Diabetes Risk Classification
About the project
The goal of this project was to analyze patient engagement levels at Clinicas de Azucar, a leading diabetes healthcare provider in Mexico, and develop targeted interventions to improve long-term patient retention in their comprehensive diabetes treatment programs.
Methodology
We employed machine learning techniques, specifically unsupervised clustering method(K-Means), to analyze over 14,000 individual patient records from Clinicas de Azucar. Patients were categorized into three groups based on their engagement levels and likelihood of dropping out from the treatment program: within one year, between one and two years, and after two years. We then identified factors contributing to lower engagement levels and designed targeted interventions to address these issues.
Background
Diabetes is a leading cause of death worldwide, with the situation in Mexico being particularly severe. Clinicas de Azucar offers a wide range of services, including education, A1C reduction, optometry, and psychology, to help patients manage their diabetes effectively. However, the clinic faces challenges in retaining patients long-term and maintaining their engagement with the treatment programs.
Example of clustering. Source: K-Means Data Clustering. In today’s world with the increased… | by Niruhan Viswarupan | Towards Data Science
Interventions Designed
One key finding was that employed patients demonstrated lower engagement levels, likely due to work commitments. We proposed offering extended hours, including evening and weekend appointments, to accommodate their schedules. A randomized controlled trial was designed to test the effectiveness of this intervention, with the aim of increasing patient engagement by 10-20% and retaining an additional 2-4% of the patient population in their second year of registration.
Let’s talk?
If you have any questions about my qualifications or would like to discuss further, please don't hesitate to get in touch!
