Using Data Mining Predict Hospital Admissions From The Emergency Department
G.Rama Rao,
Ch. Ajay,
B. Sruthika,
D. Naga Bhavana,
D. Kapil
Emergency departments (EDs) are critical entry points to hospitals, yet unpredictable admission
rates often strain resources and delay patient care. This study explores the use of data mining and
classification techniques to predict hospital admissions based on ED patient records. Historical datasets
containing demographic details, triage information, and clinical variables were processed using data
mining methods to uncover hidden patterns and correlations. Classification algorithms were then
applied to categorize patients into “admission” or “discharge” outcomes, with performance evaluated
across multiple models.
To ensure the confidentiality and integrity of sensitive patient data, cryptographic methods were integrated
throughout the workflow, enabling secure storage, transmission, and analysis of medical records. The
combined approach demonstrates that predictive accuracy can be achieved without compromising data
privacy. Results indicate that classification models, when supported by robust cryptographic safeguards,
provide reliable forecasts of admission likelihood, offering hospitals a practical tool for proactive
resource allocation and improved patient flow management.
