New criteria for early electrodiagnosis of acute inflammatory demyelinating polyneuropathy

Abstract

A variety of electrodiagnostic methods are used to confirm the diagnosis of acute inflammatory demyelinating polyneuropathy (AIDP), but difficulties are frequent during the first few weeks of weakness. We compared the nerve conduction studies (NCS) of patients with AIDP to those with critical illness polyneuropathy (CIP), a subacute axonal polyneuropathy. New electrodiagnostic criteria with graded certainty (normal, nondiagnostic, suggestive, highly suggestive, and definite) were designed and applied in a blinded manner to both groups. Among the AIDP patients, 64% met the highly suggestive and definite criteria (specificity 95–100%, P < 0.01), whereas 80% of the CIP group fell in the nondiagnostic criteria (P < 0.001). The relative preservation of the sural sensory response in spite of at least two abnormal sensory NCS in the upper limb suggested acute demyelination (sensitivity 48%, specificity 96%, P < 0.001) and was even more conclusive when associated with absent or prolonged F waves. Motor and sensory response amplitudes were lower in the CIP group, with comparable mean motor and sensory distal latencies and motor conduction velocities. Motor conduction blocks were present in 10% of nerves in AIDP and were not encountered in CIP. The frequency of absent or delayed F waves and absent H reflex was similar in both groups. The correlation coefficient of the cerebrospinal fluid protein concentration with the designed criteria was higher in the AIDP group (r = 0.9). We conclude that a new criterion with graded certainty is of higher specificity in the majority of patients with early AIDP. Muscle Nerve, 2005