Predictive Score for In-Hospital Mortality of Myasthenic Crisis: A Retrospective Chinese Cohort Study

Abstract

Introduction: Myasthenic crisis (MC) is a life-threatening condition usually occurred in patients with myasthenia gravis (MG). Objective: On the basis of a retrospective case series review, we try to develop a scoring system to evaluate the probability for in-hospital death in MC patients. Methods: We retrospectively reviewed 78 MC patients who were hospitalized from January 2014 to December 2018. Clinical and laboratory data including 17 variables were analyzed univariately. The main clinical outcome was defined as the in-hospital death. Then eligible variables were evaluated by a stepwise multivariate regression and a scoring system was then generated. Calibration and discrimination methods were used to evaluate and validate the model performance. Results: The overall in-hospital mortality was 11.5% (9/78) in the MC cohort. Five clinical variables including Myasthenia Gravis Foundation of America (MGFA) classification at onset, septic shock, thymoma classification, cardiac arrest, and the lowest serum albumin were found to be associated with in-hospital mortality. Further 3 variables entered the final regression, and internal validation showed that the area under the curve was 0.919 (95% CI 0.788–1.000). A scoring system with a full credit of 7 points was generated to predict the in-hospital mortality (MGFA at onset, 2 points; septic shock, 4 points; and cardiac arrest, 3 points). Conclusions: A 7-point scoring model was established on the basis of a retrospective review of MC patients to predict the in-hospital mortality.