A Case Study on Optimizing HCC 21 (Cachexia) Code Capture
The High Cost of Missed Opportunities
A large health system faced significant revenue leakage due to the inefficiencies of manual chart
reviews. The process was slow, prone to error, and unable to keep up with the volume of data,
leading to substantial under-capture of critical HCC codes.
Inaccurate RAF Scores
Failure to capture all valid diagnoses led to risk scores that didn't reflect the true patient population's acuity.
Keyword Blindness
Coders missed diagnoses documented with synonyms (e.g., "wasting", "malnourished") instead of explicit terms.
Inefficient Workflows
The time-consuming nature of manual reviews created significant backlogs and hindered coder productivity.
Inaccurate RAF Scores
Each missed HCC code represented a direct loss of appropriate reimbursement for care provided.
The Al-Powered NLP Solution
To combat these challenges, the health system deployed a sophisticated Al engine to read and
interpret unstructured physician notes, automating the identification of high-probability missed
diagnoses.
Analyzes Notes
Al ingests millions of unstructured clinical records.
Applies Logic
Identifies synonyms, BMI, and other clinical indicators.
Validates MEAT
Ensures documentation meets criteria (Monitored, Evaluated, Assessed, Treated).
Flags for Coder
Creates a prioritized list for human review and validation.
The Financial & Operational Impact
The Al solution produced immediate, measurable results by analyzing 1 million medical records, demonstrating a clear return on investment.