A large medical research association needed to improve model performance on broader demographics to catch pre-diabetic patients earlier.
Challenge
Smaller demographic groups lacked available historical data to train prediction models.
Individual medical history data was dispersed across hundreds of hospitals and, therefore, inaccessible for analytics.
Centralizing the disbursed medical data would expose the organization to HIPPA and CCPA/CPRA scrutiny and come with long delays and high costs
Solution
Devron collaborated with the association’s analytics team to deploy on historical data wherever it resided.
Devron sped up model training and increased accuracy with medical claims data from sources like internal, clearinghouse, public, and health data companies.
Increased data access meant faster model training and rapid predictions of likely pre-diabetics, which helped address causes earlier and ultimately save lives.
