AI in Clinical Trials Quiz

Answers and explanations appear as you take the quiz.

1. Which long-standing problem in clinical trials does AI most directly target?
   • Slow and difficult patient recruitment
   • The color of medication packaging
   • Hospital cafeteria menus
   • Parking availability at trial sites
2. How can natural-language processing help screen patients for clinical trials?
   • By skipping safety reviews
   • Extract clinical information from electronic health records and compare it with trial eligibility criteria
   • By paying patients more than competitors
   • By shortening the trial to one day
3. What is an external control arm in a clinical trial?
   • A robotic prosthetic limb
   • A computer-generated drug molecule
   • A comparison group drawn from data outside the trial, such as historical or concurrent external data
   • A backup data server
4. What can propensity-score matching do when constructing an external-control comparison?
   • Alphabetical sorting
   • Random number generation alone
   • Spell-checking
   • Balance measured baseline covariates related to treatment assignment
5. Why do predictive dropout and site-performance models matter to trial sponsors?
   • They let sponsors intervene before a trial stalls
   • They design the building architecture
   • They write the marketing slogans
   • They choose staff uniforms
6. What does FDA’s January 2025 draft AI guidance address for drugs and biologics?
   • To slow down all research permanently
   • AI used to generate information supporting regulatory decisions about safety, effectiveness, or quality
   • Because AI is illegal in medicine
   • To reduce the number of doctors
7. What can a Bayesian adaptive clinical-trial design change as evidence accumulates?
   • The original trial question after results are known
   • The identity of past participants
   • Trial elements such as dosing or allocation to study arms under prespecified rules
   • The inclusion criteria without documentation
8. Which pharmacovigilance tasks does FDA identify as possible uses for emerging AI?
   • Only advertising click-through rates
   • Satellite images of clinical sites
   • Unrelated social-media sentiment about packaging
   • Adverse-event intake, data entry, processing, and safety-signal detection or evaluation