Concussions are one of the most prevalent traumatic neurologic conditions occurring in children, adolescents, and young adults.1,2 In the United States, an estimated 1.6 to 3.8 million sport-related concussions occur annually.1 During a 3-year study of male and female high school athletes in 10 different contact sports, 5.5% of all injuries were sport-related concussions.2 In a study of collegiate athletes, 5.8% of all injuries were concussions.3 Another study noted that sport-related concussion represented 8.9% of all high school sport-related injuries.4 Sport-related concussions are second only to motor vehicle accidents as a cause of traumatic brain injuries in those aged 15 to 24 years.5
Because there is no gold standard to diagnose concussions, these injuries are often assessed with a variety of clinical measures, focusing on cognition, balance, and symptoms. Evaluation of oculomotor function through cranial nerve assessment is an integral aspect of a thorough clinical examination6; however, recent reports7,8 have emphasized the need for objective assessment of oculomotor function due to the importance of the visual system with respect to vestibular and cranial nerve function. The cerebral structures involved with control of voluntary eye motion are complex but well understood.9 The visual system comprises approximately half of the pathways within the brain,10 and the complex anatomical structures associated with vision are susceptible to injury in an adolescent who sustains a concussion.11 Visual disruption is reported in 65% to 90% of patients who suffer a concussion.9,12 Visual-motor disruptions in these patients include difficulty with saccades, accommodation, smooth pursuit, and fixation.10 Fortunately, disruptions of the visual system can be objectively measured and managed by a health care professional.13 However, objective and validated field-expedient measurements of oculomotor deficits are underused by health care providers.11
The King–Devick test has been used to evaluate the interactions of the visual and vestibular sensory systems following concussion.14,15 This test assesses visual performance as mediated through attention and verbal report, which may both be compromised by a concussion.10 The advantages of the King–Devick test are that health care providers can be quickly trained to discern normal versus abnormal verbal performance deficits secondary to delayed visual motion and it does not require expensive, highly technical equipment.10 Further, this screening tool requires less than 2 minutes to administer and has been shown to be reliable in mixed martial arts fighters who are susceptible to concussion.14
However, as is the case with any assessment that is administered serially, clinicians must be cognizant of practice (or learning) effects, which have been reported in previous studies of the King–Devick test in healthy athletes.14–16 The distinction between practice and learning effects has not been a point of extended discussion in the King–Devick test literature; researchers tend to use these terms interchangeably. It has been suggested that a practice effect occurs when improvement in performance occurs between concurrent testing sessions, whereas a learning effect occurs when improvement in performance is retained over a period of time.17 In the interest of consistency, and with all due respect, the term “practice” effects will be adopted in this article to describe the influence of practice on performance, regardless of the term chosen by the authors of the research cited.
Galetta et al.15 noted a 2.5-second improvement in time to complete the test among healthy athletes who were tested twice in a 15-minute period and a 0.72-second improvement from preseason to postseason testing. More recently, a pilot study of 5 athletes who were tested five times with minimal rest between trials18 showed a practice effect after two trials of the King–Devick test.17 Although these studies demonstrate a practice effect with repeated testing, the study participants were predominantly adult males,14,15,18 limiting generalization of results to younger populations and female athletes. Recent studies recommend using two trials of the King–Devick test for interpretation,14,15,18 but there is insufficient evidence that completion of two trials achieves a plateau for the King–Devick score.18 When a practice effect is identified in physical performance test–retest scenarios in baseline testing, Weir19 advocates that trials be repeated until a plateau is achieved. He also suggests that the first test serve as practice.
