AI in Healthcare Revenue Cycle Management Quiz

Answers and explanations appear as you take the quiz.

1. Which X12 transaction pair checks a patient's insurance eligibility?
   • 837/835
   • 276/277
   • 270/271
   • 278 only
2. In the charge capture example, what did the AI find?
   • A documented implanted device that was never billed
   • A duplicate patient record
   • A payer that paid twice
   • An expired provider license
3. Where does autonomous coding work best, according to the guide?
   • Complex inpatient stays with many complications
   • High-volume, narrow specialties such as radiology
   • Any claim, as long as the model is large enough
   • Only for patient statements
4. Why is automating a flawed revenue cycle process not enough?
   • Payers reject automated claims
   • AI tools cannot read claims
   • HIPAA forbids automation
   • It makes the same mistakes faster; the gains come from fixing root causes
5. How should a confidence-threshold design handle uncertain items?
   • Send them to human work queues, and audit a random slice of the automated items
   • Submit them anyway to save time
   • Delete them
   • Send them back to the patient
6. Why does the guide advise preferring standard transactions and APIs over portal bots?
   • Portal bots are illegal
   • APIs are always free
   • Portal bots break when payer websites change
   • Portals cannot show eligibility
7. How should staff follow-up on unpaid accounts be prioritized?
   • Oldest account first, always
   • Alphabetically by patient name
   • Smallest balance first
   • By expected value: the balance times the chance of collecting, divided by the effort
8. How does the guide say to measure the impact of revenue cycle AI?
   • Vendor dashboards alone
   • Before-and-after comparisons of standard metrics over complete reporting periods
   • The number of bots deployed
   • Staff satisfaction surveys only