Scheduling Smarter: Scheduling Decision Impact on Nurse-Aide Turnover
研究分析了长期护理机构中兼职护理助手的排班决策（工作时长和同事组合）对其离职率的影响，发现工作时长与离职率呈U型关系，高同事变动性增加离职率，管理者可通过优化排班降低离职率。
Research problem: High turnover rates in long-term nursing facilities exacerbate the existing shortage of caregivers, a trend that will only worsen as the population of the United States ages. Part-time certified nursing assistants (CNAs) provide a significant amount of patient care in these facilities. CNAs also have high annual rates of turnover, which can harm health outcomes and increase the cost of care. In this study, we empirically analyze the effect of scheduling decisions on part-time CNA turnover. We explore three research questions that examine both how much and with whom a CNA is scheduled to work. We seek to empirically answer how (1) hours worked and (2) coworker variability influence turnover for part-time CNAs and if (3) coworker variability moderates the relationship between hours worked and CNA turnover. Methodology/results: Using novel data from one of the nation’s largest nursing home organizations, which includes data for 6,221 part-time CNAs at 157 facilities in the United States over a 26-month period, we identify two scheduling levers managers may be able to use to influence turnover. As hours worked increase, turnover first decreases and then increases, demonstrating a nonlinear U-shaped relationship between hours worked and turnover. We also find that high coworker variability increases turnover while also moderating the effects of hours worked on turnover. In post hoc analyses, we demonstrate that high CNA turnover has negative impacts on patient health. Managerial implications: These findings suggest that managers may be able to leverage part-time CNA scheduling to reduce turnover, improving both the quality and cost of care. Specifically, we demonstrate that managers can reduce CNA turnover by increasing hours worked, scheduling coworkers together consistently and doing both simultaneously. Supplemental Material: The online appendices are available at https://doi.org/10.1287/msom.2021.0533 .