Examination of the Test-Retest Reliability of a Computerized Neurocognitive Test Battery.

BACKGROUND: Test-retest reliability is a critical issue in the utility of computer-based neurocognitive assessment paradigms employing baseline and postconcussion tests. Researchers have reported low test-retest reliability for the Immediate Post Concussion Assessment and Cognitive Testing (ImPACT) across an interval of 45 and 50 days. PURPOSE: To re-examine the test-retest reliability of the ImPACT between baseline, 45 days, and 50 days. STUDY DESIGN: Descriptive laboratory study. METHODS: Eighty-five physically active college students (51 male, 34 female) volunteered for this study. Participants completed the ImPACT as well as a 15-item memory test at baseline, 45 days, and 50 days. Intraclass correlation coefficients (ICCs) were calculated for ImPACT composite scores, and change scores were calculated using reliable change indices (RCIs) and regression-based methods (RBMs) at 80% and 95% confidence intervals (CIs). RESULTS: The respective ICCs for baseline to day 45, day 45 to day 50, baseline to day 50, and overall were as follows: verbal memory (0.76, 0.69, 0.65, and 0.78), visual memory (0.72, 0.66, 0.60, and 0.74), visual motor (processing) speed (0.87, 0.88, 0.85, and 0.91), and reaction time (0.67, 0.81, 0.71, and 0.80). All ICCs exceeded the threshold value of 0.60 for acceptable test-retest reliability. All cases fell well within the 80% CI for both the RCI and RBM, while 1% to 5% of cases fell outside the 95% CI for the RCI and 1% for the RBM. CONCLUSION: Results suggest that the ImPACT is a reliable neurocognitive test battery at 45 and 50 days after the baseline assessment. The current findings agree with those of other reliability studies that have reported acceptable ICCs across 30-day to 1-year testing intervals, and they support the utility of the ImPACT for the multidisciplinary approach to concussion management. CLINICAL RELEVANCE: This study suggests that the computerized neurocognitive test battery, ImPACT, is a reliable test for postconcussion serial assessments. However, when managing concussed athletes, the ImPACT should not be used as a stand-alone measure.

Injury patterns in big ten conference football.

BACKGROUND: In 1998, the National Collegiate Athletic Association legislated a decrease in the number of scrimmages and full-contact practices allowed during the spring season. HYPOTHESIS: A significantly increased risk of injury faced by the same player during the spring versus fall practice sessions does exist, but the National Collegiate Athletic Association regulations will not have the intended effect of reducing injury rates to equal or below those of the fall practice sessions. STUDY DESIGN: Retrospectively analyzed, descriptive study of prospectively collected injury surveillance data followed by a prospective, controlled, repeated-measures study after the rule change. METHODS: Phase 1: the Big Ten Conference Sports Injury Surveillance System database was reviewed from spring 1992 through fall 1997 for all reportable injuries. Phase 2: a prospective investigation was instituted from spring 1998 through fall 2000. This protocol was modified to also document updated individual player position descriptions and string as well as exposures for the old fall and spring practice categories (scrimmages, full contact, and limited contact) and the new spring practice designations (full pads with and without tackling, helmets only, spring game, and other scrimmages). RESULTS: Phase 1: There were 3950 fall injuries and 1007 spring injuries, with 469 of the 1007 attributable to the "spring risk factor." There was a statistically significant increase in injury rate in the spring (19.8) versus fall (10.6). Scrimmages (incidence density ratio = 2.4) and limited-contact practices (incidence density ratio = 2.5) showed more than twice the fall injury rate. Phase 2: There were 1502 fall injuries and 648 spring injuries, with 192 attributable to the spring risk factor. There was a 3-fold increase in injury rate in the spring (incidence density ratio = 3.2). Although the noncontact, helmets-only practices produced the lowest injury rates, the nonspring game scrimmages and the limited-contact practices defined as "practice with pads but without tackling" displayed the highest injury rates. In both spring and fall, nonplayers exhibited the highest injury rates for all string groups. CONCLUSIONS: There was and still is a statistically significant increase in spring practice injury risk. The 1998 rule change resulted in an even greater increase in spring practice injury risk. If the goal is to minimize the number of spring practice injuries, it will be best accomplished by limiting the number of scrimmages and limited-contact sessions.