Diagnosis of carpal tunnel syndrome: interobserver reliability of the blinded scratch-collapse test

The reliability of the scratch-collapse test for diagnosis of carpal tunnel syndrome (CTS) has not been tested by independent investigators. This study measured the reliability of the scratch-collapse test comparing the treating hand surgeon and blinded evaluators. We performed a prospective observational study of 41 patients with a provisional diagnosis of CTS or a combination of CTS and cubital tunnel syndrome and prescribed electrodiagnostic testing. The treating hand surgeon performed the scratch-collapse test. Next, the test was administered by one of the six observers, unaware of the patient's symptoms and the diagnosis made by treating hand surgeon. The kappa statistic (κ) was used to calculate the interrater reliability between the treating hand surgeons and blinded scratchers. The agreement between the blinded observers and the hand surgeons on the scratch-collapse test was substantial 0.63 (95 % CI, 0.34-0.87; p < 0.001). The sensitivity of the blinded scratch test in our sample was 32 %. In a small study with a spectrum bias favoring electrophysiologically confirmed CTS the reliability was lower than that reported by the inventors of the test, but was still substantial. We propose a larger study of patients with a greater variety of electrodiagnostic test results using fewer observers with more experience.