AI in Antibiotic Stewardship Quiz

Answers and explanations appear as you take the quiz.

1. Why is a hospital antibiogram limited as a guide for one specific patient's empiric therapy?
   • It only lists antiviral drugs
   • It is updated every hour and hard to read
   • It reports hospital-wide susceptibility averages and ignores the individual's history
   • It is only available after cultures return
2. Which pair correctly matches each prescribing error with its cost, as the guide describes?
   • Too narrow risks treatment failure; too broad drives resistance and C. difficile
   • Too narrow causes C. difficile; too broad causes treatment failure
   • Both errors only raise costs
   • Neither error affects patient outcomes
3. What is the goal of applying machine learning to MALDI-TOF mass spectra in the microbiology lab?
   • To identify the patient's blood type
   • To estimate resistance hours before conventional susceptibility testing finishes
   • To measure antibiotic levels in blood
   • To replace the need to grow bacteria at all
4. Why do resistance models trained on culture results suffer from selection bias?
   • Cultures are only done on children
   • Labs report results in different languages
   • Only patients who were cultured get labels, and they are often sicker or more complex
   • Resistant bacteria cannot be cultured
5. Why does the guide recommend temporal validation for resistance models?
   • Resistance patterns and CLSI or EUCAST breakpoints change over time
   • It makes models train faster
   • Regulators forbid random splits
   • Older data is always more accurate
6. Which decision can a resistance-prediction model NOT make, according to the guide?
   • Which drug is likely to be ineffective
   • The probability of MRSA
   • The narrowest drug likely to work
   • Whether the patient needs antibiotics at all, such as telling true infection from asymptomatic bacteriuria
7. In prospective audit and feedback, what happens?
   • Pharmacists or infectious disease physicians review active prescriptions and recommend changes
   • Patients rate their antibiotics after discharge
   • Drugs are blocked until a lab result arrives
   • Software automatically cancels all broad-spectrum orders
8. What did the 2020 Kanjilal study on outpatient urinary tract infections report?
   • The algorithm eliminated the need for urine cultures
   • Tested retrospectively, the algorithm could reduce second-line antibiotic use while keeping ineffective therapy similar to or lower than clinicians' choices
   • Clinicians refused to use the algorithm in a randomized trial
   • The algorithm increased resistance rates