Takizawa T, Ito T, Kosuge M
Natl Inst Anim Health Q (Tokyo). 1977 Winter;17(4):171-8.
Based on the Reed-Frost model (Model I), the authors conducted computer simulation of an epizootic model (Model II) constructed on the assumption that any infected animal in a group, after a given time-period of infectivity, would be removed from the group at the beginning of the next time-period. Models I and II were simulated 100 times for each of the different conditions, viz. the initial size of group, 100 and 1,000, the five steps of contact rate or contact size, and the five more steps of contact rate for the group of 1,000 animals in Model I. From the results obtained, it is believed that as a constant parameter, contact size may be preferably used instead of contact rate in these models. Model II mostly gave higher morbidities than Model I, and earlier termination of epizootics, except the simulation with the smallest contact size. This fact may be due to the effect of herd immunity involved only in Model I. The long duration of epizootic was demonstrated in two of the 100 simulations of Model II with 1,000 individuals and contact size 1. This is characteristic of probabilistic models which are really instructive to studying the flow of epizootic.
基于里德 - 弗罗斯特模型（模型I），作者对一个动物流行病模型（模型II）进行了计算机模拟，该模型的构建假设是，一组中的任何受感染动物在给定的传染期后，会在下一个时间段开始时从该组中移除。模型I和模型II针对不同条件各模拟了100次，即组的初始规模为100和1000、接触率或接触规模的五个级别，以及模型I中1000只动物组的接触率的另外五个级别。从获得的结果来看，人们认为作为一个常数参数，在这些模型中接触规模可能比接触率更适合使用。除了接触规模最小的模拟外，模型II的发病率大多高于模型I，并且动物流行病的终止更早。这一事实可能是由于仅在模型I中涉及的群体免疫效应。在模型II针对1000个个体且接触规模为1的100次模拟中有两次显示出动物流行病的持续时间很长。这是概率模型的特点，对研究动物流行病的传播很有指导意义。