The King–Devick test appears to hold promise as a complimentary measure of impairment resulting from concussion. Galetta et al.15 reported excellent test–retest reliability of the King–Devick test in boxers and mixed martial arts fighters with a concussion (intraclass correlation coefficient [ICC]: 0.97; 95% confidence interval [CI]: 0.90 to 1.0). However, the King–Devick test's measurement properties have not been adequately examined in young populations. In particular, the minimal detectable change has not been established for the King–Devick test. Establishing a minimal detectable change for the King–Devick test allows health care professionals to determine if differences in King–Devick composite scores reflect a real and meaningful change. Therefore, the purpose of the current study was to examine the measurement properties of the King–Devick test in a young population, including test–retest reliability, minimal detectable change, and potential practice effects. The sex and age of participants were also examined as potential effect modifiers.
The current study used a prospective research design to evaluate King–Devick test trials in 60 healthy young participants. The analyses presented below are part of a larger study that included postural stability tests that were completed between each of the three trials of the King–Devick test.
For inclusion in the current study, participants were required to be 14 to 24 years old and possess sufficient English language skills to complete all tasks. Exclusion criteria were lower extremity musculoskeletal injuries in the past 3 months; history of a head injury in the past year; or a diagnosis of visual, vestibular, or balance disorders. Participants were also asked if they had ever been diagnosed as having a learning disorder, attention deficit disorder, or dyslexia, and if they had ever participated in sports. An initial telephone screening was used to determine eligibility, and those who met inclusion criteria were given information about the purpose of the research, the potential risk, and the liability and referral procedures in case of injury. Participants provided informed consent or parental permission and child assent, as appropriate. All experimental procedures were approved by the local institutional review boards associated with the study. Participants completed a personal/medical history form before testing to verify eligibility. Potential participants were recruited through flyers distributed to public and private school systems in a metropolitan community.
The King–Devick test is a screening tool that can be administered by medical professionals in less than 2 minutes.14 The test captures impairment of eye movements, attention, and language.10,14,15 Administration of the King–Devick test requires following standardized instructions to read aloud a series of single-digit numbers from left to right on one practice (demonstration) card and three test cards.3,5,16 The summed time to complete the three test cards constitutes the summary (composite) score for the entire test, the King–Devick score. Completion of all three cards is considered one trial. The King–Devick test increases in difficulty from the first test card to the third test card. In the current study, we refer to the test cards, based on sequence, as card 1, card 2, and card 3 for clarity.
Participants were recruited between May and September 2014. All testing was conducted at a university research laboratory by the same investigator and performed at the same location within the laboratory to provide adequate lighting. A comfortable seat was provided. Practice effects were mitigated by using a standard script describing the task, and a single investigator proctored every session. Participants were permitted to use glasses or contacts if they wished. In the current study, the King–Devick test was performed with the participant in a seated position at a self-selected distance for reading the test cards, similar to recent studies.3,5,16 All participants received standardized instructions before performing the King–Devick test, and none of the participants had previously been exposed to the King–Devick test. The study examiner started the stopwatch timer for the King–Devick score when the participant read the first number on card 1, and stopped the timer when the participant completed the last number on card 3. This procedure was repeated three times, constituting three trials for each participant, and the King–Devick score was recorded. As a safety precaution, participants were also asked to rate symptoms of dizziness prior to and throughout the testing session using the University of California–Los Angeles Dizziness Questionnaire 2 (UCLA-DQ2) for perceived intensity of dizziness. The UCLA-DQ2 is a 6-point Likert scale with “0” indicating no dizziness and “5” indicating severe dizziness.
A sample size analysis indicated that 56 participants would be required for a power of at least 0.80 to identify an effect size as small as f = 0.20 (partial η2 = 0.04) for a practice effect across repeated trials and to detect a correlation as low as r = 0.37 between age and King–Devick scores and a difference between sexes (effect size = 0.36); all α = .05, two-tailed. Putting this into context, assuming a standard deviation as high as 3 seconds, a sample size of 56 participants would allow detection of test– retest differences as small as 1.2 seconds. The power analysis was conducted using PASS software version 12 (NCSS Statistical Software, Kaysville, UT).
