In reply
Crittenden and Shah raise concerns about “the real-world effectiveness of patient-centered decision aids in improving measurable outcomes and what those measures should be.” This is an important comment relevant to widespread implementation and applicability of patient-centered decision aids across different clinician, patient, and practice contexts. The Diabetes Medication Choice trial aimed to evaluate “efficacy” of a decision aid to improve patient knowledge, involvement in making a decision, and adherence and glycemic control in a single practice context.1 However, it was not designed to document real-world effectiveness across different practice contexts or how to normalize their use into routine usual care.2