Blythe S P, Anderson R M
Department of Pure and Applied Biology, Imperial College, London University, UK.
IMA J Math Appl Med Biol. 1988;5(4):237-60. doi: 10.1093/imammb/5.4.237.
A proportionate mixing one-sex model of sexual transmission of HIV is described, in which sexual activity (new partners per unit time) is defined as a continuous variable in a set of integro-partial-differential equations. A discrete activity-class approximation is developed by matching equilibrium state and rate variables as closely as possible with the continuous-variable model, and consists only of ordinary differential equations. Activity-class boundaries are arbitrary, and each class is characterized by a single level of activity. If there are N classes, the level of activity in N - 1 of them is such that the steady-state susceptible class sub-population is equal to the population in the equivalent section of the continuous model. The activity level for the remaining class is chosen such that the condition for endemicity of the infection in the approximation is equal to that for the equivalent continuous-variable model; this minimizes errors in the steady-state population. The relationship between the discrete and continuous-variable models is explored, via numerical and analytical studies, in order to evaluate the accuracy of the approximation.
描述了一种成比例混合的艾滋病病毒(HIV)性传播单性别模型,其中性活动(单位时间内的新伴侣数量)在一组积分偏微分方程中被定义为一个连续变量。通过尽可能紧密地将平衡状态和速率变量与连续变量模型相匹配,开发了一种离散活动类别近似法,它仅由常微分方程组成。活动类别边界是任意的,并且每个类别由单一活动水平表征。如果有N个类别,其中N - 1个类别的活动水平使得稳态易感类别亚群体等于连续模型等效部分中的群体。选择其余类别的活动水平,使得近似中感染流行的条件等于等效连续变量模型的条件;这使稳态群体中的误差最小化。通过数值和分析研究探索离散和连续变量模型之间的关系,以评估近似的准确性。