Establishing Test-Retest Reliability and Reliable Change for the King-Devick Test in High School Athletes

Objective: To establish test-retest reliability in nonconcussed high school athletes and compare absolute change, reliable change indices (RCIs), and minimal detectable change (MDC) methods for classifying impairment after sport-related concussion. Design: Prospective, repeated measures. Setting: High schools from the Midwest and Central regions of the United States. Participants: A convenience sample of 772 nonconcussed high school athletes (n = 546 men) completed preseason K-D testing. In addition, 69 athletes completed a second postseason K-D test, and 54 athletes sustained a concussion and completed postconcussion K-D tests. Intervention: K-D test. Main Outcome Measures: Data for age, sex, and concussion history were determined using preseason K-D test outcomes. Test-retest reliability, RCIs, and MDCs were calculated using postseason K-D tests (M = 98.9, SD = 9.1 days). Postinjury K-D assessments within 5 days of injury (M = 1.5, SD = 1.5 days) were used to classify impairment on K-D using absolute change, RCI, and MDC methods. Results: Significant effects for age (P < 0.001) and history of concussion (P = 0.001) were supported on baseline K-D time, with no sex differences (P = 0.21). Preseason to postseason reliability for K-D times was 0.60 (intraclass correlation coefficient, 95% CI, 0.43-0.73), although 38% of athletes exhibited slower postseason K-D times compared with baseline. Impairment on K-D exhibited for 72% of the concussed sample using absolute change, 48% using MDC, and 44% using RCI methods. Conclusions: K-D exhibited moderate test-retest reliability across 1 season. Absolute change yielded the highest sensitivity for preinjury to postinjury impairment on the K-D compared with RCI and MDC methods.