Department of Management Sciences, University of Iowa, Iowa City, IA 52242, USA.
Anesth Analg. 2010 Jul;111(1):196-203. doi: 10.1213/ANE.0b013e3181cd6e88. Epub 2010 Mar 23.
An anesthesia group staffing agreement with a hospital often specifies the minimum number of operating rooms (ORs) to be covered during evening or weekend hours. Frequently, 1 anesthesiologist works in-house and others take call from home, coming in if multiple cases are waiting to be done. The anesthesiologist in-house sometimes does not bring in colleagues from home when the number of ORs he can supervise is less than the number specified in the agreement with the hospital (i.e., assignment is less than staffing). Queuing occurs even if managers have selected an appropriate number of ORs to be run during evenings and weekends (i.e., the number of cases [jobs] exceeds the number of ORs [identical machines] that are actually run).
In our study, we used a dataset without trend over 3 years to determine whether we could identify differences among anesthesiologists in queuing behavior. Negative binomial regression was used to model the number of cases waiting while the number of ORs running was fewer than allocated. Multiple variable regressions modeled waiting times among cases. The easy to construct Shewhart chart and the more complicated cumulative sum chart were then compared to test for outliers in the number of cases waiting.
Queuing behavior was found to be homogeneous among the 33 anesthesiologists (P > 0.28 among different statistical models). Weekend days differed from evenings. Shewhart and cumulative sum charts were comparable for detecting unusual variations.
Anesthesiologists sometimes do not notify call team members when cases are waiting and the number of ORs running is less than allocated. The number of cases waiting, rather than the amount of waiting, is more appropriate for monitoring trends over time. Simple Shewhart charts can be used for monitoring contractually specified staffing.
医院与麻醉科之间的人员配备协议通常规定了晚间或周末时段需要覆盖的手术室（OR）数量下限。通常情况下，1 名麻醉师在院内工作，其他麻醉师则在家待命，一旦有多个病例需要处理，他们就会前来医院。当院内麻醉师可监督的手术室数量少于与医院协议中规定的数量（即，分配人数少于人员配备人数）时，他有时不会从家中召回同事。即使管理人员选择了在晚间和周末期间运行的适当数量的手术室（即，病例数量[作业]超过实际运行的手术室数量[相同机器]），也会出现排队现象。
在我们的研究中，我们使用了 3 年无趋势的数据集来确定是否可以识别麻醉师在排队行为方面的差异。我们使用负二项回归来对运行的手术室数量少于分配数量时等待的病例数量进行建模。多元变量回归对病例之间的等待时间进行建模。然后，比较易于构建的休哈特图和更复杂的累积和图，以测试等待病例数量中的异常值。
我们发现 33 名麻醉师的排队行为具有同质性（不同统计模型之间 P>0.28）。周末与晚上不同。休哈特图和累积和图在检测异常变化方面具有可比性。
当等待病例数量超过运行的手术室数量，而麻醉师数量低于协议规定的人数时，他们有时不会通知待命的团队成员。与等待时间相比，等待病例的数量更适合监测随时间的趋势。简单的休哈特图可用于监测合同规定的人员配备情况。