A linear mixed effects model, with random effects for participants, was used for all analyses. A full factorial model was specified to evaluate all possible interactions to determine if a practice effect was present within difficulty levels (card number) of the King–Devick test and to determine if age and sex affected King–Devick composite and trial scores. Age was defined in yearly increments. An α of .05, two-tailed, was used, and 95% CIs are reported for the outcome measure. Time to complete trials was log-transformed to accommodate a moderate skew. Results are reported as geometric means and geometric standard deviations.
Post-hoc comparisons were conducted using Bonferroni adjustment. An additional analysis was conducted to assess possible differences in practice effect across the three cards. Effect size was calculated for simple comparisons using Cohen's d. A two-way, random effects ICC and the minimal detectable change and its associated standard error of measurement (SEM) were calculated for all three trials of the King–Devick test, for trials 2 and 3 only, and for male and female participants separately. For the ICC, agreement across trials (rather than consistency) was calculated because of the anticipated practice effect (ie, systematic differences across trials were considered potentially relevant). For the current study, an ICC less than 0.40 was interpreted as poor reliability, an ICC between 0.40 and 0.75 indicated moderate to good reliability, and an ICC greater than 0.75 indicated excellent reliability.20 Interpretation of the King–Devick test depends on the composite scores summed across multiple trials, so the ICC for the summed score was calculated. The minimal detectable change and SEM were calculated according to Weir,19 where minimal detectable change = SEM × 1.96 × √2 and SEM = SDDifference / √2. SDDifference is the standard deviation of the difference scores.
SPSS software (version 22.0; IBM, Armonk, NY) was used for the analyses. Although our primary analysis focused on the composite score, we also analyzed the time to complete each of the cards within the three trials.
There were no dropouts or excluded/missing data, so data from 60 participants (30 men, 30 women; mean age = 19.9 ± 3.74 years) were analyzed. None of the participants reported dizziness before, during, or after King–Devick testing as measured by the UCLA-DQ2. The demographic characteristics of the participants are presented in Table 1.
Demographic Characteristics of Participants (N = 60)
Test–Retest Reliability and Minimal Detectable Change
The average interval between trials was approximately 10 minutes. Table 2 presents the ICC, minimal detectable change, and SEM for all three trials of the King–Devick test and the ICC, minimal detectable change, and SEM for trials 2 and 3 only. Test–retest reliability was excellent for all participants across all trials (ICC: 0.95; 95% CI: 0.78 to 0.98) and improved when trial 1 was excluded (ICC: 0.97; 95% CI: 0.91 to 0.99).
ICCs, MDC, and SEM for the King–Devick Test in Participants (N = 60)
The minimal detectable change (SEM) for all participants across all trials was 6.35 ± 2.29 seconds (Table 2). When trial 1 was excluded, the minimal detectable change (SEM) decreased (5.55 ± 2.00 seconds). Minimal detectable change (SEM) was larger for women (6.66 ± 2.40 seconds) than men (3.96 ± 1.43 seconds).
King–Devick Composite Score
Analysis of a full factorial model using the King– Devick composite score as the dependent variable across the three trials (three scores per participant) identified no significant interactions, so all interactions were removed from the model.
Significant main effects were identified for trial (F(2, 175) = 6.19, P < .001), age (F(1, 175) = 22.31, P < .001), and sex (F(1, 175) = 30.94, P < .001). Post-hoc comparisons showed decreases in mean (95% CI) time from trial 1 (45.33 seconds [range: 43.08 to 47.68 seconds]) to trial 2 (41.80 seconds [range: 39.75 to 44.00 seconds]) to trial 3 (range: 40.08 seconds [38.10 to 42.17 seconds]). All pairwise differences were significant (P < .001). For each 1-year increase in age, the mean King–Devick composite score decreased by 2%. Age accounted for 5% of the variability in time to completion. Females (46.01 seconds [range: 42.79 to 49.43 seconds]) required more time than males (39.01 seconds [range: 36.28 to 41.91 seconds]) to complete the test (P < .001).
