Predicting protracted recovery from sports-related concussion using computerized neurocognitive test scores and symptom cluster scores
Background: A previous study showed that symptom evaluation, in conjunction with computerized neurocognitive testing, improved predictions of protracted recovery in a group of male high school football players. The determination of prognosis following sports-related concussion can be used to facilitate return-to-play and academic decision making in concussion management. ^ Purpose: The purpose of this study was to replicate and extend previous research in determining the accuracy, sensitivity, specificity, and predictive value of computerized neurocognitive test scores and symptom cluster scores in predicting protracted recovery following sports-related concussion. Study Design: Systematic replication, Cohort study (prognosis) ^ Methods: 30 concussed collegiate athletes were followed clinically using the ImPACT test, and were released to return to play once they were asymptomatic and their neurocognitive performance returned to baseline/normative data. The athletes were retrospectively classified as having a short (≤ 14 days) or protracted (> 14 days) recovery based on the number of days they took to return to play. Discriminant function analysis was used to determine how well neurocognitive test scores and symptom cluster scores can predict group membership, and to identify combinations of variables with the highest classification accuracy. Follow-up phone calls were conducted to determine how clinically accurate current return-to-play standards are in predicting recovery of function. ^ Results: Replication of the Lau et al. (2011) study revealed that computerized neurocognitive test scores and symptom cluster scores classified athletes in the current study into short and protracted recovery groups with 70% accuracy. Two classification models were identified as having the strongest predictive power, one of which resulted in increased classification accuracy compared to the combination of variables used in the Lau et al. study. Follow-up phone calls revealed that 26.7% of athletes experienced symptoms during the return-to-play protocol. ^ Conclusions: This study extended the Lau et al. (2011) study to a different population of athletes. Four variables identified in the Lau et al. study and two variables identified in the current study were used to identify two models for predicting protracted recovery with the highest classification accuracy, which can be used to facilitate clinical decision making in concussion management.^
Sepulveda, Amanda, "Predicting protracted recovery from sports-related concussion using computerized neurocognitive test scores and symptom cluster scores" (2015). ETD Collection for University of Texas, El Paso. AAI1591996.