Identifying Persistent Postconcussion Symptom Risk in a Pediatric Sports Medicine Clinic David R. Howell, * yz§ PhD, Roger Zemek, k MD, Anna N. Brilliant, §{ BS, Rebekah C. Mannix, # ** MD, Christina L. Master, yyzz MD, and William P. Meehan III, §{ ** MD Investigation performed at the Sports Concussion Clinic, Boston Children’s Hospital, Boston, Massachusetts, USA Background: Although most children report symptom resolution within a month of a concussion, some patients experience per- sistent postconcussion symptoms (PPCS) that continues for more than 1 month. Identifying patients at risk for PPCS soon after an injury can provide useful clinical information. Purpose: To determine if the Predicting Persistent Post-concussive Problems in Pediatrics (5P) clinical risk score, an emergency depart- ment (ED)–derived and validated tool, is associated with developing PPCS when obtained in a primary care sports concussion setting. Study Design: Cohort study; Level of evidence, 3. Methods: We conducted a study of patients seen at a pediatric sports medicine concussion clinic between May 1, 2013, and October 1, 2017, who were \19 years of age and evaluated within 10 days of a concussion. The main outcome was PPCS, defined as symptoms lasting .28 days. Nine variables were used to calculate the 5P clinical risk score, and we assessed the association between the 5P clinical risk score and PPCS occurrence. The secondary outcome was total symptom duration. Results: We examined data from 230 children (mean age, 14.8 6 2.5 years; 50% female; mean time from injury to clinical assess- ment, 5.6 6 2.7 days). In univariable analyses, a greater proportion of those who developed PPCS reported feeling slowed down (72% vs 44%, respectively; P \ .001), headache (94% vs 72%, respectively; P \ .001), sensitivity to noise (71% vs 43%, respec- tively; P \ .001), and fatigue (82% vs 51%, respectively; P \ .001) and committed 4 errors in tandem stance (33% vs 7%, respec- tively; P \ .001) than those who did not. Higher 5P clinical risk scores were associated with increased odds of developing PPCS (adjusted odds ratio [OR], 1.62 [95% CI, 1.30-2.02]) and longer symptom resolution times (b = 8.40 [95% CI, 3.25-13.50]). Among the individual participants who received a high 5P clinical risk score (9-12), the majority (82%) went on to experience PPCS. The area under the curve for the 5P clinical risk score was 0.75 (95% CI, 0.66-0.84). After adjusting for the effect of covariates, fatigue (adjusted OR, 2.93) and 4 errors in tandem stance (adjusted OR, 7.40) were independently associated with PPCS. Conclusion: Our findings extend the potential use for an ED-derived clinical risk score for predicting the PPCS risk into the sports concussion clinic setting. While not all 9 predictor variables of the 5P clinical risk score were independently associated with the PPCS risk in univariable or multivariable analyses, the combination of factors used to calculate the 5P clinical risk score was sig- nificantly associated with the odds of developing PPCS. Thus, obtaining clinically pragmatic risk scores soon after a concussion may be useful for early treatments or interventions to mitigate the PPCS risk. Keywords: mild traumatic brain injury; pediatric; recovery; postconcussion syndrome; assessment A sports-related concussion prognosis is challenging because of the heterogeneity of clinical presentation among patients and a lack of well-validated objective diagnostic methods available. In addition, little is currently known about how a patient prognosis varies within different points of health care entry settings, as most studies focus on those who are adolescents or older or take place in the emergency department (ED) setting. 1 As a result, clinicians have highly variable practice patterns related to testing methods, treatment options, and criteria used for diagnosis and prognosis. 4,29 Contributing to the man- agement difficulty, no consensus exists regarding what measurements can be used to determine diagnoses, prog- noses, and recovery timelines using robust and easily administered methods. In particular, child athletes are a challenging population because recovery timelines have been established primarily from adult populations 20 and they may be more vulnerable to prolonged recovery com- pared with their adult counterparts. 2,16 Despite most chil- dren reporting that their symptoms resolve within a month of the injury, 18,23 a substantial number may continue to The American Journal of Sports Medicine 1–8 DOI: 10.1177/0363546518796830 Ó 2018 The Author(s) 1