Alcohol Dependent Patient’s Case Automation Tool with Embedded Predictive Models

In response to the growing need for efficient healthcare data management, paper introduces a patient data retrieval system tailored foraccessing to the patient records by healthcare providers. The system is fortified with predictive models for alcohol risk, Age of Alcohol Initiation, and Sobriety Period, facilitating in-depth analysis to inform tailored treatment strategies for alcohol-addicted patients. Emphasizing scalability to accommodate escalating patient data volumes, the proposed web application caters to two users’, administrators and counsellors. These users can seamlessly add, edit, and visualize patient information using Excel sheets, alongside the integration of three predictive models. This dual-functionality system provides a user-friendly interface, enhancing accessibility and decision-making for effective alcohol addiction treatment.The data utilized for the predictive models was sourced from the SDM De-addiction and Research Centre in Ujire, India. In the evaluation of Risk Level and Age of Alcohol Initiation (AAI) prediction models, the Random Forest classifier demonstrated superior performance, achieving an accuracy of 92.10% and 94.88%, respectively. Additionally, the GaussianNB model outperformed others in the Sobriety Period prediction, attaining an accuracy rate of 98.68%.So here the application serves dual purposes: firstly, it mitigates the need for manual entry of alcohol rehabilitation patient data into Electronic Health Records (EHR), streamlining and automating the process. Secondly, the integration of predictive models enhances the accuracy and efficiency of patient assessment. By leveraging these models, the application facilitates a more precise and efficient evaluation of alcohol addicted patients, contributing to an advanced level of health care delivery.