โ† All CRC Flashcard Decks

Data Mining and Predictive Analytics Flashcards

6 cards from real CRC practice questions. Tap to flip, then mark Knew It or Still Learning โ€” missed cards come back until you master them.

Read the first 6 Data Mining and Predictive Analytics flashcards as text
  1. What is the recommended minimum sample size for validating a suspected documentation gap pattern?

    Answer: A statistically significant sample with 95% confidence and 5% margin of error

    Statistical rigor requires a sample calculated based on population size, confidence level, and margin of error.

  2. What is the most critical NLP limitation for risk adjustment applications?

    Answer: NLP may identify negated conditions as positive findings

    Negation detection is critical. Clinical notes frequently mention conditions in negative context like 'ruled out' or 'no evidence of.'

  3. A plan's RAF scores increase 3% year-over-year for five years while FFS benchmarks remain flat. What analytical framework applies?

    Answer: Apply coding intensity adjustment analysis to distinguish true health changes from coding pattern changes

    Persistent RAF growth exceeding FFS benchmarks suggests coding practice changes rather than true population health changes.

  4. Which data visualization is most effective for presenting risk adjustment patterns to non-technical stakeholders?

    Answer: Heat maps showing capture rates compared to benchmarks by provider

    Heat maps provide intuitive visual comparison that non-technical stakeholders can immediately understand.

  5. A model identifies 500 members as likely having undocumented CKD. After outreach, only 200 are confirmed. What metric describes this?

    Answer: 40% positive predictive value (precision)

    PPV measures the proportion of positive predictions that are true positives: 200/500 = 40%.

  6. 12% of submitted codes come from encounters where the provider specialty does not typically diagnose the condition. What does this suggest?

    Answer: The finding warrants investigation for possible copy-forward artifacts

    Provider specialty-diagnosis mismatches warrant investigation but are not automatically invalid. Common causes include copy-forward of problem lists.