Development and preliminary evaluation of a decision coach training module for nurses in Norway

We developed a decision coach training module to be added to the Ready for SDM meta-curriculum. The purpose of the decision coach training was to build knowledge and skills needed to provide decision coaching to patients. There were 10 specific objectives for Part A in the classroom instruction and five objectives for Part B on the self-appraisal of decision coaching (see Table 2). To accomplish these objectives, we selected relevant modules from the meta-curriculum (Basic SDM course and SDM skills in situ), components from other evaluated coaching trainings and created a new module specific to decision coaching.

Rationale for the training also considered the nurses’ motivation to engage in becoming a decision coach. Motivation might be strong due to nurses’ closeness to the patient, their particular focus on trust and relationship building and their central roles as providers of treatment and counselling [31]. On the other hand, nurses might be reluctant to become committed to tasks related to decision-making, which they might feel belong to the traditional responsibilities of the physicians. The training was designed to target nurses from quite heterogeneous medical contexts. To equip trainees with skills for specific contexts while training them in context-heterogeneous groups puts particular emphasis on the need to facilitate the transfer from generic knowledge and skills to the individual decision settings and medical contexts. The development was focused on nurses working in a medical domain with existing PDAs to make sure that the nurses would have access to evidence-based information for a specific decision for use with patients, which is known to improve outcomes of decision coaching [11].

In line with the purpose of the “knowledge creation funnel” of the KTA framework, the draft for the decision coaching training was adapted from our group's previous studies evaluating SDM training [18, 21, 22, 26, 32‐34]. The module's design was shaped by Knowles' Adult Learning Theory [35], which highlights four crucial planning steps for adult learning: 1) determining learning needs, 2) creating appropriate learning strategies and resources, 3) implementing these strategies and 4) evaluating outcomes. This framework aligns seamlessly with our development process, where learning needs were initially assessed through stakeholder feedback, during the piloting of components and a brief literature review of learning objectives for nurses in SDM [36]. Resources were then tailored to meet these needs and finally, outcomes were evaluated using Kirkpatrick's model.

The curriculum is designed for a structured progression from foundational knowledge to advanced competencies in SDM and decision coaching. Initially, participants grasp basic SDM principles and coaching concepts. As they progress, they develop practical techniques and communication skills through role-play and simulations in Part A of the module. Part B advances to independent application and self-appraisal in clinical settings. Each stage builds on the previous, enhancing participants' abilities to apply learned skills in real-world scenarios. This progression aligns with Bloom's Taxonomy, moving systematically from basic knowledge and comprehension to advanced application [37]. Table 2 clearly outlines this progression, aligning learning objectives with corresponding learning/teaching methods and materials to ensure a coherent educational progression.

Two established modules from the meta-curriculum were incorporated in the new decision coaching training module: 1) a generic “basic SDM course” (Fig. 1), previously evaluated for this target group and 2) the proven efficacious observation- and feedback-based “SDM skills training in situ” from the doktormitSDM module (new name: SDM skills in situ). In addition, we created a new module specific to decision coaching. The single components in the “decision coaching” part, regarding the nurse’s role in interprofessional SDM, were developed or adjusted, inspired by a Canadian decision coach training approach [32, 38] in the associated Framework for Decision Coach-Mediated Shared Decision Making (based on the Ottawa Decision Support Framework) [17].

Single components of the decision coach training, such as role-play exercises and a moderated discussion of nurse-led decisions relevant for SDM, were piloted within several training sessions for nurses guided by the KTA. Although the action cycle from the KTA framework has a stepwise flow, the phases were revisited several times before moving on to later phases. For example, after piloting the role-play exercises, we returned to the “knowledge creation funnel” to develop new methods for overcoming the barriers that were encountered. This resulted in the development of standardised patient cases (Table 3). Exercises to facilitate the training of practical skills were adjusted from the Canadian approach to fit the Norwegian context. Field observations were taken by the instructors to assess the appropriateness and comprehensibility of the components.

Learning resources and BCTs [29] used in the Ready for SDM meta-curriculum [22, 39] were adapted for nurses providing decision coaching. Throughout the entire training, learning activities were used with reference to evidence-based BCTs. BCTs are considered the “active ingredients” in the implementation of an intervention and are defined as the observable, replicable components of behaviour-change interventions that can be used individually or in combination with other BCTs [29, 40]. For example, the BCT “action planning guidance” in the meta-curriculum was operationalised using pocket reminder cards (see Table 1) [24].

