Inter-Rater Reliability for a Recently Developed Cluster of Headache Assessment Tests

Objectives: An important component of assessment-based physiotherapy management of patients with headache is a thorough physical examination to identify musculoskeletal dysfunctions. A recently published international consensus study identified 11 clinically useful headache assessment tests (HATs). Test properties are not yet documented for all HATs. Different rating methods include absolute values and a 0-100 visual analog scale (VAS) to indicate the clinical relevance of a test.

Methods: To evaluate the inter-rater reliability of different scales, two manual therapists, blinded towards the diagnosis, examined 25 patients with headache and 25 headache-free controls. Inter-rater reliability was evaluated using the intraclass correlation coefficient and Bland-Altman plots for interval data; Cohen‘s kappa and Gwet´s AC for categorical data.

Results: Substantial to excellent reliability (Gwet´s AC or ICC > 0.7) was observed for flexion-rotation test, muscle strength, upper cervical quadrant, the examination of active range of motion and reproduction and resolution of symptoms. Forward-head posture, cranio-cervical flexion test, passive accessory intervertebral movements and the observation of latent trigger points, showed moderate levels of interrater agreement (Gwet´s AC or ICC> 0.5), all other tests showed only low levels of agreement.

Discussion: We recommend a 0-100 scale to indicate the clinical relevance of a test result. Substantial to excellent reliability (ICC > 0.7) was observed for active range of motion, trigger point palpation and upper cervical quadrant. Reliability was higher in the headache group. VAS can be handled as interval data; for clinical purposes it can be dichotomised at a cut-off point at 20 mm to indicate a positive or negative test result.