King–Devick Card Number
The analysis reported above was repeated, but time to complete each card individually was used as the outcome, rather than King–Devick score (summed time to complete all three cards) (ie, card number was added as a factor). Significant effects were found for card number (F(2, 97.4) = 3.42, P = .03), trial (F(2,73.03) = 89.08, P < .001), age (F(1, 64.79) = 11.05, P < .001), and sex (F(1, 64.79) = 14.14, P < .001). Card number × trial was the only interaction that achieved significance (F(4, 2) = 3.52, P < .01).
Table 3 presents the means and 95% CIs for the King–Devick score for each of the three trials by card number (difficulty). The average time to complete decreased 1.16 seconds per card between trials 1 and 2 and 1.73 seconds per card between trials 1 and 3. The tests for trends within trial, across cards, were all significant (P < .001). All pairwise comparisons were significant (P < .004), except for the trial 3 comparison between cards 2 and 3. Table 4 presents the results of pairwise comparisons across trials by card numbers expressed as mean within-participant differences.
King–Devick Times (Seconds) to Complete the 3 Trials of the 3 Cards for Participants (N = 60)
Pairwise Comparisons Within Each Card of the King–Devick Test for Time to Complete (Seconds) of Participants (N = 60)
Our findings suggest that in a young, healthy population, the King–Devick test has excellent test–retest reliability. However, a substantial practice effect was identified across repeated trials, and both age and sex of the participant influence time to completion. Estimates of minimal detectable change and standard error of measurement are provided for this population. A practice effect has been described previously,11,14,15,21 but the influence of age and sex on time to completion has not.
To our knowledge, the current study is the first to evaluate the measurement properties of the King–Devick test in a younger healthy population. Our results suggest that the King–Devick test has excellent test–retest reliability in healthy 14 to 24 year olds, for both the composite score and each of the three trials. Previous studies have reported similar results, with ICC values of 0.90 (95% CI: 0.84 to 0.97) for boxers tested by parents who were not health care professionals,22 0.92 (95% CI, not reported) for professional Finnish ice hockey players tested by health care professionals,18 and 0.97 (95% CI: 0.90 to 1.0) for mixed martial art fighters tested by health care professionals.14 We report that eliminating the first trial—effectively providing an additional practice trial—improves the reliability of the King–Devick composite score.
Minimal Detectable Change
The minimal detectable change for all participants was 6.35 seconds using all three trials of the King– Devick test. For participants in the current study, the minimal detectable change and associated SEM decreased when trial 1 was eliminated from the King– Devick composite score analyses. Males in particular showed a decrease in these indices.
Galetta et al.15 suggested a 5.5-second increase in King–Devick composite scores as indicative of head trauma in their study of mixed martial arts fighters. The authors suggested using 5.5 seconds as a cut-off score to stop play pending medical evaluation for a concussion. In a different study, Galetta et al.14 noted a 5.9-second increase in King–Devick composite scores in male and female collegiate athletes with concussion. Although the time differences noted in Galetta et al.'s14,15 studies are similar, the authors included only a few female athletes in their studies. King et al.21 noted a 6.8-second increase in King–Devick composite time in male rugby players with concussion. Although studies have suggested a 5-second rule,14,15,21 additional evidence is required before this rule is applied to both sexes of all ages. In the current study, the minimal detectable change was 3.96 seconds for males and 6.66 seconds for females when considering trials 2 and 3 only. This difference suggests the 5-second rule may not be universal. Future research should include more females to determine if the difference we identified is replicable.
The participants of the current study demonstrated a practice effect consistent with studies investigating the King–Devick test in samples drawn from various contact sports, including rugby, ice hockey, boxing, mixed martial arts, soccer, football, and basketball.11,14,15,18,21 Unlike our study, the majority of previous studies included only male athletes aged 12 to 53 years.11,14,18,21 The mean age in our study was 19.9 years, and we included male and female participants. The practice effect seen in our participants was consistent across males and females.
