Garcia Gian-Gabriel P, Yang Jing, Lavieri Mariel S, McAllister Thomas W, McCrea Michael A, Broglio Steven P
Department of Industrial and Operations Engineering, University of Michigan, Ann Arbor, Michigan.
Department of Psychiatry, Indiana University School of Medicine, Indianapolis, Indiana.
Neurosurgery. 2020 Oct 15;87(5):971-981. doi: 10.1093/neuros/nyaa150.
The Sport Concussion Assessment Tool (SCAT) could be improved by identifying critical subsets that maximize diagnostic accuracy and eliminate low information elements.
To identify optimal SCAT subsets for acute concussion assessment.
Using Concussion Assessment, Research, and Education (CARE) Consortium data, we compared student-athletes' and cadets' preinjury baselines (n = 2178) with postinjury assessments within 6 h (n = 1456) and 24 to 48 h (n = 2394) by considering demographics, symptoms, Standard Assessment of Concussion (SAC), and Balance Error Scoring System (BESS) scores. We divided data into training/testing (60%/40%) sets. Using training data, we integrated logistic regression with an engineering methodology-mixed integer programming-to optimize models with ≤4, 8, 12, and 16 variables (Opt-k). We also created models including only raw scores (Opt-RS-k) and symptom, SAC, and BESS composite scores (summary scores). We evaluated models using testing data.
At <6 h and 24 to 48 h, most Opt-k and Opt-RS-k models included the following symptoms: do not feel right, headache, dizziness, sensitivity to noise, and whether physical or mental activity worsens symptoms. Opt-k models included SAC concentration and delayed recall change scores. Opt-k models had lower Brier scores (BS) and greater area under the curve (AUC) (<6 h: BS = 0.072-0.089, AUC = 0.95-0.96; 24-48 h: BS = 0.085-0.093, AUC = 0.94-0.95) than Opt-RS-k (<6 h: BS = 0.082-0.087, AUC = 0.93-0.95; 24-48 h: BS = 0.095-0.099, AUC = 0.92-0.93) and summary score models (<6 h: BS = 0.14, AUC = 0.89; 24-48 h: BS = 0.15, AUC = 0.87).
We identified SCAT subsets that accurately assess acute concussion and improve administration time over the complete battery, highlighting the importance of eliminating "noisy" elements. These findings can direct clinicians to the SCAT components that are most sensitive to acute concussion.
通过识别能使诊断准确性最大化并消除低信息元素的关键子集，可对运动性脑震荡评估工具（SCAT）进行改进。
确定用于急性脑震荡评估的最佳SCAT子集。
利用脑震荡评估、研究与教育（CARE）联盟的数据，我们通过考虑人口统计学、症状、脑震荡标准评估（SAC）和平衡误差评分系统（BESS）得分，将学生运动员和学员的伤前基线（n = 2178）与伤后6小时内（n = 1456）以及24至48小时（n = 2394）的评估结果进行了比较。我们将数据分为训练/测试（60%/40%）集。利用训练数据，我们将逻辑回归与一种工程方法——混合整数规划相结合，以优化包含≤4、8、12和16个变量的模型（Opt-k）。我们还创建了仅包含原始分数的模型（Opt-RS-k）以及症状、SAC和BESS综合分数（汇总分数）的模型。我们使用测试数据对模型进行评估。
在<6小时以及24至48小时时，大多数Opt-k和Opt-RS-k模型包括以下症状：感觉不对劲、头痛、头晕、对噪音敏感，以及身体或精神活动是否会使症状加重。Opt-k模型包括SAC注意力和延迟回忆变化分数。与Opt-RS-k（<6小时：BS = 0.082 - 0.087，AUC = 0.93 - 0.95；24 - 48小时：BS = 0.095 - 0.099，AUC = 0.92 - 0.93）和汇总分数模型（<6小时：BS = 0.14，AUC = 0.89；24 - 48小时：BS = 0.15，AUC = 0.87）相比，Opt-k模型的布里尔分数（BS）更低，曲线下面积（AUC）更大（<6小时：BS = 0.072 - 0.089，AUC = 0.95 - 0.96；24 - 48小时：BS = 0.085 - 0.093，AUC = 0.94 - 0.95）。
我们确定了能准确评估急性脑震荡并比完整量表缩短施测时间量的SCAT子集，突出了消除“噪声”元素的重要性。这些发现可引导临床医生关注对急性脑震荡最敏感的SCAT组件。
