Survival Outcomes of Medicare-covered Elderly US Population with Myasthenia Gravis
Yuebing Li1, David Bruckman1, Jesse Schold2, Benjamin Claytor1, Nicholas Silvestri3, Michael Hehir4, Ikjae Lee5
1Cleveland Clinic, 2University of Colorado, 3UBMD Neurology, 4Larner College of Medicine at the University of Vermont, 5neurology, Columbia University
Objective:
We performed a retrospective longitudinal study using Medicare claims data (2006 - 2019).
Background:
Survival outcomes and causes of death have been inadequately studied among myasthenia gravis (MG) patients in the United States (US).
Design/Methods:
Incident MG cohort in 2010-2011 was identified; age ≥65 years, at least one month of FFS Parts A/B coverage, no health maintenance organization insurance coverage, initial and subsequent MG claims within 2010-2011 separated by at least 28 days. The non-MG control cohort was randomly selected 5 times the number of the MG cohort matched with age, sex, region of residence, and start of Medicare coverage. Overall and cause-specific mortality was compared between cohorts in the following 7-year period using the Kaplan-Meier plots and Cox proportional hazard model, adjusted for the Charlson comorbidity index (CCI). Mortality rates were estimated with Poisson regression offset by total enrollment (person-year).
Results:
Cohorts of 6,024 incident MG and 30,083 control beneficiaries were identified. Baseline demographics were well-matched between cohorts. Average 7-year mortality rate was higher in the MG cohort compared to the control cohort (66.8 vs 57.1 per 1,000-person-year, p<0.0001). 6% of the MG cohort died of MG as the primary cause. Compared to the non-MG control cohort, mortality rates (per 1,000-person-year) specific to MG and infection as the cause of death were significantly higher among the MG cohort (<0.01 vs 11.0 and 1.2 vs 2.0 respectively) while malignancies and dementia-specific mortality rates were significantly lower in the MG cohort (12.5 vs 10.3 and 7.2 vs 4.7). The CCI was higher in the MG cohort at enrollment and during the follow-up periods. After adjusting for the time-varying CCI, there was no significant difference in 7-year survival between the MG and the control cohorts.
Conclusions:
Long-term mortality is slightly increased in elderly MG patients compared to non-MG counterparts, driven by MG-related deaths and their higher comorbidities.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.