New Research Publication: AI-Based Surgery Scheduling Cuts Hospital Overtime and Patient Waiting Times

A new study from our chair demonstrates how a novel data-driven planning approach that integrates case-specific information—such as treatment type and individual patient characteristics—can significantly improve the scheduling of surgeries. By accounting for this variability, the approach reduces patient waiting times and staff overtime by more than 35%.

The article, titled “Case-Individual Data-Driven Optimization for Surgery Planning”, was published in the journal OR Spectrum (Springer Nature). The research is based on a collaboration with the Department of Gastroenterology at the University Hospital Würzburg, which provided access to detailed, anonymized data on surgeries performed in 2021 and 2022. The study was conducted by:

• Janine Rottmann, Research Associate, Chair of Information Systems and Systems Development

• Kai Günder, Research Associate, Chair of Logistics and Quantitative Methods

• Prof. Dr. Richard Pibernik, Chair of Logistics and Quantitative Methods

Why surgery scheduling matters

Operating theatres account for more than 40 % of a hospital’s total costs, so even small inefficiencies have outsized financial and clinical repercussions. The biggest headache is the extreme variability of surgery durations, driven by, for example, the specific procedure type and individual patient factors. Uncertainty in the schedule leads to:

• Prolonged waiting times for patients

• Expensive overtime for surgical teams

• Under- or overutilised operating rooms

A two-step, data-driven solution

The Würzburg team combines machine-learning models and optimization to tame this variability:

1. Predictive analytics: Electronic health record data are used to estimate the uncertainty surrounding the duration of each individual surgery, factoring in detailed case-specific information such as patient characteristics, the attending medical team, and specifics of the type of surgery being performed. 

2. Optimized sequencing and room allocation: Both operating room assignments and surgery start times are determined automatically using an optimization approach. The goal is to jointly minimize patient waiting times and staff overtime, balancing the trade-off between the two. The optimization is based on the estimated uncertainty of each surgery scheduled for the day.

Real-world impact

When tested against current scheduling practice, the new method achieved:

• 35 % reduction in patient wait times and staff overtime

• Noticeably smoother daily workflows thanks to variance-based sequencing

• More flexible use of existing OR capacity through smarter room assignments

Outlook

By explicitly modelling case-specific uncertainty, the researchers demonstrate that hospitals can deliver better care and use resources more sustainably—without adding extra staff or facilities. 

For full details, you can read the article in Open Access.