OBJECTIVES: Myasthenia gravis (MG), a neuromuscular disorder, results in muscle weakness that fluctuates daily, leading to difficulty in determining overall disease burden. This study investigates the feasibility of collecting patient reported outcomes virtually to quantify disease burden and severity.

METHODS: U.S. Patients with MG were enrolled in PicnicHealth’s research platform and consented to collection of their medical records. Structured and unstructured data were abstracted using human-validated machine learning. Eligible patients completed the MG Activities of Daily Living Profile (MG-ADL) on PicnicHealth's virtual patient portal within 90 days of enrollment. Baseline MG-ADL responses (scored 0-normal to 24-most severe) were used to categorize disease burden and severity (0-4, mild; 5-9, moderate; 10-24, severe). MG-related medications that were either ongoing, or started within 90 days of baseline, were used to determine treatment patterns. Demographics were summarized with descriptive statistics.

RESULTS: Of the 249 eligible patients (70% female, 77% White), median (IQR) ages at diagnosis were 39(26-, 53) for females and 58(44-, 70) for males.The mean total MG-ADL score was 8, with 17.3%, 47.8%, and 34.9% classified as mild, moderate and severe, respectively. Males reported less severe symptoms (mean total MG-ADL 7 vs. 9 , p<0.001 and 22% vs. 40% severe, p=0.005). Mild patients attributed breathing difficulties, eyelid droop, and rising from a chair most strongly to their total score. Comparatively, moderate and severe patients reported symptoms more evenly across domains and the contribution of double vision and impairment with brushing teeth and hair increased. Mild patients utilized fewer unique treatments than moderate and severe patients (1.63 vs. 2.13 vs. 2.20 mean medications, p=0.036).

CONCLUSIONS: Patients’ daily disease burdens can be effectively monitored virtually. Given frequent variations in these burdens, future studies will aim to quantify trends in disease burden over time.