We applied to the Norwegian Nursing Association (NSF) for six hours of Continuing Nursing Education Credits. A certificate of completion of the module (parts A and B) was approved by the Norwegian Nursing Association (NSF) for six hours of Continuing Nursing Education Credits.

The second part of the training focused on practical application and evaluation. Three primary goals were for the nurses: to apply their knowledge and skills by providing decision coaching to an actual patient with decisional needs in their clinical practice, to analyse their own decision coaching performance and to receive individualised feedback on their recorded coaching session.

To facilitate this practical component, participants received an audio recorder, memory stick and patient informed consent form. They were instructed to record their decision coaching session and analyse it using the MAPPIN'SDM (Multifocal Approach to Sharing in SDM) HCP scale (see Additional file 3) [42, 45].

A follow-up meeting with the course instructor (SK) was arranged to occur either at the clinical department or via a virtual call. For those who submitted recordings, a 15-min interactive feedback session was provided, tailored to each individual nurse based on MAPPIN'SDM analysis of their audio taped decision coaching [26]. The MAPPIN'SDM inventory, which demonstrates good sensibility, reliability and validity [46], was used by SK to assess the extent of patient involvement in decision-making in the recorded coaching sessions. Rather than providing quantitative feedback, the instructor identified and discussed specific areas where further skill development would be most beneficial.

The study was conducted at two study sites (hospitals) within the Northern and the South-Eastern Norway Regional Health Authorities. The organisers were informed that the target group for the decision coach training was nurses working in departments with SDM relevant decisions (i.e. PDAs available such as breast cancer surgery or PDAs in development). The participants needed to be able to understand and communicate in Norwegian.

Recruitment was pursued through SDM ambassadors/facilitators who had previously undertaken the “Ready for SDM – train-the-trainer course”. They were informed about the accompanying evaluation of the training and they organised practicalities regarding the course. The training programme was designed for a maximum of ten nurses per course and was approved as regular working hours for the nurses.

Data collection was conducted at two time-points:

The training module was documented in a training protocol to ensure standardised delivery in both locations. All training was delivered by SK, a registered nurse with a master’s degree in Health and Empowerment and a PhD student. She was trained in the using the Ottawa Decision Support Framework at the University of Ottawa in a graduate course and is a certified MAPPIN´SDM observer.

Participants reported their demographics, including gender, age, clinical departments and years of clinical practise in a short questionnaire (Table 4). Variables measured to evaluate the module were based on three of Kirkpatrick's four levels [47], incorporating both quantitative and qualitative measures. The primary assessment tool was a self-administered online questionnaire comprising 13 items, adapted from previously validated evaluations used in other Ready for SDM curriculum modules (see Additional file 4). This was supplemented by the MAPPIN'SDM observer instrument, which measures the extent of patient involvement in decision-making. The items, provided at Kirkpatrick's Level one (Reaction), assessed relevance and satisfaction using four 4-point Likert-scale items (1 = strongly disagree, 4 = strongly agree) and one question had the response options "yes," "no" and "unsure". Additionally, indicators of applicability and suggestions for improvements were collected from the participants using open-ended questions and field notes were taken by the co-host.

At Level two (learning), we assessed perceived understanding, attitude, self-rated SDM skills, self-confidence about providing decision coaching using five items (4-point Likert scale, 1 = strongly disagree, 4 = strongly agree) and one item regarding the commitment to providing decision coaching in clinical practice (5-point Likert response scale, 1 = Very likely, 5 = Very unlikely).

At Level three (behaviour/practical application of decision coaching), we assessed whether 1) the participants completed Part B and 2) the nurses’ perceived barriers and facilitators for implementing decision coaching in their context and their intention to implement it along with PDAs (two open-ended questions at the end of Part A).

Guided by the KTA framework, this study demonstrated that it was possible to develop a module on decision coach training for nurses that could be used alongside the "SDM basic module" and the "SDM in situ" modules. Our findings revealed three key messages. First, the classroom component proved both feasible and acceptable, with nurses participating actively and 90% rating it as highly relevant to their practice. Second, while participants expressed high confidence and positive attitudes immediately after training, the module could not yet demonstrate behavior change in clinical practice. Third and most notably, we uncovered a significant barrier to skill development: nurses' strong reluctance to audio record their coaching sessions, with only one of 19 participants completing this crucial practical component.

