STATISTICAL METHODOLOGY IN THE NATURAL HISTORY OF AIDS

本提案的目标是制定和评估 研究中使用的统计方法 获得性免疫缺陷综合症(艾滋病)的自然史。一个 来自加州大学伯克利分校的协作团队和 加州大学旧金山分校将参与 项目。有三个主题将特别令人感兴趣。第一个问题与 详细分析利用艾滋病发病率数据的方法 提供对人类免疫缺陷时间分布的估计 病毒(HIV)感染和潜伏期的估计 分发。第二个是关于估计的方法。 与多种病毒相关的病毒的传染性特性 基于接触者数量和感染状况的数据 已知感染者的伴侣。特别值得关注的是 处理包括缺乏信息在内的问题的统计技术 关于任何一方感染的日期，以及哪种模式 受感染者传染性在以下期间的变化 感染到艾滋病的发病。最后，进一步的方法论工作是 建议与孵化期间进化的标记变量有关 部分衡量免疫系统的衰退。统计模型 是以如下方式充分利用这些变量所必需的 临床试验中艾滋病发作的代孕反应与 已知的感染时间，特别是流行的 样本，并了解其在艾滋病发病危险中的作用。 所有方法都将在适当情况下应用于各种数据集。这 该项目的一部分将包括关于建筑的统计工作 标记物及基于随访数据的标记物模型评价 一群受感染的人。

The objective of this proposal is the development and evaluation of statistical methods for use in research related to understanding the natural history of Acquired Immunodeficiency Syndrome (AIDS). A collaborative team from University of California Berkeley, and the University of California, San Francisco, will be involved in the project. Three topics will be of specific interest. The first relates to detailed analysis of methods for the use of data on incidence of AIDS to provide estimates of the distribution of time to Human Immunodeficiency Virus (HIV) infection and estimates of the incubation period distribution. The second is concerned with methodology for estimation of properties of infectivity of the virus associated with various kinds of contacts based on data on the number of contacts and infection status of partners of known infected individuals. Of particular interest are statistical techniques which handle issues including lack of information on the dates of infection of either partner and which elucidate patterns of change in infectiousness of infected individuals over the period from infection to onset of AIDS. Finally, further methodological work is proposed relating to marker variables which evolve during incubation and measure in part the decline of the immune system,. Statistical models are necessary to make full use of these variables in such ways as surrogate responses for onset of AIDS in clinical trials, correlates for times of infection when these are known, particularly for prevalent samples, and in understanding their role in the hazard of onset of AIDS. All methods will be applied to various data sets where appropriate. This part of the project will include statistical work on the construction of markers and evaluation of marker models on data arising from follow-up cohorts of infected individuals.