Modélisation Epidémiologie et Surveillance des Risques Sanitaires (MESuRS) Laboratory, Conservatoire National des Arts et Métiers (Cnam), Paris, France.
PLoS One. 2021 May 14;16(5):e0251779. doi: 10.1371/journal.pone.0251779. eCollection 2021.
High turnover among healthcare workers is an increasingly common phenomenon in hospitals worldwide, especially in intensive care units (ICUs). In addition to the serious financial consequences, this is a major concern for patient care (disrupted continuity of care, decreased quality and safety of care, increased rates of medication errors, …).
The goal of this article was to understand how the ICU-level nurse turnover rate may be explained from multiple covariates at individual and ICU-level, using data from 526 French registered and auxiliary nurses (RANs).
A cross-sectional study was conducted in ICUs of Paris-area hospitals in 2013. First, we developed a small extension of a multi-level modeling method proposed in 2007 by Croon and van Veldhoven and validated its properties using a comprehensive simulation study. Second, we applied this approach to explain RAN turnover in French ICUs.
Based on the simulation study, the approach we proposed allows to estimate the regression coefficients with a relative bias below 7% for group-level factors and below 12% for individual-level factors. In our data, the mean observed RAN turnover rate was 0.19 per year (SD = 0.09). Based on our results, social support from colleagues and supervisors as well as long durations of experience in the profession were negatively associated with turnover. Conversely, number of children and impossibility to skip a break due to workload were significantly associated with higher rates of turnover. At ICU-level, number of beds, presence of intermediate care beds (continuous care unit) in the ICU and staff-to-patient ratio emerged as significant predictors.
The findings of this research may help decision makers within hospitals by highlighting major determinants of turnover among RANs. In addition, the new approach proposed here could prove useful to researchers faced with similar micro-macro data.
医护人员的高离职率是全球医院，尤其是重症监护病房（ICU）日益普遍的现象。除了严重的财务后果外，这也是患者护理的主要关注点（护理连续性中断、护理质量和安全性下降、药物错误发生率增加等）。
本文旨在通过来自法国注册护士和助理护士（RAN）的 526 名数据，从个人和 ICU 层面的多个协变量来解释 ICU 级别护士离职率。
2013 年在巴黎地区医院的 ICU 中进行了一项横断面研究。首先，我们对 Croon 和 van Veldhoven 于 2007 年提出的一种多水平建模方法进行了小型扩展，并通过全面的模拟研究验证了其性质。其次，我们将该方法应用于解释法国 ICU 中 RAN 的离职率。
基于模拟研究，我们提出的方法允许对组级因素的回归系数进行估计，其相对偏差低于 7%，对个人级因素的相对偏差低于 12%。在我们的数据中，观察到的 RAN 离职率的平均值为每年 0.19（标准差=0.09）。根据我们的结果，同事和主管的社会支持以及在该职业中的工作经验与离职呈负相关。相反，子女数量和由于工作量而无法跳过休息时间与较高的离职率显著相关。在 ICU 层面，床位数、ICU 中是否存在中级护理床位（连续护理单元）以及员工与患者的比例是显著的预测因素。
本研究的结果可能通过突出 RAN 离职的主要决定因素来帮助医院的决策者。此外，这里提出的新方法可能对面临类似微观-宏观数据的研究人员有用。
