Multimodal Assessment of Sport-Related Concussion.

Abstract

The purpose of this study was to determine which assessments best identify athletes with sport-related concussion (SRC) from healthy controls in the acute/early subacute phase (within 10 days of SRC) of injury. Prospective, cohort study. Specialty concussion clinic. Sixty-four athletes with SRC (52% male) and 59 matched (age and sex), healthy controls (56% male) aged 12 to 20 years (Mean [M] = 15.07, Standard Deviation [SD] = 2.23). Participants completed symptom, cognitive, vestibular/oculomotor, near point of convergence (NPC), and balance assessments. Univariate analyses were conducted to compare athletes with SRC to healthy controls across all assessments. Assessments that significantly differed between the SRC group and healthy controls were used as predictors in an enter method logistic regression (LR) model and subsequent forward stepwise LR. Results of LR analyses indicated that symptom inventory and symptom provocation on vestibular/oculomotor assessments significantly predicted athletes with SRC versus controls. The forward stepwise LR accurately classified 84.6% of the overall sample (78.3% of athletes with SRC and 91.2% of controls were accurately predicted) and accounted for 60.5% of the variance in predicting athletes with SRC versus controls. Total symptom inventory score (P = 0.003) and vestibular/oculomotor symptom provocation (P < 0.01) were the most sensitive and specific measures in a comprehensive, multimodal assessment for distinguishing athletes with SRC from healthy controls within 10 days of injury. Elements within a multimodal evaluation that are the most robust at discriminating athletes with SRC from healthy controls in the acute/early subacute phase of injury include symptom report and provocation of symptoms on vestibular/oculomotor assessment. These assessments should be considered in conjunction with other objective assessments (ie, NPC measurement and cognitive testing) as part of a comprehensive evaluation of SRC.