Risk Adjustment Models Overview Flashcards
6 cards from real CRC practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 6 Risk Adjustment Models Overview flashcards as text
What is the specific time relationship between diagnosis capture and payment in the CMS-HCC prospective model?
Answer: Diagnoses in the current calendar year affect next year's capitation payments
Diagnoses documented in the data collection year determine the following year's payment, creating a one-year lag.
What role do 'payment condition counts' play in newer CMS-HCC versions?
Answer: They add incremental payment for members with many HCCs, reflecting disproportionately higher costs of complex patients
Payment condition counts recognize that patients with many concurrent HCCs have costs exceeding the sum of individual weights.
How does the CMS-HCC model handle institutional vs community-dwelling beneficiaries?
Answer: Separate model segments with different coefficients for each population
Separate segments with different coefficients reflect dramatically different cost structures between populations.
A plan's risk scores systematically under-predict costs by 8%. What might explain this?
Answer: The population has characteristics not fully captured by the model, such as functional limitations or social determinants
The CMS-HCC model explains about 12% of individual cost variation. Factors like functional status and social determinants are not captured.
What is the ESRD model in CMS risk adjustment?
Answer: A separate model with unique HCCs, dialysis and transplant status factors, and specific cost calibration
CMS maintains a separate model including ESRD-specific HCCs, dialysis modality factors, transplant status, and disease duration.
What does 'predictive ratio' measure in risk adjustment model evaluation?
Answer: The ratio of predicted costs to actual costs for a subgroup, where 1.0 is perfect prediction
A ratio below 1.0 means under-prediction; above 1.0 means over-prediction. This reveals model fairness across populations.