Validation of a guideline to reduce variability in diagnosing cervical dystonia

Abstract

Background: Cervical dystonia is characterized by a variable pattern of neck muscle involvement. Due to the lack of a diagnostic test, cervical dystonia diagnosis is based on clinical examination and is therefore subjective. The present work was designed to provide practical guidance for clinicians in confirming or refuting suspected cervical dystonia. Methods: Participants were video recorded according to a standardized protocol to assess 6 main clinical features possibly contributing to cervical dystonia diagnosis: presence of repetitive, patterned head/neck movements/postures inducing head/neck deviation from neutral position (item 1); sensory trick (item 2); and red flags related to conditions mimicking dystonia that should be absent in dystonia (items 3–6). Inter-/intra-rater agreement among three independent raters was assessed by k statistics. To estimate sensitivity and specificity, the gold standard was cervical dystonia diagnosis reviewed at each site by independent senior neurologists. Results: The validation sample included 43 idiopathic cervical dystonia patients and 41 control subjects (12 normal subjects, 6 patients with isolated head tremor, 4 with chorea, 6 with tics, 4 with head ptosis due to myasthenia or amyotrophic lateral sclerosis, 7 with orthopedic/rheumatologic neck diseases, and 2 with ocular torticollis). The best combination of sensitivity and specificity was observed considering all the items except for an item related to capability to voluntarily suppress spasms (sensitivity: 96.1%; specificity: 81%). Conclusions: An accurate diagnosis of cervical dystonia can be achieved if, in addition to the core motor features, we also consider some clinical features related to dystonia mimics that should be absent in dystonia.