Wilmoth Kristin, Brett Benjamin L, Emmert Natalie A, Cook Carolyn M, Schaffert Jeffrey, Caze Todd, Kotsonis Thomas, Cusick Margaret, Solomon Gary, Resch Jacob E, Cullum C Munro, Nelson Lindsay D, McCrea Michael
Departments of Psychiatry and Physical Medicine & Rehabilitation, UT Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX, 75390-9055, USA.
Departments of Neurosurgery and Neurology, Medical College of Wisconsin, Milwaukee, WI, USA.
Neuropsychol Rev. 2023 Dec;33(4):675-692. doi: 10.1007/s11065-022-09553-4. Epub 2022 Aug 30.
Athletic programs are more frequently turning to computerized cognitive tools in order to increase efficiencies in concussion assessment. However, assessment using a traditional neuropsychological test battery may provide a more comprehensive and individualized evaluation. Our goal was to inform sport clinicians of the best practices for concussion assessment through a systematic literature review describing the psychometric properties of standard neuropsychological tests and computerized tools. We conducted our search in relevant databases including Ovid Medline, Web of Science, PsycINFO, and Scopus. Journal articles were included if they evaluated psychometric properties (e.g., reliability, sensitivity) of a cognitive assessment within pure athlete samples (up to 30 days post-injury). Searches yielded 4,758 unique results. Ultimately, 103 articles met inclusion criteria, all of which focused on adolescent or young adult participants. Test-retest reliability estimates ranged from .14 to .93 for computerized tools and .02 to .95 for standard neuropsychological tests, with strongest correlations on processing speed tasks for both modalities, although processing speed tasks were most susceptible to practice effects. Reliability was improved with a 2-factor model (processing speed and memory) and by aggregating multiple baseline exams, yet remained below acceptable limits for some studies. Sensitivity to decreased cognitive performance within 72 h of injury ranged from 45%-93% for computerized tools and 18%-80% for standard neuropsychological test batteries. The method for classifying cognitive decline (normative comparison, reliable change indices, regression-based methods) affected sensitivity estimates. Combining computerized tools and standard neuropsychological tests with the strongest psychometric performance provides the greatest value in clinical assessment. To this end, future studies should evaluate the efficacy of hybrid test batteries comprised of top-performing measures from both modalities.
体育项目越来越频繁地采用计算机化认知工具，以提高脑震荡评估的效率。然而，使用传统神经心理测试组合进行评估可能会提供更全面、个性化的评价。我们的目标是通过系统的文献综述，向体育临床医生介绍脑震荡评估的最佳实践，该综述描述了标准神经心理测试和计算机化工具的心理测量特性。我们在包括Ovid Medline、Web of Science、PsycINFO和Scopus在内的相关数据库中进行了检索。如果期刊文章评估了纯运动员样本（受伤后30天内）认知评估的心理测量特性（如信度、敏感性），则将其纳入。检索共得到4758个独特结果。最终，103篇文章符合纳入标准，所有这些文章都聚焦于青少年或年轻成年参与者。计算机化工具的重测信度估计值在0.14至0.93之间，标准神经心理测试的重测信度估计值在0.02至0.95之间，两种方式在处理速度任务上的相关性最强，尽管处理速度任务最容易受到练习效应的影响。采用双因素模型（处理速度和记忆）以及汇总多个基线检查可提高信度，但在一些研究中仍低于可接受限度。受伤后72小时内对认知能力下降的敏感性，计算机化工具为45%-93%，标准神经心理测试组合为18%-80%。认知能力下降的分类方法（常模比较、可靠变化指数、基于回归的方法）影响敏感性估计。在临床评估中，将计算机化工具和心理测量性能最强的标准神经心理测试相结合能提供最大价值。为此，未来的研究应评估由两种方式中表现最佳的测量方法组成的混合测试组合的有效性。