Authors have described the learning effect for the King–Devick test as mild,15 slight,23 subtle,18 or moderate, but they rarely refer to a metric when using these terms. In the current study, we found the extent of the practice effect was inversely related to the difficulty of the cards.
In the current study, we noted a 3.53-second improvement from trial 1 to trial 2 and a 5.25-second improvement from trial 1 to trial 3. Similarly, other researchers have found a practice effect using the standard of two testing trials. When comparing baseline scores with scores after fighting matches and including two testing trials, Galetta et al.14 found a 1.9-second improvement in the composite King–Devick score in mixed martial art fighters who did not sustain an observable concussion. In another study, Galetta et al.15 found a 2.8-second improvement in the composite King–Devick score in multiple athletic populations and a median 0.72-second improvement from pre-season to postseason scores after two testing trials. Vartiainen et al.18 found a 2.1-second improvement in professional male hockey players in two trials during a single testing session.
Vartiainen et al.18 reported normative reference values for male hockey players, classifying King–Devick scores from extremely low to superior. They used reference values for two trials for each of the three cards and for the King–Devick composite scores.18 Using the reference values of Vartiainen et al.,18 participants in the current study would be classified as average, with composite scores ranging from 37.7 to 46.2 seconds for trial 1 and from 36.4 to 43.9 seconds for trial 2. Obtaining King–Devick scores to compare across multiple sports for younger and older males and females would be useful for health care professionals if individual baseline scores are not available.
Only one study investigating the King–Devick test reported the ICC of the three levels of difficulty (individual test cards) (ICC: 0.88, 0.88, and 0.83, respectively).18 Vartianien et al.18 also noted that 88% (163 of 185) of their study participants performed the second trial faster than the first trial. In our study, we examined the three individual trials to determine the influence of a practice effect, and we also observed that subsequent trials were completed faster.
Research suggests that participants can improve their test scores on physical performance tests because of practice effects. Clinically, a baseline King–Devick test involves the participant completing a practice test (demonstration) card before completing the three test cards, which increase in level of difficulty. The King– Devick test is usually performed twice, and the average of the two trials is recorded as the athlete's baseline King–Devick composite score. However, Weir19 suggested that practice effects may be reduced by adding trials to the measurement schedule until a performance plateau occurs. This suggestion could be accomplished during baseline testing so the performance plateau is reached. Including an additional trial will add time to baseline testing and this burden may be considerable if many athletes are being tested, but the precision gained may justify the additional effort.
We identified an inverse relationship between the age of participants and time to complete the King–Devick test in the current study. Each 1-year increase in age was associated with a decrease of 0.29 seconds in time to complete, while controlling for sex. In Vartiainen et al.'s18 study of professional male ice hockey players, the authors found that age did not influence King–Devick composite scores. However, participants in this study were considerably older than ours. Nonetheless, the age of the athlete should be considered when interpreting the King–Devick test.
Most of the studies of the King–Devick test to date have been conducted in sports played predominantly by males, so females are not well represented (55 of 518 [10.6%] female athletes).11,14,15,18,21,24 A study by Tjarks et al.11 had the largest number of female participants, representing cheerleading, gymnastics, soccer, and basketball (18 of 35 [51.4%] females). Several studies investigating sex differences in baseline testing for concussions have reported that females performed better than males on verbal memory and perceptual motor speed,25,26 but males performed better than females on visual-spatial tasks, mental rotation, and quantitative problem solving.27,28 In the current study, we found a difference in the King–Devick composite scores based on sex: males had faster completion times than females. To our knowledge, this difference has not been previously reported. Female participation in sports and the incidence of females incurring a concussion have increased over the past decade.29 If the King– Devick test is used as a clinical measure to identify oculomotor dysfunction, then the interpretation of King–Devick scores for females requires further study.
