The objective of this study was to evaluate how the number of demyelinating findings (DF) on nerve conductions affects sensitivity and specificity of electrodiagnostic criteria for chronic inflammatory demyelinating polyneuropathy (CIDP).
Electrodiagnostic findings of 26 consecutive patients with CIDP were compared with amyotrophic lateral sclerosis (ALS) and diabetic polyneuropathy controls. Patients with CIDP were divided into typical and atypical CIDP, as defined elsewhere.
Depending on the minimal required number (MRN) of DF on nerve conductions, sensitivities decreased from an arbitrary 100% to 58% and 54%, for an MRN of 1, 2, and 3, respectively, as specificities increased, from 48% to 81% and 95%, respectively. The number of DF per patient was higher in typical CIDP than in atypical CIDP.
The considerable gap between specificity and sensitivity is the reason for controversy regarding the MRN for the diagnosis of CIDP. Requiring 2 or more DFs to identify CIDP increases specificity from 48% to 81% but lowers sensitivity from 100% to 58%. For patients with other potential causes of neuropathy, the requirement of 2 or more DFs could further increase specificity.