The never-ending note

Finishing a patient visit in primary care isn't just about the medicine — it's a gauntlet of clicks. After documenting the encounter, confirming orders, and writing up an assessment and plan, physicians are then prompted to verify Hierarchical Condition Category (HCC) codes: a Medicare system designed to capture how sick and complex a patient truly is, so insurers can reimburse accordingly. The catch? Many of these conditions are managed by specialists, yet primary care doctors bear the burden of re-confirming them — sometimes annually — even for permanent conditions.

Dr. Fred Pelzman argues this system is both inefficient and illogical. Insurers already receive claims for dialysis sessions, chemotherapy, ICU stays, and defibrillator placements — so why do physicians need to manually re-enter what the data already shows? He makes the case that AI could do this work far better, scanning EHRs, pharmacy records, imaging, and claims data to automatically generate accurate patient risk profiles.

Key Takeaways

• HCC coding is required (often annually) to help Medicare calculate risk-adjusted reimbursements for primary care practices
• Primary care physicians must verify codes for conditions they may not be actively managing — adding significant documentation burden
• AI-driven automation of risk scoring could reduce redundant clicking and produce more accurate complexity assessments
• Under value-based care models like ACOs, reimbursement bonuses are partly tied to HCC code capture, creating financial pressure on PCPs

Why it matters: Documentation burden is a leading driver of physician burnout. If AI can automate HCC risk scoring by pulling from existing data sources, it could meaningfully reduce administrative load for primary care physicians — and potentially improve the accuracy of reimbursement across the board.