Investigating the Risk Factors for Medication Nonadherence in Myasthenia Gravis Patients: Establishing a Nomogram Model.

OBJECTIVE

Many studies have shown that MG patients often show poor adherence to their medications. Poor adherence may affect disease control, and also put a heavy burden on patients' physical functions and medical costs. Therefore, this study aimed to develop and verify a nomogram for predicting medication adherence in MG patients.

METHODS

We prospectively reviewed clinical information, demographic data, and medication adherence records for MG patients at Shijiazhuang People's Hospital from January to December 2025. Logistic regression models were used to analyse risk factors related to medication nonadherence among MG patients; based on this analysis, we constructed a nomogram. Internal validation involved the Bootstrap method, while discrimination and accuracy of the nomogram were evaluated through the C-index, area under the receiver operating characteristic (ROC) curve (AUC), and calibration curves. Potential clinical use was confirmed by decision curve analysis (DCA) and clinical impact curves.

RESULTS

Our study included 376 patients; 194 (51.60%) had poor compliance. There were 263 MG patients in the training cohort, of which 136 (51.71%) had poor compliance. We included five factors in our model: disease recurrence, insurance type, BMQ necessity, BIPQ cognition, and BIPQ emotion. Our model showed an area under the ROC curve of 0.94 (95% CI: 0.91-0.97), with a sensitivity of 87.5% and specificity of 85.8%. Additionally, DCA and clinical impact curves confirmed the effectiveness of the developed model for predicting medication adherence in MG patients.

CONCLUSION

A successful development and validation of a decision curve model based on five risk factors was achieved, which can be utilised for the home management of MG patients.