Sensitivity and specificity of repetitive nerve stimulation with lower cutoffs for abnormal decrement in myasthenia gravis
Christopher J. Lamb, Devon I. Rubin
Abstract
INTRODUCTION: The sensitivity of repetitive nerve stimulation (RNS) in myasthenia gravis (MG) is dependent on the cutoff for abnormal decrement. METHODS: RNS data of adults with and without MG from 2014 to 2017 were reviewed retrospectively. The maximum reliable RNS amplitude/area decrement before and after exercise from facial, spinal accessory (SA), ulnar, and fibular nerves was recorded. Sensitivity/specificity using 5%, 7%, and 10% cutoffs were calculated. RESULTS: Seventy-nine of 141 patients had MG (46 generalized, 21 ocular, 12 bulbar). A total of 608 unique RNS recordings were analyzed. Overall RNS sensitivity/specificity at ≥5%, ≥7%, and ≥10% amplitude cutoffs were as follows: SA, 65.6%/86.3%, 49.2%/94.1%, and 29.5%/96.1%; facial, 51.0%/82.5%, 43.1%/95.0%, and 37.3%/100%; ulnar, 43.6%/100%, 41.0%/100%, and 41.0%/100%; and fibular, 52.6%/89.5%, 42.1%/94.7%, and 42.1%/100%. DISCUSSION: Lowering amplitude cutoff from 10% to 7% increased or maintained sensitivity with little loss in specificity. Post-exercise and area analysis resulted in increased sensitivity in some circumstances.