Medical Tests for the Diagnosis of Myasthenia Gravis: A Statistical Analysis of Jitter on Single Fiber Electromyography

• Activity Number: 187
• Type: Topic Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Norberto Pantoja*+ and Rebecca A. Betensky and Pushpa Narayanaswami and Seward Rutkove
• Companies: U.S. Food and Drug Administration and Harvard School of Public Health and Harvard Medical School and Harvard Medical School
• Keywords: Early stopping ; futility analysis ; conditional power design ; medical diagnostic tests ; myasthenia gravis

Abstract

Single Fiber Electromyography (SFEMG) is an electromyographic technique that identifies and records action potentials from individual muscle fibers. The major clinical application of SFEMG has been in the diagnosis of Myasthenia Gravis.

During SFEMG at least 20 muscle fiber action potential pairs are recorded from a given muscle. An abnormal study can be declared if at least 10% of potential pairs have jitter that exceeds the upper limit of normal for paired jitter in that muscle.

SFEMG is a time consuming procedure associated with discomfort. If it were possible to predict that the test was likely to be normal after a certain number of muscle fiber pairs had been analyzed, the test could be terminated earlier without the extra time and discomfort associated with collecting 20 fiber pairs. There would be little motivation to continue the procedure if the probability of declaring an abnormal test by the end of the 20 pairs given the interim data is low.

We propose a statistical analysis that suggests early stopping of the SFEMG procedure if the conditional probability of declaring an abnormal test by the end of the procedure given the interim data is sufficiently low