The current study has several limitations. Perhaps most importantly, our test–retest interval was brief, averaging 10 minutes. There is no guarantee that the statistical indices calculated will translate to a much longer interval of weeks or months. Our data were also collected under optimal circumstances, in a controlled laboratory environment. Again, there is no guarantee that our findings will translate into a real life, clinical scenario. The findings presented should be interpreted only as an early step in the validation of the King–Devick test for clinical use with young patients who potentially have concussion. Finally, our participants were not selected to represent an athletic population; therefore, our results may not be generalizable to other athletes.
The results of our study suggested that using trials 2 and 3 improved the reliability of the King–Devick test. Therefore, during baseline testing, we recommend that three trials be performed and that the first trial be considered a practice test to decrease systematic error and provide a more accurate King–Devick composite score. From the trained health care professional's perspective, the additional minute of testing required to meet this recommendation would be acceptable because a better approximation of the athlete's baseline status would be obtained. Future studies should investigate how many trials are needed to establish a performance plateau in multiple athletic populations to determine normative scores and if practice effects differ among different athletic groups.
Implications for Clinical Practice
In the current study, the King–Devick test demonstrated a practice effect with repeated measurements. The three trials of the complete test, as opposed to the two trials commonly reported in the literature, may provide better information for health care professionals. The current study highlights the importance of precise and thorough baseline testing to assist the health care professional with clinical decision-making if there is a subsequent concussion. Further, it suggests that the sex and age of the athlete should be considered if the King–Devick test is used as a concussion screening tool. Males and older participants tend to complete the test more quickly, so normative and baseline values must respect the potential influence of these variables. Translated clinically, this indicates that males and females may not have the same normative population distributions, and that baseline testing should be repeated annually because the King–Devick composite score may decrease as athletes mature.
- Langlois JA, Rutland-Brown W, Wald MM. The epidemiology and impact of traumatic brain injury: a brief overview. J Head Trauma Rehabil. 2006;21:375–378. doi:10.1097/00001199-200609000-00001 [CrossRef]
- Dompier TP, Kerr ZY, Marshall SW, et al. Incidence of concussion during practice and games in youth, high school, and collegiate American football players. JAMA Pediatr. 2015;169:659–665. doi:10.1001/jamapediatrics.2015.0210 [CrossRef]
- Dick R, Ferrara MS, Agel J, et al. Descriptive epidemiology of collegiate men's football injuries: National Collegiate Athletic Association Injury Surveillance System, 1988–1989 through 2003–2004. J Athl Train. 2007;42:221–233.
- Gessel LM, Fields SK, Collins CL, Dick RW, Comstock RD. Concussions among United States high school and collegiate athletes. J Athl Train. 2007;42:495–503.
