The effects of fatigue and injury on the King-Devick test across a field hockey season
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The King-Devick test, a relatively new vision-based assessment, has recently been incorporated into the concussion battery. However, the determinants of King-Devick testing have received limited attention in the literature thereby limiting the test’s evidence-based utilization by clinicians. The purpose of this study was to evaluate potential determinants of the King-Devick test under ecologically valid conditions of collegiate student-athletes. This study was prospective longitudinal, taking place in the University of Delaware Athletic Training Room and Field Hockey field. The participants were 16 collegiate Division I field hockey student-athletes (Age: 19.5±1.0 years, Height: 165.0±6.4 cm, Weight: 61.8±5.1 kg, Previous concussions: 0.8±1.1). The King-Devick test was administered on six occasions beginning at the start of the spring 2015 season. The first testing session (T1) was administered in February and served as the overall baseline test and was the first exposure to King-Devick for all participants. The second testing session (T2) was administered at the mid-point of the season immediately after a normal practice session. The third testing session (T3) was administered post-season, within one week of the season ending. The same procedure was followed for the fall season, with a baseline test prior to the start of the season (T4), mid-season after-practice session (T5) and a post season test (T6). For repeat administration, a repeated measures ANOVA found a significant main effect for time (P= .018). No significant differences ix were found between T1 and T2, T3, and T4. However, a simple contrast post-hoc test revealed a significant difference between baseline and T5 (P=.018) and T6 (P=.001). A multiple regression analysis measuring acute fatigue found no significant correlations between either measure of acute fatigue and K-D performance at T2 (P = .688) and T5 (P = .554). Post priori, participants were classified as false-positives in three manners; 1) any worsening from baseline (T1), 2) greater than or equal to 3 seconds worse than T1, and 3) greater than or equal to 5 seconds worse than T1 as all approaches have been utilized in the literature. Overall regarding false positives, 35.4% (28/79) of subsequent tests were slower than baseline, 17.7% (14/79) were at least 3 seconds slower than baseline, and 11.4% (9/79) were at least 5 seconds slower than baseline which were misclassified as impaired. Therefore, a 5th repeat administration of the King Devick test within a calendar year demonstrated a significant practice effect with a moderate to large effect size (0.7). Further, an exertional practice was also not associated with any changes in King-Devick test performance. However, between 31.3-68.8% of participants were misclassified as impaired on the King-Devick test despite no evidence to suspect a concussive injury. These results suggest concern in regards to specificity of the King-Devick test suggesting the possibility of false positive test results. This reiterates that a concussion is a clinical diagnosis made by a healthcare provider and supported by a multifaceted testing battery.