Background: Internationally approved electrodiagnostic criteria for Guillian Barre syndrome(GBS) are lacking in children.

Aims: To compare the diagnostic accuracy of the different electrophysiological criteria in childhood GBS at the time of sentinel assessment.

Methodology: In this single center study, data of children diagnosed with GBS between January 2013 to December 2017 was retrieved. Patient charts were reviewed for clinical features, electrophysiological recordings. The electrodiagnostic results (4 motor nerves and 2 sensory nerves in upper limbs + lower limbs) were reanalyzed and were classified based on Dutch group; Ho; Hadden; Hughes and Rajabally criteria for GBS. All criteria were compared against the Hadden criteria.

Results: 215 children with GBS were evaluated. 15 children had incomplete electrophysiological data and 4 children were excluded due to missing data. Of the 186 children, the mean age of onset was 77 months; median duration from onset of symptoms to electrodiagnostic evaluation was 7 days; pure motor and motor-sensory form of GBS was seen in 71 and 115 children. Based on Hadden, demyelination was noted in 57 children; axonal in 37; Inexcitable in 84 and Equivocal in 8 children. The sensitivity of the various criteria ranged from 71% to 100% for demyelination; 97% to 100% for axonal. The Hughes criteria had the lowest specificity (85% [95% CI 79-90%]) for axonal forms. The diagnostic accuracy of Rajabally criteria was > 98% for axonal, demyelinating and inexcitable forms of GBS.

Conclusions: There is wide variation in the classification of GBS subtypes using different criteria excepting Rajabally criteria