headache

Importance: Concussion is a mild traumatic brain injury with associated abnormalities in brain function, rather than structural injury. An estimated 1.1 million to 1.9 million pediatric concussions occur annually in the US.

Objective: To determine the accuracy of clinical history and physical examination findings for identifying concussion in children and adolescents who have had a plausible mechanism of injury.

Data sources and study selection: PubMed, Embase, ClinicalTrials.gov, Cochrane Library, CINAHL, Web of Science, and Google Scholar were searched from January 2002 through December 2025 without language restrictions. Observational studies including patients aged 2 to 18 years evaluated for concussion in outpatient, emergency, or inpatient settings were included.

Data extraction and synthesis: Four reviewers independently extracted study characteristics and diagnostic accuracy data and assessed study quality with the Rational Clinical Examination levels of evidence.

Main outcomes and measures: Sensitivity, specificity, and likelihood ratios (LRs) for symptoms and physical signs associated with concussion were calculated using random-effects meta-analysis when summary measures were appropriate.

Results: Of 7110 screened abstracts, 23 studies (level 4 evidence; case-control design) met inclusion criteria. The presence of mental fog (LR, 11.8-12.0; specificity, 0.96), noise sensitivity (LR, 6.9; 95% CI, 3.6-13.1; specificity, 0.94), nausea (LR, 6.7; 95% CI, 3.1-14.6; specificity, 0.93), and light sensitivity (LR, 6.4; 95% CI, 2.1-19.7; specificity, 0.93) were most useful for increasing the likelihood of a concussion diagnosis. The absence of headache was the most useful symptom for decreasing the likelihood of concussion (LR, 0.20; 95% CI, 0.10-0.39; sensitivity, 0.74). Signs that increased the likelihood of concussion were abnormal near-point convergence, which is the inability to maintain ocular convergence on a near target (LR, 7.0; 95% CI, 2.0-24.9; specificity, 0.97); abnormal smooth pursuits, which are jerky, irregular eye movements when tracking a target (LR, 6.5; 95% CI, 2.4-17.5; specificity, 0.96); and saccades, which are inaccurate or slow eye movements with overshooting or undershooting when looking between 2 or more targets (LR, 4.8; 95% CI, 1.8-13.1; specificity, 0.92); however, none of these findings had a sensitivity of more than 0.40. A consensus statement by the International Conference on Concussion in Sport recommends the Sport Concussion Assessment Tool to systematize the comprehensive evaluation of patients with symptoms concerning for concussion.

Conclusions and relevance: While no single finding was sufficient to confirm or exclude concussion, the presence of mental fog, noise and light sensitivity, nausea, or ocular abnormalities were most useful to identify patients more likely to have had a concussion, while absence of a headache made a concussion less likely. These symptoms and signs are integrated into structured clinical assessments to support the clinical diagnosis and management of pediatric concussion.