Dynamic Scheduling for Veterans Health Administration Patients using Geospatial Dynamic Overbooking

Area of Science:

• Health Informatics
• Operations Research
• Mobile Computing

Background:

• Veterans Health Administration (VHA) medical outpatient clinics face significant productivity losses due to high patient no-show and cancellation rates.
• Inefficient scheduling systems contribute to unused appointment slots and reduced healthcare access.

Purpose of the Study:

• To develop and evaluate a dynamic scheduling system utilizing mobile computing and geo-location data to mitigate VHA clinic no-show and cancellation rates.
• To enhance the efficiency and productivity of medical outpatient clinics through real-time schedule optimization.

Main Methods:

• A dynamic scheduling system was designed using geo-location data to estimate patient arrival probabilities.
• Real-time schedule updates were implemented, with appointments canceled if arrival probability fell below a threshold.
• Canceled appointments were reassigned to prioritized patients from a waitlist based on their arrival probability.
• Synthetic patient data were generated for feasibility testing; arrival prediction models were validated using real-world taxicab data.

Main Results:

• The dynamic scheduling system demonstrated potential in reducing unused appointments.
• The system showed minimal risk of double booking, even with prediction inaccuracies.
• Geo-location data-driven scheduling offers a feasible approach to optimize clinic operations.

Conclusions:

• Dynamic scheduling leveraging geo-location data can significantly decrease appointment no-shows and cancellations in VHA clinics.
• The proposed system offers a practical solution for improving clinic efficiency and patient access.
• Privacy concerns associated with location data require careful consideration and mitigation strategies.