Concussion: Quantitative Electroencephalogram and the Comparison with Established Measures

Burchett, Gabrielle (2021) Concussion: Quantitative Electroencephalogram and the Comparison with Established Measures. Masters thesis, Northern Arizona University.

Text Burchett_2021_concussion_quantitative_electroencephalogram_comparison_.pdf - Published Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

The present study aimed to investigate whether the EEG Concussion Index (CI) is a stronger predictor over established measures of concussion in relationship to a physician diagnosis. A cohort of 9 concussed participants and 15 age- and gender-matched non-concussed control participants from Northern Arizona University were exposed to five concussion assessment tools in a cross-sectional design. The assessment tools range from cognitive to neurological measures: Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT), Vestibular Oculomotor Screening (VOMS), King-Devick (KD), Post-Concussion Symptom Scale (PCSS), and Quantitative Electroencephalogram (QEEG) CI. The data obtained from these assessment tools were examined in a series of logistic regression analyses to determine which measure is the single best predictor of medically diagnosed concussion, and which pair of predictors most accurately predict concussion when compared to other pairs. Due to the obtained sample size, the data for analysis were bootstrapped for 1000 samples. It was hypothesized that 1) QEEG will serve as a stronger predictor for medically diagnosed concussion compared to each established measure individually and that 2) QEEG partnered with the KD will be superior to the other grouped measures of VOMS and KD, VOMS and PCSS, QEEG and PCSS, and ImPACT and QEEG. Both hypotheses were rejected, as the QEEG CI was not a significant predictor of concussion. Results showed that the PCSS was the strongest single predictor of medically diagnosed concussion, B = 7.40, p = .010, 95% CI [2.38, 9.01]. When assessing pairs of predictors, only two pairs yielded -2 log likelihood values that were not equal to .000: the KD and QEEG, explaining between 21% (Cox and Snell R2) and 28% (Nagelkerke R2) of the variance in concussion, and the ImPACT and QEEG which explained between 13% (Cox and Snell R2) and 17% (Nagelkerke R2) of the variance in concussion. However, when these pairs were entered into a test of the full model against a constant-only model, they were not able to distinguish concussed and non-concussed participants, resulting in a non-significant model. With these observed outcomes, no conclusion can be made about which are the best pair. It is important to note that because this was an underpowered and preliminary analysis, these outcomes should be interpreted with caution. It is likely that results may be different in a larger or more severe TBI sample and/or with a different neurological measure.

Item Type:	Thesis (Masters)
Publisher’s Statement:	© Copyright is held by the author. Digital access to this material is made possible by the Cline Library, Northern Arizona University. Further transmission, reproduction or presentation of protected items is prohibited except with permission of the author.
Keywords:	concussion/mTBI; ImPACT; KD; PCSS; QEEG/Quantitative Electroencephalogram; VOMS; Northern Arizona University
Subjects:	R Medicine > RC Internal medicine > RC0321 Neuroscience. Biological psychiatry. Neuropsychiatry
MeSH Subjects:	E Analytical,Diagnostic and Therapeutic Techniques and Equipment > E01 Diagnosis
NAU Depositing Author Academic Status:	Student
Department/Unit:	Graduate College > Theses and Dissertations College of Social and Behavioral Science > Psychological Sciences
Date Deposited:	23 Feb 2022 17:33
Last Modified:	23 Feb 2022 17:33
URI:	https://openknowledge.nau.edu/id/eprint/5736

Actions (login required)

IR Staff Record View

Downloads