AI for Medical Coding Audits Quiz

Answers and explanations appear as you take the quiz.

1. A hospital's AI reviewed every outpatient claim and flagged 400 errors. What is the status of those flags?
   • They are leads until a qualified auditor confirms them
   • They are final findings that must be refunded right away
   • They can be extrapolated across all claims
   • They satisfy a RAC audit automatically
2. Why can't a targeted review of high-risk claims be used to estimate an overpayment across all claims?
   • Targeted reviews are illegal
   • Targeted claims are chosen on purpose, not at random, so they do not represent the whole group
   • RAT-STATS does not support outpatient claims
   • Only payers may review high-risk claims
3. What free tool does the OIG offer for statistical sampling?
   • PEPPER
   • NCCI
   • RAT-STATS
   • CERT
4. The hospital reran Medically Unlikely Edits on paid outpatient claims. What were those checks looking for?
   • Missing patient signatures
   • Out-of-network providers
   • Invalid diagnosis codes
   • Billed units above the maximum usually reported for a service
5. Why does the overpayment rule raise the stakes of AI auditing?
   • It bans AI in compliance programs
   • Once an overpayment is identified, it must be reported and returned within set deadlines, so surfaced problems need a resolution process
   • It requires auditing 100 percent of claims
   • It applies only to Medicare Advantage plans
6. In the inpatient example, what triggered the AI's flag?
   • A missing admission order
   • A duplicate claim number
   • A major complication supported by one mention of acute respiratory failure with no clinical indicators
   • A patient older than 90
7. Which document lists auditing and monitoring as a core compliance element?
   • The OIG's General Compliance Program Guidance from November 2023
   • The ICD-10-CM Tabular List
   • The X12 835 standard
   • The CPT E/M guidelines
8. Why should coding error rates be reported by claim, by code line and by dollar?
   • CMS requires exactly three reports
   • It lets the AI run faster
   • Dollar rates are always zero
   • The three measures can tell different stories about the size and impact of errors