How H2O’s OR Block Scheduling solution freed up 12 beds during peak demand, saving up to $20K per night
September 2022
Problem & Motivation
Accurate planning for surgeries, especially during peak demand, can become very complicated given the different work schedules of each surgeon and the varying patient recovery time post-surgery. H2O’s OR Block Scheduling solution uses data-driven formulation to find the optimal schedule.
Block Optimization
During a successful test across 200 surgeons, the H2O optimization algorithm was able to shift planned surgeries in a way that minimized the number of beds during peak time and distributed them more evenly across the week. The result was a decrease in the average peak census by 12.6 beds which meant up to $20K savings per night for the hospital, by shifting the schedule of only 10 surgeons. The algorithm took into account the level of scheduling flexibility for each surgeon to ensure that the outputted calendar was practical.
Descriptive Analysis
The H2O OR Block Scheduling solution includes a user-friendly interface with many charts and tables that aids management in their decision-makring and in viewing the hospital from a different perspective. Below are some examples.
Rooms: Below the manager can view the average operation time and room utilization trends to decide where there is room to better distribute surgeries and maximize the utility of the rooms throughout the week.
Department: Below the manager can look at a detailed view of census for a specific department to get a better understanding of trends and be able to decide where to make scheduling changes.
Surgeons: Below the manager can see the difference between the current schedule of surgeries compared to the optimized schedule and decide if they want to accept the modified schedule to improve the peak hours of the week.
Services: Below is a high level view of a surgeon’s profile for managers to get a better sense of the services most in demand and where there may be hiring needs or a need for additional support.
