Objective: To assess the diagnostic accuracy of electrodiagnostic (EDX) criteria for the early detection and characterization of Guillain-Barré syndrome (GBS) in clinical practice.
Methods: We conducted a prospective study in patients referred for an EDX exam with clinical suspicion of GBS. We evaluated four sets of neurophysiological criteria and four neurophysiological tests among those recently proposed for the early diagnosis of GBS.
Results: We recruited 84 patients. Acute inflammatory demyelinating polyneuropathy (AIDP) was the final diagnosis in 23 patients. No axonal forms were found. The best sensitivity was obtained using Rajabally et al.'s criteria (82.1%), whereas the specificity was 90.0% for Ho et al.'s and Hadden et al.'s criteria and 100% for the Dutch GBS study group and Rajabally's criteria. Regarding the neurophysiological tests proposed for early diagnosis, the sensitivity ranged from 16.6 to 100%, whereas specificity ranged from 73.1 to 98.3%.
Conclusion: The Dutch GBS study group and Rajabally et al.'s criteria showed an optimal combination of sensitivity and specificity for clinical practice, although with a slightly higher sensitivity for Rajabally et al.'s criteria. None of the neurophysiological parameters recently proposed for early diagnosis have good diagnostic accuracy for clinical application.
Significance: In a real clinical setting with patients referred by neurologists and emergency doctors, an EDX study performed within a week of symptom onset supports the diagnosis of AIDP in 82% of cases.
Keywords: Acute inflammatory demyelinating polyneuropathy; Cohort study; Electrodiagnosis; Guillain-Barré syndrome; Polyradiculoneuropathie aiguë inflammatoire démyélinisante; Syndrome de Guillain-Barré; Électroneuromyographie; Étude de cohorte.
Copyright © 2016 Elsevier Masson SAS. All rights reserved.