Quantitative ocular tests for myasthenia gravis: a comparative review with detection theory analysis.

Many reports in the literature describe a variety of ocular signs in myasthenia gravis. To determine the utility of laboratory recordings of ocular signs in the evaluation for myasthenia, we reviewed all previous studies of quantitative measures of eye movements or intra-ocular pressures. We selected those studies with data presented for both myasthenic and non-myasthenic ocular palsies. Signal detection theory was used to evaluate the discriminative power of each variable. The characteristics of saccades and quick phases of optokinetic nystagmus at the start of recording were poor at distinguishing between myasthenic and non-myasthenic palsies, except when the comparison was solely between myasthenia and chronic progressive external ophthalmoplegia. The effects of fatigue on saccadic parameters were also not discriminative, though there was insufficient data to evaluate this adequately. Changes induced by edrophonium in the amplitude or peak velocity of saccades or optokinetic quick phases were good diagnostic tests, retaining high sensitivities when criteria were set to yield a specificity of 95%. Most of these parameters were less efficient as screening tests, with modest specificities when criteria were set to yield a sensitivity of 95%. The change in optokinetic quick phase amplitude recorded by infrared oculography was the best test, with ideal characteristics of 97% specificity and sensitivity at a criterion of zero change. This analysis suggests that eye movement recordings of saccades or optokinetic nystagmus have potential as useful and inexpensive tests for myasthenia, and warrant further study.