Dunja Dreesens

73 professional needs to ‘translate’ it together with the patient to the patient characteristics and preferences (199), whereas decision aids aim to acquaint patients with information about choices and available options, including watchful monitoring. Of late, clinical practice guidelines (recommendations) are created or re-written to include these elements as well, but it is not a common practice yet (180, 200, 201). As one of the participants involved in the meeting noted: “Looking through a future lens and decision making view: [there will be] a set of guidelines that don’t [provide room for contemplation], and ‘innovator’ guidelines that do list options.” [quote on reconciling clinical practical guidelines and shared decision-making] (202) Furthermore, some participants argued using other sources than CPGs as a basis for patient- directed knowledge tools, such as evidence summaries or systematic reviews, and questioned how to include qualitative data in the evidence base (203). We also discussed whether the framework should be discretionary or prototypical. Discretionary meaning that the core elements mentioned in the framework are optional; and prototypical indicating that we should aim to include the core elements listed in the framework in patient-directed knowledge tools. For example, current tools mostly disregard core elements such as harms and the need to elicit patient’s values (204). Often only one option is stated, instead of giving a complete overview of options, including watchful waiting when relevant (69). We decided that it would be a descriptive and prototypical framework. This decision resulted in one participant to say: “We are wasting time on patient education material without options; we need to be bold.” [quote on importance of a prototypical framework] Finalizing the draft framework entailed checking and rewriting the wording used in the framework, and reorganizing the core element categories (171, 205-207). The category ‘outcomes’, which refers to the effect(s) of the tools, was not included in the framework. Possible effects proposed and discussed during the meeting were: improved knowledge acquisition by the patient, more realistic or informed expectations, decreased decision burden/ stress or less regret. As most of these outcomes have not been confirmed by research (yet), we felt it premature to include this category in the framework (208). The draft framework at the end of the meeting contained eight tools (see table 2), of which two – ‘care map’ and ‘decision box’ – were conditionally included as we were not entirely sure if they were patient-directed knowledge tools. The framework in the article contains five tool types, because when looking further into the tools ‘question prompt list’ and ‘decision box’, we decided that these tools belong to the tool type ‘encounter PDAs, and they were therefore removed from the framework as separate tool types. The tool type ‘care map’ was left out as well, as they are also called ‘care pathways’ (70, 143, 178), and we excluded the latter at the beginning of the meeting. It was also suggested in the feedback rounds via email to leave out the tool ‘decision tree’. Due to renewed attention to this tool type – e.g. digitalising CPGs into decision trees to enhance use by healthcare professionals and patients (107, 209, 210), and because the tool was not included conditionally, it is still part of the draft framework. The final draft of the framework contains five patient-directed tool types, and consists of two Chapter 4