考虑工作倦怠的护士夜班指派问题研究

Markov Decision Process-based Nurse Night Shift Assignment with Working Fatigue Considered

面对死亡和疾病,不可避免经常性地加班,护士很容易遭受护理工作倦怠综合征的困扰。在辽宁省某三甲医院开展的试验表明,护士每月夜班数与其工作倦怠程度有显著的正相关关系。在一定的假设前提下,建立了有限时域的马尔科夫决策支持模型,研究了不同倦怠程度护士每周最优夜班数的指派问题,目标是最小化护士工作倦怠惩罚、夜班成本和护士空闲惩罚。通过解析分析,为不同倦怠程度的护士指派最优夜班数是一个阀值型的控制策略。数值试验结果表明,模型中不同参数的组合会对最优策略产生影响。最后,敏感度分析考虑了工作倦怠成本,状态转移概率和决策末期各状态成本对最优控制策略的影响。

Exposed to death and illness and faced with inevitable regular shift work, nurses are more likely to suffer from burnout syndrome. A cross-sectional experiment conducted in a tertiary hospital in Liaoning Province shows that there is a significantly positive relationship between the number of night shifts and nurse burnout level. Under the assumption of transition probability, a finite-horizon Markov Decision Process model was proposed to determine the optimal number of night shifts for individual nurse with the objective to minimize the weighted cost of burnout, employment, and underutilization. At each burnout level, the optimal number of night shift is proved to be a threshold-type policy. Numerical experiments and sensitivity analysis show the influence of different parameters on the optimal schedule and cost functions.