Leveraging the Enhanced Sports as a Laboratory Assessment Model to Cross-Validate a Formula Used to Predict Clinical Recovery From Concussion in Collegiate Athletes.

Context: The Sports as a Laboratory Assessment Model (SLAM+) allows for point-of-care data collection to improve outcomes after concussion and assist with clinical research. Objective: To cross-validate a prognostic formula to predict clinical recovery in collegiate athletes with concussion leveraging the SLAM+. Design: Cross-sectional Study. Patients or Other Participants: Collegiate athletes (n = 133 [42.9% female]) who were, on average, 20.4 ± 1.57 years of age. Main Outcome Measures: Collegiate athletes with concussion were administered the Sport Concussion Assessment Tool-5 (SCAT-5) Symptom Evaluation and the Revised Head Injury Scale (HIS-r) within 48 hours of injury. The HIS-r was administered daily until a participant reported symptom free. The SCAT-5's total symptom severity score and the HIS-r's duration of drowsiness, neck pain, nervousness, and tingling were used to predict days until clinical recovery. Spearman rho (ϱ) was calculated between predicted and actual days until clinical recovery. Sensitivity, specificity, positive, and negative predictive values were calculated to determine the formula's ability to correctly classify participants who achieved clinical recovery within 7 (n = 57), 8-14 (n = 45), 15-21 (n = 17), and 22+ (n = 14) days. Results: A weak and nonsignificant correlation (ϱ = 0.14, ϱ 2 = 0.02, p = 0.10) was observed between actual (median = 8.0 [interquartile range = 8.0]) and predicted 17.9 (7.0) days until clinical recovery. Approximately 30.4% and 80.6% of collegiate athletes with concussion were correctly identified to recover within 14 or 15 days or longer, respectively. Sensitivity was highest (100%) for athletes with concussion who recovered within 7 days and lowest (4.2%) when recovery occurred after 22 days. Specificity was highest (88.1%) when athletes recovered beyond 22 days and lowest (58%) when recovery occurred within 7 days. The positive and negative (+/-) predictive values were 7 (3.5%/100%), 8-14 (34.3%/66.3%), 15-21 (15.0%/90.3%), and 22+ days (7.1%/80.7%). Conclusions: Upon cross-validation, our prognostic formula had limited sensitivity and specificity when predicting clinical recovery after concussion. This formula may be used, with caution, when accompanied by a multidimensional assessment to inform clinical decision making after concussion.