Related Experiment Video
Updated: Jun 4, 2026
A Multi-Modal Approach to Assessing Recovery in Youth Athletes Following Concussion
Published on: September 25, 2014
Reliability of a graded exercise test for assessing recovery from concussion
John J Leddy1, John G Baker, Karl Kozlowski
1Department of Orthopaedics and Sports Medicine Institute, University at Buffalo, State University of New York, Buffalo, NY 14214, USA. leddy@buffalo.edu
Objective:
To evaluate a graded treadmill test for retest reliability (RTR) and interrater reliability (IRR) in the evaluation of the physiologic effects of symptom exacerbation from concussion.
Design:
Prospective case series (RTR) and blinded rater assessment of 10 actors portraying patients with and without symptom exacerbation (IRR).
Setting:
University Sports Medicine Concussion Clinic.
Participants:
For RTR, 21 refractory concussed patients (11 athletes and 10 nonathletes) and 10 healthy subjects; for IRR, 32 raters representing a variety of health care disciplines.
Intervention:
For RTR, a Balke protocol treadmill test to symptom exacerbation before and after 2 to 3 weeks. For IRR, video recordings of actors during the treadmill test viewed by raters blinded to condition.
Main Outcome Measures:
For RTR, agreement of the tests for maximal heart rate (HR), systolic blood pressure, diastolic blood pressure, and rating of perceived exertion. For IRR, presence or absence of symptom exacerbation and the symptom exacerbation HR.
Results:
Raters achieved a sensitivity of 99% for identifying actors with symptom exacerbation and a specificity of 89% for ruling out concussion symptoms and agreed on 304 of 320 observations (accuracy of 95%). The intraclass correlation coefficient for the symptom exacerbation HR was large at 0.90 (95% confidence interval, 0.78-0.98). The treadmill test had good RTR for maximum HR (intraclass correlation coefficient, 0.79) but not for systolic blood pressure, diastolic blood pressure, or rating of perceived exertion.
Conclusions:
The Balke exercise treadmill protocol has very good IRR and sufficient RTR for identifying patients with symptom exacerbation from concussion.
