Dynamic-programming-based heuristic for multi-objective operating theater planning

Abstract

<p style='text-indent:20px;'>This study focuses on the assignment of surgical operations to operating room (OR) blocks to minimize not only the utilization cost of the OR blocks but also the number of patterns generated. We propose a dynamic-programming-based heuristic to solve this problem. Using an iterative formulation, we separate the patterns individually. The computational results show that the proposed heuristic is efficient. Compared with "quasi-optimal" plans, the average cost gap ranges from 0.022 to 0.066 for various scenarios. A lower bound based on column generation procedure is developed. Results show that the average absolute difference of three scenarios are respectively 0.045, 0.081 and 0.115. According to evaluations based on performance indicators from the extant literature, the utilization ratio of the operating theater (OT) varies from 1% underutilization to 2% overutilization in the solutions yielded by the proposed heuristic. This study deals with a multi-objective optimization problem, the second objective of which plays a significant role in reducing the preparation cost, error probability, and staff fatigue in medical systems, where security and human issues are far more emphasized than in other sectors. To the best of our knowledge, our study is the first to investigate such a real-world multi-objective OT planning problem.</p>