- Sosin DM, Sniezek JE, Thurman DJ. Incidence of mild and moderate brain injury in the United States, 1991. Brain Inj. 1996;10:47–54. doi:10.1080/026990596124719 [CrossRef]
- Broglio SP, Cantu RC, Gioia GA, et al. National Athletic Trainers' Association Position Statement: management of sport concussion. J Athl Train. 2014;49:245–265. doi:10.4085/1062-6050-49.1.07 [CrossRef]
- Mucha A, Collins MW, Elbin RJ, et al. A brief Vestibular/Ocular Motor Screening (VOMS) assessment to evaluate concussions: preliminary findings. Am J Sports Med. 2014;42:2479–2486. doi:10.1177/0363546514543775 [CrossRef]
- Ventura RE, Balcer LJ, Galetta SL. The neuro-ophthalmology of head trauma. Lancet Neurol. 2014;13:1006–1016. doi:10.1016/S1474-4422(14)70111-5 [CrossRef]
- Heitger MH, Anderson TJ, Jones RD, Dalrymple-Alford JC, Frampton CM, Ardagh MW. Eye movement and visuomotor arm movement deficits following mild closed head injury. Brain. 2004;127:575–590. doi:10.1093/brain/awh066 [CrossRef]
- Ventura RE, Jancuska JM, Balcer LI, Galetta SL. Diagnostic tests for concussion: is vision part of the puzzle?J Neuro-Opthalmol. 2015;35:73–81. doi:10.1097/WNO.0000000000000223 [CrossRef]
- Tjarks BJ, Dorman JC, Valentine VD, et al. Comparison and utility of King–Devick and ImPACT® composite scores in adolescent concussion patients. J Neurol Sci. 2013;334:148–153. doi:10.1016/j.jns.2013.08.015 [CrossRef]
- Heitger MH, Jones RD, Macleod AD, Snell DL, Frampton CM, Anderson TJ. Impaired eye movements in post-concussion syndrome indicate suboptimal brain function beyond the influence of depression, malingering or intellectual ability. Brain. 2009;132:2850–2870. doi:10.1093/brain/awp181 [CrossRef]
- Guskiewicz KM, McCrea M, Marshall SW, et al. Cumulative effects associated with recurrent concussion in collegiate football players: the NCAA Concussion Study. JAMA. 2003;290:2549–2555. doi:10.1001/jama.290.19.2549 [CrossRef]
- Galetta KM, Barrett J, Allen M, et al. The King–Devick test as a determinant of head trauma and concussion in boxers and MMA fighters. Neurology. 2011;76:1456–1462. doi:10.1212/WNL.0b013e31821184c9 [CrossRef]
- Galetta KM, Brandes LE, Maki K, et al. The King-Devick test and sports-related concussion: study of a rapid visual screening tool in a collegiate cohort. J Neurol Sci. 2011;309:34–39. doi:10.1016/j.jns.2011.07.039 [CrossRef]
- Braun PA, Kaminski TW, Swanik CB, Knight CA. Oculomotor function in collegiate student athletes with a previous history of sport-related concussion. Athl Train Sports Health Care. 2013;6:282–288. doi:10.3928/19425864-20131108-01 [CrossRef]
- Valovich McLeod TC, Perrin DH, Guskiewicz KM, Shultz SJ, Diamond R, Gansneder BM. Serial administration of clinical concussion assessments and learning effects in healthy young athletes. Clin J Sport Med. 2004;14:287–295. doi:10.1097/00042752-200409000-00007 [CrossRef]
- Vartiainen MV, Holm A, Peltonen K, Luoto TM, Iverson GL, Hokkanen L. King-Devick test normative reference values for professional male ice hockey players. Scand J Med Sci Sports. 2015;25:e327–e330. doi:10.1111/sms.12307 [CrossRef]
- Weir JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res. 2005;19:231–240.
- Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice, 3rd ed. Upper Saddle River, NJ: Pearson Prentice Hall; 2009.
- King D, Clark T, Gissane C. Use of a rapid visual screening tool for the assessment of concussion in amateur rugby league: a pilot study. J Neurol Sci. 2012;320:16–21. doi:10.1016/j.jns.2012.05.049 [CrossRef]
- Leong DF, Balcer LJ, Galetta SL, Liu Z, Master CL. The King–Devick test as a concussion screening tool administered by sports parents. J Sports Med Phys Fitness. 2014;54:70–77.
