Statistical Methods for HIV/AIDS Studies

描述（由申请人提供）：在过去的20年里，在艾滋病相关研究方面做出了巨大的努力。结果提供了有用的信息，关于病人的疾病管理基于个体病人的特点。然而，大多数此类“个性化医疗”策略都是在临时基础上获得的。有许多亚组分析无法验证的例子。需要新颖的定量方法来系统地在个人层面上制定合理的治疗策略。例如，平均而言，高效抗逆转录病毒疗法（HAART）是治疗艾滋病毒感染的一种非常有效的疗法。然而，并不是每个人都能从HAART中受益。利用临床试验和观察性研究的数据，建立预测模型以确定未来将受益于这种有效治疗的患者是很重要的。当前研究计划的主要主题之一是系统地开发定量方法，以对未来患者的反应做出有效的推断。提案中有五个研究目标。前三个专门针对个性化医疗问题。最后两个目标是开发新的方法来设计、监测和分析具有事件时间反应的HIV研究，事件时间反应是评估新疗法的常用终点。该计划组建了一个研究小组，在过去的资助期内，该小组在设计、监测和分析与艾滋病毒有关的临床试验和观察性研究方面，在开发理论上合理和实用的定量方法方面非常有成效。PI和共同调查人员也是哈佛大学生物统计学艾滋病研究中心的高级统计学家。他们通过艾滋病临床试验小组（ACTG）和神经学艾滋病研究联盟（NARC）密切参与艾滋病毒研究。他们可以获得过去和当前ACTG和NARC研究的数据，并且可以直接将开发方法的结果应用于正在进行和未来的艾滋病毒研究。公共卫生相关性：将开发定量方法，利用临床试验和观察性研究的数据，为患者管理艾滋病毒疾病的个性化医疗方法。还将开发新的统计程序，用于设计、监测和分析评估艾滋病毒感染新疗法的临床试验。

DESCRIPTION (provided by applicant): Great efforts have been made over the past 20 years in HIV-related research. The results have provided useful information regarding the patient's disease management based on the individual patient characteristics. However, most of such "personalized medicine" strategies were obtained on an ad hoc basis. There are numerous examples of subgroup analyses that could not be validated. Novel quantitative methods are needed to systematically develop sound treatment strategies at a personal level. For example, on average, the highly active antiretroviral therapy (HAART) is a very potent therapy for treating HIV-infection. However, not everyone benefits from HAART. Using data from clinical trials and observational studies, it is important to establish prediction models that can identify future patients who would benefit from such potent therapies. One of the main themes for this current research plan is to systematically develop quantitative methods to make valid inferences for future patient responses. There are five research aims in the proposal. The first three specifically address the personalized medicine issues. The last two aims are to develop new methods for designing, monitoring, and analyzing HIV studies with event time responses, which are commonly used endpoints for evaluating new therapies. The PI has assembled a research team which has been very productive in developing theoretically sound and practical quantitative methods for designing, monitoring and analyzing HIV- related clinical trials and observational studies over the past funding periods. The PI and co-investigators are also senior statisticians for the Center for Biostatical AIDS Research Center at Harvard University. They are intimately involved in HIV studies via the AIDS Clinical Trails Group (ACTG) and Neurologic AIDS research Consortium (NARC). They have access to the data from past and current ACTG and NARC studies, and can directly apply the results from developed methodologies to the on-going and future HIV studies. PUBLIC HEALTH RELEVANCE: Quantitative methods will be developed for personalized medicine approaches for patient management of HIV disease using the data from clinical trials and observational studies. New statistical procedures will also be developed for designing, monitoring, and analyzing clinical trials for evaluating new therapies for HIV-infection.