Modeling for Change of Daily Nurse Calls After Surgery in an Orthopedics Ward Using Bayesian Statistics
Modeling for Change of Daily Nurse Calls After Surgery in an Orthopedics Ward Using Bayesian Statistics
复制标题
DOI：
10.1097/cin.0000000000000712
复制
发表时间：
2021-07-01
影响因子：
1.3
通讯作者：
Mori, Taketoshi
中科院分区：
文献类型：
--
作者：
Noguchi, Hiroshi;Miyahara, Maki;Mori, Taketoshi
护士访视数据可以用来评价护理质量。然而，传统的基于频率的统计可能不容易适用于护士来电，因为个体差异很大，每天的来电变化很大。我们打算提出一个基于贝叶斯统计的护士来电的概率模型。我们构建了包括护士呼叫日常变化、个体变异性和根据特征(年龄和性别)进行调整的模型。根据2014年4月至2017年10月骨科病房的数据，分析手术后护士呼叫的差异。结果显示，从术后第1天到第10天，接受过骨科手术的患者和接受过肿瘤手术等其他手术的患者之间的护士呼叫存在差异。此外，在使用额外止痛药的患者和没有使用止痛药的患者之间，从术后第1天到第8天的护士电话也有差异。虽然分析需要关于每天护士呼叫变化和每天固定数据样本的多次比较，但我们使用贝叶斯统计的方法可以检测出周期和显著差异。这表明基于贝叶斯统计的护士呼叫模型可以用于分析护士呼叫的变化。
Nurse call data may be used to evaluate the quality of nursing. However, traditional frequency-based statistics may not easily apply to nurse calls due to the large individual variability and daily call changes. We intended to propose a probabilistic modeling of nurse calls based on Bayesian statistics. We constructed the model including nurse call daily changes, individual variability, and adjustment according to characteristics (age and sex). Nurse call differences after surgery were analyzed based on data from the orthopedic ward from April 2014 to October 2017. Results show that there were differences in nurse calls from day 1 to day 10 after surgery between patients who had undergone orthopedic surgery and those who had undergone other surgeries such as tumor surgery. Furthermore, there were differences in nurse calls from day 1 to day 8 after surgery between patients who used extra pain relief medicine and those who did not. Although the analysis required multiple comparisons regarding daily nurse call changes and fixed data samples per day, our approach using Bayesian statistics could detect the periods and significant differences. This indicates that our nurse call modeling based on Bayesian statistics may be used to analyze nurse call changes.