Performance of individual items of the quantitative myasthenia gravis score

We evaluated the quantitative myasthenia gravis score to determine which items better discriminate among patients and are most responsive to treatment. Current scoring ranges were compared to updated normative data when appropriate. Data from a previous study of intravenous immunoglobulin versus placebo in myasthenia gravis were analyzed. Each item of the quantitative myasthenia gravis score was analyzed for distribution, responsiveness and inter-item correlations. Items on swallowing, speaking and vital capacity had marked floor effects as did grip strength based on current normative data. The timed measures were responsive to treatment but diplopia, vital capacity and grip strength were not. Ocular items were not correlated with bulbar and generalized items (p>0.05), suggesting multi-dimensionality. Contralateral arm and leg tests were highly correlated (r>0.9, p<0.001), indicating redundancy. In summary, we found that the bulbar items and grip strength did not perform well. Timed items from the Quantitative Myasthenia Gravis Score, representing fatigability, are responsive to change and we propose new ranges based on updated normative data.