AI in Hospital Patient Flow and Bed Management Quiz

Answers and explanations appear as you take the quiz.

1. In patient flow, what does 'boarding' mean?
   • Checking patients in at the front desk
   • Moving patients between hospitals by ambulance
   • Admitted patients waiting in the emergency department because no inpatient bed is available
   • Discharging patients to nursing homes
2. Which prediction task estimates when each current inpatient is likely to leave?
   • Length-of-stay and discharge prediction
   • Demand forecasting
   • Sepsis risk scoring
   • Ambulance diversion planning
3. Why can't capacity forecasts treat all beds as interchangeable?
   • All beds are identical in modern hospitals
   • Beds are counted only once a year
   • Only emergency department beds matter
   • Specialized beds such as ICU or telemetry serve different patient needs
4. Which company partnered with Johns Hopkins Hospital on its 2016 capacity command center?
   • Qventus
   • GE Healthcare
   • LeanTaaS
   • Epic
5. According to the guide, what is a key misconception about patient-flow AI?
   • That it can forecast admissions
   • That it uses historical data
   • That the model itself frees beds without anyone acting on its predictions
   • That it is used in command centers
6. Why should simple baselines such as seasonal averages always be reported in hospital demand forecasting?
   • Strong weekly and seasonal patterns make them hard to beat
   • They are required by law
   • They always outperform machine learning
   • They eliminate the need for uncertainty estimates
7. Why do managers need prediction intervals, not just a single forecast number?
   • Intervals make dashboards look more detailed
   • Intervals remove the need for staff
   • Point forecasts cannot be calculated for hospitals
   • They need to judge the range of likely demand when deciding whether to open a surge unit
8. What is a feedback loop in patient-flow prediction?
   • Patients giving feedback on their stay
   • Acting on predictions changes the outcomes the model was trained to predict
   • A model automatically retraining every hour
   • Staff sending comments to the vendor