Bioanalytics Core

生物分析核心

• 批准号: 9900811
• 负责人: TIMOTHY T HOULE
• 金额: $ 10.35万
• 依托单位: WAKE FOREST UNIVERSITY HEALTH SCIENCES
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至 2022-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9900811
• 关键词: Address; Animals; Data; Data Analyses; Data Display; Data Set; Elements; Ensure; Equation; Goals; Growth; Human; Individual; Individual Differences; Linear Models; Measurement; Measures; Methods; Modeling; Nature; Operative Surgical Procedures; Pain; Pain Measurement; Pattern; Polynomial Models; Predictive Factor; Procedures; Program Research Project Grants; Publications; Recovery; Research; Research Personnel; Specific qualifier value; Statistical Models; Structure; Time; To specify; Visualization; Work; base; data dictionary; data sharing; data visualization; disability; experience; healing; innovation; novel; pain behavior; satisfaction; simulation

Summary The Bioanalytics Core (BC) will support the implementation of a uniform, innovative approach to the analyses of the Program's research Projects. The Core consists of Dr. Timothy Houle and Carol Aschenbrenner, both of whom are experienced biostatisticians with expertise in the proposed methods. The goals of the BC are to evaluate the statistical inferences posed in each of the Projects using the most appropriate statistical model. To accomplish these goals we will utilize several approaches including: growth curve modeling, generalized linear models, generalized estimating equations and other multivariate procedures (e.g., multidimensional scaling). The BC will work closely with the individual investigators of the Projects (PIs: Peters, Martin, Eisenach) to condition the data and address the specific hypotheses inherent in the research. All of the Projects involve studies that repeatedly measure some element of recovery over time after a surgical/experimental insult. Growth curve modeling, sometimes referred to as mixed-effects modeling, or hierarchical linear modeling, will allow us to specify a change trajectory (i.e., healing) that is unique to each individual/animal. The nature of the common form of changes in pain/behavior will be modeled using curvilinear forms (e.g., polynomial regression). Through the use of fixed and random effects, we will then be able to examine the influences on the changes in pain measurements in both the human and animal studies. In this way, we can examine the predictors immediately after surgery, the factors that predict delayed/absent recovery, patterns in these changes across individuals, and examine the relationships between initial measurements and expected change. This approach to the data analysis is highly innovative in this setting, and will extract novel information from these important data.

摘要 生物分析核心(BC)将支持实施统一、创新的方法来 对该计划的研究项目进行分析。核心由蒂莫西·豪尔博士和卡罗尔组成 Aschenbrenner，他们两人都是经验丰富的生物统计学家，拥有拟议方法的专业知识。这个 BC的目标是评估在每个项目中提出的统计推断，使用最 适当的统计模型。为了实现这些目标，我们将使用几种方法，包括：增长 曲线建模、广义线性模型、广义估计方程和其他多变量程序 (例如，多维缩放)。BC将与项目的个人调查员密切合作(PI： 彼得斯、马丁、艾森纳赫)对数据进行条件调整，并解决研究中固有的具体假设。 所有的项目都涉及反复测量在一段时间后恢复的某些元素的研究 外科/实验性侮辱。增长曲线建模，有时称为混合效果建模，或 分层线性建模，将允许我们指定一个变化轨迹(即，修复)，这是每个 个体/动物。疼痛/行为变化的常见形式的性质将使用 曲线形式(例如，多项式回归)。通过使用固定和随机的效果，我们将被 能够在人类和动物研究中检查对疼痛测量变化的影响。在……里面 这样，我们可以立即检查手术后的预测因素，即预测延迟/缺席的因素 恢复，这些变化在个人之间的模式，并检查之间的关系最初 测量结果和预期变化。这种数据分析方法在这种情况下是非常创新的， 并将从这些重要数据中提取新的信息。