Delon G L, Smalley H E
Health Serv Res. 1970 Fall;5(3):196-209.
A stochastic model is developed to describe the traffic among ten hospital departments or functional centers that generate 75 percent of all personnel trips to and from two nursing units with a total of 120 beds. The effects of five factors on travel frequencies are investigated by analysis of variance, as are the interactions among the factors. Multiple regression and correlation analyses are used to test the feasibility of predicting travel frequencies from the patient census in three levels of care and from the size of the nursing unit staff, which are found not to be very useful predictors. A Poisson distribution is found to be a good descriptor of the observed frequency distributions for seven of the ten departments studied.
开发了一个随机模型来描述十个医院科室或功能中心之间的交通流量，这些科室或功能中心产生了往返两个护理单元的所有人员出行的75%，这两个护理单元共有120张床位。通过方差分析研究了五个因素对出行频率的影响，以及这些因素之间的相互作用。使用多元回归和相关分析来测试根据三级护理中的患者普查和护理单元工作人员规模预测出行频率的可行性，结果发现这些并不是非常有用的预测指标。对于所研究的十个科室中的七个科室，发现泊松分布是观察到的频率分布的良好描述。
