Presentation at the American Medical Society for Sports Medicine 2017 Annual Meeting

Relationship between the King-Devick Test (KDT) and Commonly Used Concussion Tests

Authors: Clugston JR, Kontos AP, Houck ZM, Asken BM, Boone JK, Hoffman NL, Chrisman SP, Buckley TA, Schmidt JD, Kaminski TW, Collins MW, Ortega JD
Publication: Presentation at the American Medical Society for Sports Medicine 2017 Annual Meeting
Date: April 25, 2017
Tags:

Purpose: To explore relationships between KDT and other concussion tests at baseline.

Methods and Study Design: Baseline concussion test scores were obtained from the Concussion Awareness, Research, and Education (CARE) Consortium for the KDT (mean=41.1seconds, standard-deviation(SD)=6.9, range=24.0-112.0); SCAT3 Symptom Evaluation [total, severity]; Brief Symptom Inventory-18 [anxiety, depression, somatization]; Balance Error Scoring System [firm-surface, foam-surface]; Standardized Assessment of Concussion (SAC) [orientation, immediate-memory, concentration, delayed-memory]; Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) [verbal-memory, visual-memory, visual-motor speed, reaction-time]; and the Vestibular-Ocular Motor Screening tool [Δ-score, dichotomized>2, mean-convergence]. Data were analyzed from athletes’ first year in CARE (n=2,331, mean-age=20.0yrs, SD=1.5, range=17.3-30.1). Fifty-five subjects were excluded as outliers on KDT and ImPACT. Pearson bivariate correlations were used to assess relationships between KDT and the other tests. SAC-immediate-memory (r=-.069, p=.002), SAC-concentration (r=-.219, p<.001), ImPACT-verbal memory (r=-.049, p=.028), ImPACT-visual-motor speed (r=-.271, p<.001) and ImPACT-reaction-time (r=.169, p<.001) significantly correlated with baseline KDT times. These were used as predictors in a hierarchical regression model (n=1,962).

Results: The regression model explained 10.5% (r2=.105, small effect size) of the variance in KDT baseline times with SAC-concentration (β=-.168, p<.001), ImPACT-visual-motor speed (β=-.221, p<.001), ImPACT-reaction-time (β=.054, p=.029) and ImPACT-verbal-memory (β=.056, p=.014) being significant predictors of baseline KDT times.

Conclusions: Better baseline SAC-concentration, ImPACT-visual-motor speed and ImPACT-reaction-time predicted better baseline KDT times. Conversely, better ImPACT-verbal-memory predicted worse KDT times. However, the effect sizes for these relationships were small.

Significance of Findings: Results demonstrate small associations at baseline between KDT and other concussion tests, but significant variance in KDT performance (almost 90%) was not explained by any of these associations, suggesting KDT may contribute unique information to baseline concussion assessments.

Summary Points:

  • K-D Test performance was significantly correlated with SAC concentration, ImPACT visual motor speed and ImPACT reaction time.
  • Regression models only explained 10% of the variability in KD baselines. Therefore, findings suggest that K-D measures a construct not related to other commonly used concussion assessments at baseline.
  • K-D Test should be used in conjunction with other measures as part of a comprehensive assessment battery.