New Models for Joint Analysis of intensive Longitudinal Data and Survival Data

密集纵向数据和生存数据联合分析的新模型

• 批准号: 8311086
• 负责人: RUNZE LI
• 金额: $ 25.25万
• 依托单位: PENNSYLVANIA STATE UNIVERSITY, THE
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/8311086
• 关键词: AIDS/HIV problem; Abstinence; Acquired Immunodeficiency Syndrome; Address; Alcohol or Other Drugs use; Alcohols; Area; Behavior; Calculi; Cellular Phone; Cessation of life; Characteristics; Collection; Complex; Computer software; Data; Data Collection; Data Set; Drops; Drug Addiction; Drug abuse; Drug usage; Ensure; Etiology; Event; Feeling; Food; HIV; HIV Infections; HIV diagnosis; Handheld Computers; Individual; Instruction; Intervention; Investigation; Joints; Knowledge Discovery; Lead; Link; Longitudinal Studies; Marijuana; Mentors; Methodology; Methods; Mind; Modeling; Moods; Natural History; Online Systems; Pattern; Performance; Pharmaceutical Preparations; Plant Roots; Postdoctoral Fellow; Prevention; Prevention Research; Procedures; Process; Psychophysiology; Public Health; Questionnaires; Relapse; Research; Research Methodology; Research Personnel; Resources; Scientist; Smoke; Smoking; Smoking Status; Source; Statistical Models; Stress; Structure; Subgroup; Substance abuse problem; Survival Analysis; Technology Transfer; Testing; Time; Tobacco; Training; Transdisciplinary Tobacco Research Center; Unsafe Sex; Woman; Work; base; data sharing; data structure; drinking; drug abuse prevention; flexibility; graduate student; insight; interest; longitudinal analysis; novel strategies; problem drinker; programs; response; simulation; smoking cessation; software development; technological innovation; theories; user friendly software; web site

PROJECT SUMMARY (See instructions): Intensive longitudinal data are becoming increasingly common in substance use research and HIV/AIDS research. Substance use researchers employ technological innovations such as hand-held computers, cell phones and web-based assessment to collect data in which many observations are collected closely spaced at time. These data collection approaches have revolutionized the process of scientific research and knowledge discovery in the substance use area. Additionally, many HIV/AIDS studies have become mature longitudinal studies that have collected many waves of data and have amassed many observations over the course of years of data collection. Thus, mature longitudinal data with many waves may be viewed as intensive Tongitudinal data as well. In theory, intensive longitudinal data can provide answers to important questions in drug abuse research and HIV/AIDS research. However, there are no appropriate statistical procedures that can be applied for such data to address important scientific questions, such as: What is the relationship between mood and urge to smoke over the period from quitting smoking to the first relapse? What factors predict first instance of unsafe sex after a diagnosis of HIV? In this project, we propose new models for joint analysis of intensive longitudinal data and time-to-event data. These models possess many valuable features which make them the most appropriate to use for testing key hypotheses in drug use research and HIV/AIDS research using intensive longitudinal data. The proposed new models allow effects to vary over time and changing across individual subjects, and keep the structure of random error process very flexible. We further propose estimation procedures for the proposed new models. We plan to develop software to implement the proposed procedure and assess the performance of the proposed procedures by extensive Monte Carlo simulations. We further plan to apply the new procedures we have developed to address important questions in substance use and HIV/AIDS research areas using intensive longitudinal data on HIV/AIDS and on tobacco, alcohol and marijuana.

项目摘要（请参阅说明）： 密集的纵向数据在药物使用研究和艾滋病毒/艾滋病中变得越来越普遍 研究。药物使用研究人员采用技术创新，例如手持计算机，电池 手机和基于Web的评估收集数据，其中收集了许多观察结果 时间。这些数据收集方法彻底改变了科学研究的过程 在物质使用区域中发现的知识发现。此外，许多艾滋病毒/艾滋病研究已经成熟 纵向研究已经收集了许多数据浪潮，并积累了许多观察结果 数据收集的年份。因此，具有许多波的成熟纵向数据可能被视为 密集的通态数据。从理论上讲，密集的纵向数据可以为重要的答案提供答案 药物滥用研究和艾滋病毒/艾滋病研究中的问题。但是，没有适当的统计 可以用于此类数据以解决重要科学问题的程序，例如：什么是 情绪与渴望在戒烟到第一次复发的时期吸烟之间的关系？ 哪些因素可以预测诊断为HIV后不安全性的初次实例？在这个项目中，我们提出了新的 密集纵向数据和事件时间数据的联合分析模型。这些模型拥有许多 有价值的功能使它们最适合于测试药物使用中的关键假设 研究和艾滋病毒/艾滋病研究使用密集的纵向数据。提出的新模型允许效果 随着时间的流逝而变化并在各个受试者之间变化，并保持随机错误过程的结构 非常灵活。我们进一步提出了针对拟议的新模型的估计程序。我们计划发展 软件以实施拟议的程序并评估拟议程序的性能 广泛的蒙特卡洛模拟。我们进一步计划应用我们开发的新程序 使用密集的纵向解决药物使用和艾滋病毒/艾滋病研究领域的重要问题 有关艾滋病毒/艾滋病以及烟草，酒精和大麻的数据。