We plan to go into more depth on how this three-item measure can be considered alongside existing measures of shared decision making in future studies. Interpretation challenges in this area are well known. As Scholl noted [1], patients often interpret attempts to measure their presumed participation in decision making as attempts to assess satisfaction. Entwistle and others have drawn attention to the difference in how patients and researchers interpret terms such as ‘involvement’ [39], [50], [51] and [52]. The reluctance of patients to step into decision making roles is also well-established [39], [53] and [54]. Examining the literature, it seems

that these issues are rarely considered during the development of measures in this field. We intend to evaluate whether the involvement of lay people and patients in the development of our items, DAPT mw through cognitive interviews,

will provide CollaboRATE with a greater degree of content validity. In contrast, all of the patient-reported measures of shared decision making to date have either included the term ‘decision’ or referred to ‘options’ in their item phrasing [1], and therefore, for the reasons already elaborated, they run the risk that patients misinterpret or measurement goal. Rather than narrow our focus on ‘decisions’, we developed items to assess broader aspects of engagement. We found that the phrase ‘what to do next’ was correctly interpreted by patients as involving situations where key determinations

mTOR inhibitor are needed. However this avoided the ambiguity surrounding the use of the term ‘decision’, as well as attributing the decision making role to either patient or provider. Such findings highlight the need to develop tools that are purposefully designed for end-users. The quality of measure development is often compromised by not paying attention to the steps of construct clarification and rigorous item development, particularly when Rebamipide completion requires end-user interpretation. Cognitive interviewing is an established technique to address this requirement [36], and because of the focus on individual responses to item phrasing, is superior to the use of focus groups methods. We also tested responses and preference to two response scales, an important yet often overlooked step in measure design [55]. A potential weakness of this work is the relative homogeneity of the participant sample, and their higher than average educational profiles. However, we noted no difference in item interpretations across the range of educational profiles but accept that further testing of these items would be required in different populations. This work has used recommended qualitative methods to arrive at a brief patient-reported measure of shared decision making that we anticipate will have acceptable content validity.