JMIR Publications

ABSTRACT

Managing chronic diseases requires ongoing monitoring of disease activity and therapeutic responses to optimize treatment plans. With the growing availability of disease-modifying treatments (DMTs), it is crucial to investigate comparative effectiveness and long-term outcomes beyond those available from randomized clinical trials (RCTs). We introduce a comprehensive pipeline for generating reproducible and generalizable real-world evidence (RWE) on disease outcomes by leveraging electronic health record (EHR) data. The pipeline links EHR data with registry information and applies algorithms based on longitudinal EHR features to evaluate therapies for chronic diseases, as illustrated through a case study of multiple sclerosis. Our approach addresses challenges in RWE generation for disease activity of chronic conditions, specifically the lack of direct observations on key outcomes and biases arising from imperfect or incomplete data. We present advanced machine learning techniques such as semi-supervised and ensemble methods to impute missing outcome data, further incorporating steps for calibrated causal analyses and bias correction.