A simulation-based genetic algorithm approach for solving no-wait two stages elective surgery scheduling problem under uncertainty

Abstract

Purpose: The surgical unit is critical for hospital revenue. This study aims to address the no-wait two-stage elective surgery scheduling problem, focusing on minimizing makespan by considering operating room (OR) availability and eligibility, surgeon dedication, and recovery bed availability.

Design/methodology/approach: We propose a Mixed-Integer Linear Programming (MILP) model that integrates stochastic parameters for surgery duration and recovery time. A Genetic Algorithm (GA) is designed for deterministic conditions, while a simulation-based GA handles uncertainty

Findings: Experimental results showed that GA provides near-optimal solutions for small instances within reasonable timeframes. While The simulation-based GA approach provided favorable outputs, with a relative gap to the lower bound of 6.531% on average and a CPU time of 56.286 seconds. The proposed model was applied to a real-world surgical suite, where results showed that the GA finds good solutions that deviate from the lower bound on an average of 4.053% and CPU time of 71 seconds. Also, results showed that simulation-based GA provides solutions with makespan value significantly less than the actual scheduling in the hospital, where the makespan reductions range from 11.771 to 25.080%.

Research limitations/implications: One of the limitations of this study is that it focused only on the surgery and recovery stage without considering the pre-operative stage. Also, the application of the proposed model was limited to small and medium-sized cases.

Practical implications: The proposed model and solution approach offer a practical and effective tool for hospital administrators to address the complexity of surgery scheduling, thereby enhancing operational efficiency within hospital settings.

Social implications: Improving the surgical scheduling process impacts on the overall performance of the hospital, as it leads to reducing overtime and waiting time, improving efficiency, increasing productivity, and thus reducing total costs.

Originality/value: This study addresses two-stage surgical scheduling while considering OR availability and eligibility, dedicated surgeons, PACU bed availability, and two probabilistic parameters: surgery durations and recovery time. Also, using a simulation-based GA to solve the problem. These interactive aspects represent a research gap as they have not been comprehensively studied in previous research.