Improving nurse scheduling using genetic algorithm and fairness criteria

Poor work schedule causes job dissatisfaction and high turnover rates among the nurses. Literature shows that a good schedule contributes an enormous impact on the working environment. Thus, the preparation for an effective nurse work schedule is very crucial as nurses play a significant role in delivering efficient healthcare systems. The previous researches on nurse work schedule mainly focus on the optimization technique of the working time and the number of nurses, but only a few focuses on the fairness criteria. Therefore, the goal of this research is to develop a nurse scheduling using Genetic Algorithm (GA) which incorporates the fairness criteria and the optimization technique based on actual hospital environment and relevant literature. The data collection of this study involved eight hospitals. This research is divided into three phases: data collection, formulation and evaluation. In the first phase, the literature is reviewed to generate the criteria of fairness desired in nurse scheduling. This follows with interviews with nurses from varying positions. Then, a questionnaire was developed, distributed to three hospitals with 100 nurses responded. In the second phase, a GA was formulated based on the collected list of fairness criteria. In the third phase, an automated nurse scheduling was developed to evaluate the formulated algorithm. A total of twenty-three original schedules were gathered from Hospital Sultan Ahmad Shah for this purpose. From the series of evaluations, this research indicated that the developed GA which incorporates the fairness criteria has great potential in providing an effective nurse scheduling for healthcare organization