AI for Continuing Medical Education Quiz

Answers and explanations appear as you take the quiz.

1. According to the guide, which type of credit are US physicians most often required to earn for CME?
   • Continuing education units issued by any employer
   • AMA PRA Category 1 Credit
   • Self-reported reading hours with no provider
   • Credit issued directly by an AI chatbot
2. Which sequence best describes point-of-care learning as an accredited CME format?
   • Attend a live lecture, then take a quiz
   • Pass a proctored board examination
   • Read a journal cover to cover each month
   • Look up the answer to a real clinical question, apply it and document the learning
3. Why might an AI tool give an outdated answer about a clinical guideline?
   • The model has a training cutoff, and retrieval may surface an older guideline version
   • AI tools can never read guidelines
   • Guidelines are never revised
   • Accreditors block access to current guidelines
4. An education team uses a language model to draft practice questions for an accredited activity. What must happen before those questions count toward CME?
   • Nothing, since the model is trained on medical text
   • The vendor must certify the model
   • Qualified human review for content validity and independence from commercial influence
   • Each question must be translated into three languages
5. In a retrieval-augmented literature tool, what happens before the answer is generated?
   • The model deletes outdated articles from the internet
   • The system retrieves the most relevant passages from its indexed guidelines and articles
   • A physician writes the answer by hand
   • The accredited provider approves the question
6. An AI summary applies a trial's finding to all adults, but the trial enrolled only patients over 75. Which failure mode does the guide describe here?
   • A retrieved passage lacked the context that limits it, such as the study population
   • The tool refused to answer
   • The citation linked to a nonexistent website
   • The model produced a correct and complete answer
7. Why does the guide say that reading an AI summary is not the same as learning for credit?
   • AI summaries are always too long
   • Credit can only be earned in person
   • Summaries are copyrighted
   • Credit formats ask the clinician to reflect on how the information changed practice
8. When a clinician earns credit after using an AI-based reference tool, who generally awards that credit?
   • The AI model itself
   • The patient
   • An accredited provider, often working with the tool's company
   • Any colleague who reviews the search