STATISTICAL METHODS IN AIDS RESEARCH

说明(申请人的摘要)这笔赠款的总体目标是 发展和应用统计学方法在艾滋病研究中的应用。重点放在 分析纵向研究的统计方法的发展 艾滋病毒疾病以改进艾滋病毒流行病学研究的科学推断， 预防和治疗。 这项提议有三个具体目标。第一个具体目标是 开发分析受信息性约束的数据的方法 纵向截尾或更一般的信息性随机粗化 艾滋病毒疾病与事件发生时间数据的研究。右审查，间隔 被审查的和当前状态数据都是事件间隔时间过长的例子 数据。这项工作将适用于艾滋病毒预防、治疗的研究 以及估计艾滋病毒发病率、相对风险、疫苗的流行病学 和治疗效果。具体目标1项下的工作将适用于 印度性病门诊高危HIV阴性患者的研究， 巴尔的摩的针头交换评估，以及 美国(活着)。第二个具体目标是发展统计方法 用于研究母婴传播。这项工作将包括回归 估计母婴传播率的模型 对于随访的损失和婴儿死亡率的差异，以及 追踪母体血清和乳房病毒载量的纵向数据方法 并将这些负荷量与母婴风险相关联 变速箱。具体目标2项下的工作将应用于海地的研究 还有津巴布韦。第三个具体目标涉及统计学的发展 重复住院等复发事件的分析方法 和感染。具体目标3下的工作将需要制定 经常性事件数据的单样本和回归模型 信息审查问题在估计疾病发生率中的应用 随时间重复发生的事件和回归模型。这些方法将被应用于 活着的研究。

DESCRIPTION (The applicant's abstract) The overall objective of this grant is to develop and apply statistical methods in AIDS research. The focus is on the development of statistical methods for the analysis of longitudinal studies of HIV disease to improve scientific inferences from studies of HIV epidemiology, prevention and treatment. There are three specific aims of this proposal. The first specific aim is the development of methods for the analysis of data that are subject to informative censoring or more generally informative stochastic coarsening in longitudinal studies of HIV disease with time to event data. Right censored, interval censored and current status data are all examples of coarsened time to event data. This work will have applicability to studies of HIV prevention, treatment and epidemiology for estimation of HIV incidence rates, relative risks, vaccine and treatment efficacy. The work under specific aim 1 will be applied to a study of high risk HIV-negative individuals attending STD clinics in India, needle exchange evaluations in Baltimore, and a cohort of IV drug users in the U.S. (ALIVE). The second specific aim is the development of statistical methods for studies of maternal-infant transmission. This work will include regression models to estimate maternal-infant transmission rates that adequately account for loss to follow-up and differential infant mortality, as well as longitudinal data methods for tracking viral loads in maternal sera and breast milk and for correlating these loads with risks of maternal-infant transmission. The work under specific aim 2 will be applied to studies in Haiti and Zimbabwe. The third specific aim concerns the development of statistical methods for the analysis of recurrent events such as repeat hospitalizations and infections. The work under specific aim 3 will entail the development of one-sample and regression models for recurrent event data that address the issue of informative censoring for estimation of the occurrence rate of recurrent events over time and regression models. The methods will be applied to the ALIVE study.