Therapy decision-making for patients with chronic diseases can be difficult. Such patients usually live with their illness(es) all their life, and therapies can only help them improve their condition by managing symptoms, not curing them. Patient-oriented approaches are common to caring for people with chronic conditions because patients’ priorities become relevant means of prioritizing therapies in the absence of a cure. While such type of approach is shown to be effective, it does not leverage evidence on the success of given therapies to achieve specific similar patient goals in the past. Evidence-Based Medicine (EBM) is a concept that was introduced to the medical field in the early 90s to invalidate previously accepted tests and therapies and replace them with new, more powerful, more accurate, more efficacious, and safer ones. Unfortunately, despite the prevalence of patient-oriented approaches for patients with chronic diseases, data collected on patients is not systematically leveraged to support therapy decisions. Combining evidence-based decision-making and patient-oriented approaches could potentially further improve patient outcomes by leveraging the most up-to-date data to recommend and discuss therapy options for patients with chronic conditions. The development and implementation of Learning Health Systems (LHS) is another solution to improving patient outcomes, one that the US Institute of Medicine strongly recommends. The development and implementation of a LHS to support therapy choice for patients with chronic conditions could improve related decisions by fostering continuous learning regarding which therapy may help better achieve which patient goals. However, a learning process that systematically leverages a relevant basis of evidence to support patient-oriented approaches has yet to be defined. As such, this study aims at articulating a learning process for therapy decision-making in the context of chronic conditions. The result is framework and a demonstration of its application using the Goal Attainment Scale (GAS) and synthetic data.