The study has several possible limitations. Firstly, it is limited by the small sample; however,

the objective to assess the reaction, learning and mode of delivery of different components was reached. Secondly, while the study relied on non-validated questionnaires for evaluating training outcomes, this approach was chosen due to the need for tailored assessment measures that aligned with the specific learning objectives of the decision coaching training, a common necessity in educational intervention research [49]. Thirdly, even though the eligibility criteria were nurses working in a medical domain with existing PDAs, not all nurses had access to a PDA. For those not having access to a PDA, it may have created a feeling of not having the available support to implement decision coaching. The main limitation of the study was the overall lack of participation in Part B and thus, the missed opportunity to evaluate participants' decision coaching sessions in their clinical practice using audio recordings. On further inquiry, the nurses revealed other barriers that had prevented them from completing the module.

From a logical point of view, it would be reasonable to distinguish two main reasons why the nurses did not pursue the second, practical part of the training. First, barriers to implementing decision coaching in clinical practice and second, barriers specific to completing the training programme, particularly obtaining individualised feedback through audio recordings.

Regarding implementation barriers, although the decision coaching module was designed to overcome known barriers from the literature, there continue to be obstacles to nurses using it in practice. Research has identified common barriers that can impede decision coaching implementation at three levels: the patient level [50], the organisational level [16, 51] and the healthcare professional level [16, 52]. For example, not feeling one has the necessary information about all the alternatives, lacking communication skills, or holding strong opinions about what is best for an individual patient are potential barriers addressing nurses.

The "Ready for SDM" training modules address many of these barriers by giving them a voice in the classroom and by equipping participants with reasonable counterarguments. However, some of them cannot easily be remedied. Implementation of SDM is still constrained by information on benefits and harms not being accessible enough in a form that facilitates informed decisions, such as patient information following criteria for evidence-based health information or PDAs.

In Norway the development of decision aids is supported by the government [53] and several PDAs have been published on the central health platform “helsenorge.no.” However, the implementation of PDAs requires involvement from the clinical environments and an effective strategy for their systematic use [54]. The existence of high-quality patient information alone might not help if procedures for its use in the clinic are not in place and agreed upon. In the current study, existence and availability of a PDA suiting the tasks in the local environment was made an inclusion criterion. However, if these were not yet part of routine care, the nurses might have seen their implementation as part of decision coaching as too challenging.

Indeed, the lack of self-confidence in using decision aids does not fully explain the failure to adopt the behaviour, neither was it mentioned in the post-hoc inquiry assessing why Part B was not completed. The responses in the post-hoc inquiry revealed a collection of barriers. The startling contrast between the initial enthusiasm, positive attitudes and self-confidence and intentions to adopt decision coaching expressed in Part A and the actual behaviour later on needs to be better understood. The item evaluating the degree of self-confidence regarding PDA use showed that 50% only partly agreed and 31% disagreed. However, subjectively reported attitudes and intentions do not necessarily represent the active agents underlying subsequent behaviour.

Regarding training programme barriers, the primary barrier was participants' reluctance to audio record their consultations. This emerged as a major barrier to completing Part B of the training programme and measuring the coaching behaviour in order to give individualised feedback. To the extent that practical training and transferring knowledge to one’s own workplace is considered an essential part of the training, the barriers towards the sharing of insight into participants’ own decision coaching skills might in fact be seen as barriers to the implementation of SDM. The reluctance to share audio recordings warrants careful interpretation.

While it might be an indicator of the nurse's lack of readiness to adopt a new role as decision coach, the qualitative feedback from our post-hoc inquiry and previous research by Lenzen, Daniëls [52] suggests that practical and psychological barriers may be more significant factors. A basic requirement for decision coaching is establishing transparency and reducing power imbalances between healthcare providers and patients, such as the information monopoly on the side of the nurse and any kind of opacity [55]. However, as our study was not designed to investigate these particular attitudes, we do not know whether and to what extent this kind of barrier might have caused the dramatic dropout rate.

