A latent variable framework for modeling dyadic measures in research on shared decision-making

Abstract

The aim of our study is to introduce a statistical framework using latent variable modeling for the investigation of correspondence between patients' and physicians' perceptions that were measured using dyadic instruments. This statistical approach combines multitrait-multimethod and measurement invariance methodologies. In an illustrative example we used a sample of 285 primary care consultations on chronic diseases to test correspondence between patients' and physicians' views on the process of shared decision-making (SDM), which was measured by the patient and physician version of the nine-item Shared Decision-Making Questionnaire (SDM-Q-9 and SDM-Q-Doc), respectively. We revealed that while patients and physicians seem to agree on what is the core of SDM, they differ in their ratings regarding to which extent it is present in a certain consultation. The described statistical approach provides important insights into correspondence between perceptions of different stakeholders that cannot be gained using traditional approaches. However, its generalizability is questionable due to the partly explorative data-driven approach and its flexibility is limited by the requirement of samples including at least 200 cases.