Craig Hospital, Englewood, CO 80132, USA.
Arch Phys Med Rehabil. 2013 Mar;94(3):579-88. doi: 10.1016/j.apmr.2012.08.197. Epub 2012 Aug 16.
Establishing accurate mathematical models of outcome measures is essential in understanding change throughout the rehabilitation process. The goal of this study is to identify the best-fitting descriptive models for a set of commonly adopted outcome measures found within the Traumatic Brain Injury Model Systems National Database where the modeling is based on data submission through 2011 and the complete range of recorded time points since injury for each individual, where time points range from admission to rehabilitation to 20 years postinjury. The statistical methodology and the application of the methodology contained herein may be used to assist researchers and clinicians in (1) modeling the outcome measures considered, (2) modeling various portions of these outcomes by stratification and/or truncating time periods, (3) modeling longitudinal outcome measures not considered, and (4) establishing models as a necessary precursor in conducting individual growth curve analysis.
建立准确的结果测量数学模型对于理解康复过程中的变化至关重要。本研究的目的是确定在创伤性脑损伤模型系统国家数据库中常用的一组结果测量的最佳拟合描述性模型,这些模型是基于 2011 年之前的数据提交以及每个个体受伤后完整的记录时间点建模的,时间点从入院到康复到受伤后 20 年不等。本研究中包含的统计方法和方法的应用可用于帮助研究人员和临床医生:(1)对所考虑的结果测量进行建模;(2)通过分层和/或截断时间段对这些结果的各个部分进行建模;(3)对未考虑的纵向结果测量进行建模;(4)建立模型作为进行个体生长曲线分析的必要前提。