Computer-Assisted Coding Explained Quiz

Answers and explanations appear as you take the quiz.

1. What separates computer-assisted coding from autonomous coding, according to the guide?
   • CAC does not use natural language processing
   • In CAC, a human validates every chart, while autonomous systems send some encounters to billing without review
   • CAC is used only for outpatient claims
   • Autonomous coding requires coders to type every code by hand
2. CAC suggests a pneumonia code in an outpatient chart based on a note reading "rule out pneumonia." Why does the coder reject it?
   • Pneumonia codes are not part of ICD-10-CM
   • CAC cannot read progress notes
   • The physician must code pneumonia personally
   • Outpatient guidelines do not allow coding uncertain diagnoses as confirmed
3. Which phrase would an NLP engine's experiencer detection keep from being coded as the patient's own condition?
   • "Patient has acute kidney injury"
   • "Status post appendectomy"
   • "Mother had breast cancer"
   • "Patient reports chest pain"
4. In CAC evaluation, what does low recall mean in practice?
   • The system misses correct codes, so coders must still hunt for them
   • Most suggested codes are wrong
   • Claims are billed too quickly
   • Coders reject every suggestion
5. What does the metric "discharged not final billed" (DNFB) days measure?
   • The number of codes per chart
   • How many physicians write queries
   • The length of a patient's hospital stay
   • How long discharged accounts wait for final coding
6. Why was CAC adoption in the United States tied to October 2015?
   • That is when CPT was retired
   • That is when HIPAA took effect
   • That is when the switch from ICD-9 to ICD-10 raised fears of lost coder productivity
   • That is when autonomous coding was banned
7. Coders accept nearly every CAC suggestion for one physician's patients. According to the guide, why should this trigger review?
   • High acceptance always proves the engine is accurate
   • High acceptance can mean coders have stopped checking rather than that the suggestions are correct
   • Acceptance logs are not allowed in audits
   • Physicians must approve every accepted code
8. Which step in the NLP pipeline keeps "no evidence of pneumonia" from producing a pneumonia code?
   • Negation detection, using approaches such as NegEx or transformer models
   • Section detection alone
   • Mapping to CPT
   • Sentence splitting alone