In the post-hoc interviews, aiming to elucidate the unexpected dropout rate, some nurses reported not pursuing the training because they were uncomfortable with audio recording their encounters; others reported lack of motivation as a key factor. These statements might also imply a lack of willingness to receive the following individualised feedback. Giving feedback on a video shared by the trainee is a well-known method and a BCT has the potential to impact positively upon building SDM skills [56, 57]. Golembiewski et al. claim that using recordings of consultations can “yield important insights about clinical practice without the inherent biases involved in asking patient or clinician participants to report the content, reflect on it, or rate the encounter retrospectively” [58]. Several studies report positive effects of audio/video feedback sessions on skill building [58‐62]. Indeed, participants in an SDM training study by Ammentorp et al. indicated that the most beneficial aspect of the course was reviewing their own videos [60]. However, recording consultations can also be experienced as stressful. While the method is widely used in training physicians [56, 63], it seems less common in training nurses. A Norwegian study investigated making videos of consultations in the field of contraception counselling and found that the stress experienced by public health nursing students depended on the degree to which they felt well prepared professionally [64]. In a study conducted by Eeckhout et al. [59] evaluating the feasibility and acceptability of video recording during real-time patient encounters performed by general practitioner trainees, almost 70% of the trainees were positive about recording consultations and almost 90% noticed an improvement of their communication skills through observation and evaluation [59]. Nevertheless, as the nurses in our study, they felt uncomfortable and stressed due to the presence of the camera and assumed that patients felt the same way. This is supported by Dobber et al. who reported that the recording of conversations may be anxiety-inducing for the nurses [65]. Video feedback has also been used in SDM training for nurses. In a study by Rahn et al. [33], nurses stated that being videotaped during coaching sessions was burdensome and a study by Berger-Hoger et al. [19] was terminated partly due to the nurses not complying with the task of taking videos.

Using simulated patients could be an alternative to audio feedback [66], with the potential to overcome some of the reported barriers in our study, such as the lack of “suitable” patients or worrying about the patient’s anxiety. Previous studies evaluating the training for decision coaches’ behaviours used simulated patients with the audio recording conducted by the simulated patients [67]. However, this method is resource-consuming and does not provide a real-world SDM example. Other alternatives are to extend the training in order to practise audio recording; this may contribute to the delivery of an audio recording no longer representing an insurmountable exercise, or using self-assessment as a method of feedback [68]. Considering the positive evaluation of the classroom training and the subjectively perceived gain of skills, self-confidence and solid attitudes regarding patient involvement, an argument could also be made for omitting the audio recording. However, as discussed earlier, we would then lose an essential, well-documented training method [69].

Results from the present study indicate that the classroom training intervention was feasible and acceptable and the nurses’ reaction towards the training was good. Self-reported learning including attitudes, intentions and confidence regarding decision coaching was overall positively reported by the participants, which are the predecessors of actual behaviour [70]. Measuring Kirkpatrick’s level one, reaction and level two, learning is consistent with other pilot and feasibility studies, also within the field of SDM [49]. Evaluation results indicate a clear progression through different competency levels. At the knowledge level, 90% of participants understood SDM concepts and 84% grasped decision coaching principles. However, skill development showed variability given 74% felt confident in decision coaching and only 16% were fully confident using patient decision aids [71]. This suggests that while foundational knowledge is well-conveyed, advancing to complex skill application is more challenging. Participation was limited in Part B, highlighting difficulties in measuring/verifying higher competency levels during practical application.

In our study, we trained nurses from different units, which may have influenced our results. The majority of decision coaching studies are conducted with homogeneous groups [11], but our study stands out in this respect. Perhaps this is one of the challenges to success? Two reviews have previously stated that the interprofessional nature of SDM should be acknowledged for optimal implementation [7, 71], meaning that training whole departments, units, or teams may be more useful than training individuals from several medical domains. Results from our study also emphasises the importance of being more than one participant from one department. A realist review aimed to understand how and under what circumstances decision coaching works for people making healthcare decisions by Zhao et al. [16] reports that “Healthcare providers are likely to improve their knowledge and skills to implement decision coaching through training and practice”. This is supported by twenty-two papers that explicitly described the training for decision coaches. This underlines that others have succeeded with training healthcare professionals in decision coaching, some have even proven effective in increasing patient participation in SDM and informed choices [19]. Further research is needed to optimise training for nurses in decision coaching to reach our end goal of improving the quality of healthcare decisions.