Medicare Risk Adjustment

Identify undiagnosed conditions for reimbursements

In the United States, the Centers for Medicare & Medicaid Services (CMS) reimburses private Medicare insurers based on beneficiary risk. Most insurers rely on claims data to calculate member risk, but that provides a narrow, point-in-time view of health. To address this gap, Databricks and John Snow Labs Solution Accelerator for Medicare Risk Adjustment automates the extraction of undiagnosed member conditions from unstructured clinical notes with NLP models, improving downstream reimbursements.

• Detect undiagnosed conditions in unstructured text with NLP
• Append ICD-10 and HCC codes with pretrained models
• Calculate revised RAF scores to improve reimbursements