1. Comparison of baseline-referenced versus norm-referenced analytical approaches for in-theatre assessment of mild traumatic brain injury neurocognitive impairment.
- Author
-
Haran, F. Jay, Dretsch, Michael N., Slaboda, Jill C., Johnson, Dagny E., Adam, Octavian R., and Tsao, Jack W.
- Subjects
BRAIN injury diagnosis ,BRAIN injuries ,CHI-squared test ,COGNITION ,DATABASES ,DEPLOYMENT (Military strategy) ,DISCRIMINANT analysis ,EMPLOYMENT reentry ,MEDICAL information storage & retrieval systems ,NEUROPSYCHOLOGICAL tests ,RESEARCH methodology ,MILITARY hospitals ,SCIENTIFIC observation ,PROBABILITY theory ,QUESTIONNAIRES ,REACTION time ,RESEARCH funding ,SELF-evaluation ,MILITARY personnel ,STATISTICS ,T-test (Statistics) ,DATA analysis ,MILITARY service ,RETROSPECTIVE studies ,RECEIVER operating characteristic curves ,RESEARCH methodology evaluation ,ACUTE diseases ,DATA analysis software ,DESCRIPTIVE statistics ,COMPUTER-aided diagnosis ,SYMPTOMS - Abstract
Primary objective: To examine differences between thebaseline-referencedandnorm-referencedapproaches for determining decrements in Automated Neuropsychological Assessment Metrics Version 4 TBI-MIL (ANAM) performance following mild traumatic brain injury (mTBI). Research design: ANAM data were reviewed for 616 US Service members, with 528 of this sample having experienced an mTBI and 88 were controls. Methods and procedures: Post-injury change scores were calculated for each sub-test: (1) normative change score = in-theater score – normative mean and (2) baseline change score = in-theater score – pre-deployment baseline. Reliable change cut-scores were applied to the change and the resulting frequency distributions were compared using McNemar tests. Receiver operator curves (ROC) using both samples (i.e. mTBI and control) were calculated for the change scores for each approach to determine the discriminate ability of the ANAM. Main outcomes and results: There were no statistical differences,p< 0.05 (Bonferonni-Holm corrected), between the approaches. When the area under the curve for the ROCs were averaged across sub-tests, there were no significant differences between either thenorm-referenced(0.65) or baseline-referenced (0.66) approaches,p> 0.05. Conclusions: Overall, the findings suggest there is no clear advantage of using thebaseline-referencedapproach overnorm-referencedapproach. [ABSTRACT FROM AUTHOR]
- Published
- 2016
- Full Text
- View/download PDF