- Straume-Naesheim TM, Andersen TE, Bahr R. Reproducibility of computer based neuropsychological testing among Norwegian elite football players. Br J Sports Med. 2005;39(suppl 1):64–69. doi:10.1136/bjsm.2005.019620 [CrossRef]
- Braun PA, Lind KE, Batliner T, et al. Caregiver reported oral health-related quality of life in young American Indian children. J Immigr Minor Health. 2014;16:951–958. doi:10.1007/s10903-013-9870-0 [CrossRef]
- Kimura D, Clarke PG. Women's advantage on verbal memory is not restricted to concrete words. Psychol Rep. 2002;91:1137–1142. doi:10.2466/pr0.2002.91.3f.1137 [CrossRef]
- Covassin T, Swanik CB, Sachs M, et al. Sex differences in baseline neuropsychological function and concussion symptoms of collegiate athletes. Br J Sports Med. 2006;40:923–927. doi:10.1136/bjsm.2006.029496 [CrossRef]
- Covassin T, Elbin RJ, Crutcher B, Burkhart S. The management of sport-related concussion: considerations for male and female athletes. Transl Stroke Res. 2013;4:420–424. doi:10.1007/s12975-012-0228-z [CrossRef]
- Broshek DK, Kaushik T, Freeman JR, Erlanger D, Webbe F, Barth JT. Sex differences in outcome following sports-related concussion. J Neurosurg. 2005;102:856–863. doi:10.3171/jns.2005.102.5.0856 [CrossRef]
- Lincoln AE, Caswell SV, Almquist JL, Dunn RE, Norris JB, Hinton RY. Trends in concussion incidence in high school sports: a prospective 11-year study. Am J Sports Med. 2011;39:958–963. doi:10.1177/0363546510392326 [CrossRef]
Demographic Characteristicsa of Participants (N = 60)
|Age, y||19.9 ± 3.74|
| Male||30 ± 50|
| Female||30 ± 50|
|Glasses/contacts||23 ± 38.3|
|Previous head injury||10 ± 16.6|
|Sports participation||57 ± 95|
ICCs, MDC, and SEM for the King–Devick Test in Participants (N = 60)
|KING-DEVICK TEST COMPOSITE SCORE||ICC (95% CI)||MDC (SEM)|
|All 3 trials|
| All participants (N = 60)||0.95 (0.78 to 0.98)||6.35 (2.29)|
| Males (n = 30)||0.95 (0.83 to 0.98)||6.45 (2.33)|
| Females (n = 30)||0.94 (0.69 to 0.98)||6.18 (2.33)|
|Trials 2 and 3 only|
| All participants (N = 60)||0.97 (0.91 to 0.99)||5.55 (2.00)|
| Males (n = 30)||0.98 (0.95 to 0.99)||3.96 (1.43)|
| Females (n = 30)||0.95 (0.81 to 0.98)||6.66 (2.40)|
King–Devick Times (Seconds) to Complete the 3 Trials of the 3 Cards for Participants (N = 60)a,b
|CARD||TRIAL||MEAN (95% CI)|
|1||1||14.60 (13.86 to 15.37)c|
|2||13.90 (13.20 to 14.64)c|
|3||13.11 (12.44 to 13.80)c|
|2||1||14.85 (14.10 to 15.63)c|
|2||13.72 (13.03 to 14.45)c|
|3||13.14 (12.47 to 13.84)d|
|3||1||15.78 (14.98 to 16.62)c|
|2||14.11 (13.39 to 14.86)c|
|3||13.78 (13.08 to 14.50)c|
Pairwise Comparisons Within Each Card of the King–Devick Test for Time to Complete (Seconds) of Participants (N = 60)
|CARDA||TRIAL (I)B||TRIAL (J)B||DIFFERENCE (I TO J) (95% CI)C||P|
|1||1||2||1.05 (1.02 to 1.08)||.001|
|2||3||1.06 (1.03 to 1.10)||< .001|
|1||3||1.11 (1.07 to 1.16)||< .001|
|2||1||2||1.08 (1.05 to 1.12)||< .001|
|2||3||1.04 (1.01 to 1.08)||.004|
|1||3||1.13 (1.08 to 1.18)||< .001|
|3||1||2||1.12 (1.08 to 1.15)||< .001|
|2||3||1.02 (0.99 to 1.06)||.211|
|1||3||1.15 (1.10 to 1.20)||< .001|