Clinical Readmission Risk Assessment

AI-assisted healthcare prediction system designed to estimate patient readmission risk using clinical history, medication patterns, diagnosis categories, and hospital utilization metrics.

Model Architecture: This platform utilizes a Logistic Regression (OvR) model with preprocessing powered by Binary Encoding and StandardScaler. Hyperparameters were optimized using RandomizedSearchCV to improve predictive performance and clinical categorization stability.

Patient Information

Basic demographic and hospitalization details associated with patient history.

Age

Time in Hospital

Clinical Activity Metrics

Historical hospital utilization and medical procedure indicators.

Lab Procedures Count

Number of Procedures

Number of Diagnoses

Emergency Visits

Inpatient Visits

Unique Medications Prescribed

Medication Changes

Diagnosis Categories

Major clinical diagnosis groups associated with patient admission history.

Primary Diagnosis

Secondary Diagnosis

Additional Diagnosis

Medication & Clinical Indicators

Medication behavior and biochemical indicators associated with diabetic monitoring.

Diabetes Medication Prescribed?

Metformin Prescribed?

Insulin Prescribed?

A1c Result

Change in Medication?

Visited by Outpatients?

Emergency Case?

Emergency at Admission?

Might Return After Discharge?

Blood Glucose Level (mg/dL)

Clinical Disclaimer: This prediction is generated using a statistical machine learning model and is intended for research and educational purposes only. It should not replace physician evaluation, medical diagnosis, or professional clinical judgment.