EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

2215. Dealing with uncertainty in surgery duration

Invited abstract in session TB-10: Surgery Scheduling and Operating Room Planning, stream OR in Health Services (ORAHS).

Tuesday, 10:30-12:00
Room: 11 (building: 116)

Authors (first author is the speaker)

1. Theresia van Essen
Delft University of Technology

Abstract

Operating rooms are one of the most expensive resources of a hospital, which makes it important to use them as efficiently as possible. This involves maximizing utilization, but also minimizing the probability of overtime, which are two conflicting objectives. The problem of scheduling surgeries is usually modelled as an integer linear program (ILP), where maximizing utilization is considered as objective, while restricting the probability of overtime. The constraint used to restrict the probability of overtime includes the lognormally distributed surgery duration, which typically leads to a nonlinear constraint where the total surgery duration in one operating room is a stochastic variable.
In the literature, various ways to deal with this nonlinearity are considered: 1) only use the expected surgery duration and limit the total surgery duration, 2) model the uncertain surgery duration using a normal distribution and linearize the resulting constraint using a piecewise linear function, 3) model the uncertain surgery duration using a discrete probability distribution resulting in a linear constraint, and 4) model the uncertain surgery duration using the lognormal distribution and use column generation and simulation to solve the resulting model.
We implemented and compared all mentioned methods on both solution quality and computation time. To determine the solution quality, we tested the resulting schedules using simulation.

Keywords

Status: accepted

Back to the list of papers