissions
4. Explore average length of stay by specialty
5. Automate data refresh using a Stored Procedure

Tools Used

Tool	Purpose
PostgreSQL	Database Management & SQL Queries
pgAdmin	Query Execution & Data Import
Python (for initial analysis)	Exploratory Analysis
GitHub	Version Control & Portfolio Showcase

Folder Structure

/hospital_readmission_analysis/
│-- README.md                    # Project Overview
│-- data/
│   └── hospital_readmissions.csv # Raw Data
│-- sql/
│   ├── schema.sql                 # Table Creation Script
│   ├── data_cleaning.sql          # Data Cleaning Queries
│   ├── analysis_queries.sql       # Analytical Queries
│   ├── views_and_ctes.sql         # CTEs & Views
│   ├── stored_procedures.sql      # Stored Procedure for Refresh
│-- insights_report.md            # Key Findings Summary

Database Schema

Column Name	Data Type	Description
age	VARCHAR(20)	Age group
time_in_hospital	INTEGER	Duration of stay
n_lab_procedures	INTEGER	Number of lab tests
n_procedures	INTEGER	Number of non-lab procedures
n_medications	INTEGER	Total medications prescribed
n_outpatient	INTEGER	Outpatient visits in history
n_inpatient	INTEGER	Inpatient visits in history
n_emergency	INTEGER	Emergency visits in history
medical_specialty	VARCHAR(50)	Treating department
diag_1, diag_2, diag_3	VARCHAR(50)	Primary, secondary, tertiary diagnosis
glucose_test	VARCHAR(10)	Glucose test done (yes/no)
A1Ctest	VARCHAR(10)	A1C test done (yes/no)
change	VARCHAR(10)	Medication change (yes/no)
diabetes_med	VARCHAR(10)	On diabetes meds (yes/no)
readmitted	VARCHAR(10)	Readmitted within 30 days (yes/no)

Analytical Questions Addressed

• Readmission Rates by Specialty, Diagnosis, and Age
• High-Risk Patients Analysis
• Length of Stay Analysis
• Diabetes Medication Impact

Key Findings

• Older patients (70-90 years) have the highest readmission rates.
• Specialties like Family/General Practice and Emergency have the highest readmissions.
• Patients on diabetes medications are readmitted more often.
• Internal Medicine and Surgery patients have the longest stays.

Stored Procedure Automation

SELECT refresh_high_risk_patients();