Ready for shared decision making: Pretesting a training module for health professionals on sharing decisions with their patients.

INTRODUCTION
While shared decision-making (SDM) training programmes for health professionals have been developed in several countries, few have been evaluated. In Norway, a comprehensive curriculum, "klar for samvalg" (ready for SDM), for interprofessional health-care teams was created using generic didactic methods and guidance to tailor training to various contexts. The programmes adapted didactic methods from an evidence-based German training programmes (doktormitSDM). The overall aim was to evaluate two particular SDM modules on facilitating SDM implementation into clinical practice.


METHOD
A descriptive mixed methods study using questionnaires and a focus group guided by the Medical Research Council Complex Interventions Framework. The training was provided as two different applications (module AB [introduction and SDM-basics] and module ABC [introduction, SDM-basics and interactive training]) with differing learning objectives, extent of interactivity, and duration (1 vs 2 hours). Groups of participants were recruited consecutively based on requests for health professional SDM training in university/college- and hospital-settings. By a focus group and a self-administered questionnaire comprehensibility, relevance and acceptance were assessed and qualitative feedback collected after the training. Data passed descriptive and content analysis, respectively. Knowledge was assessed twice using five multiple-choice items and analysed using paired t-tests.


RESULTS
In 11 (six AB and five ABC) training sessions, 357/429 (296 AB and 133 ABC) eligible nurses, physicians and health professional students with varying clinical backgrounds and previous levels of SDM-knowledge participated. SDM-knowledge increased from 25-78% (range pretest) to 85-95% (range post-test) (P ≤ .001). The training was rated easy to understand, acceptable and relevant for practice. Findings to improve the education suggest higher emphasis on interprofessional teaching methods.


CONCLUSIONS
The two SDM training modules met the basic requirements for use in a broader SDM implementation strategy and can even improve knowledge.

Internationally, several SDM training programmes for HCP have been developed, but few have been evaluated. An international environmental scan, 4,5 identified a total of 148 programmes developed in 18 countries targeting licenced HCPs in various levels of training. 5 Of the 148, 43 training programmes have been systematically evaluated, but only 37 evaluation reports are available.
And few of these studies measured efficacy of the educational program(s). 4,5 The programmes vary greatly in what training is delivered, the manner of delivery, and the length of time. In addition, the review finds a lack of detail and transparency with regard to the didactic contents. 4,5 In a recent update of the review, increasing activity in producing SDM trainings is indicated, however, the level of evaluation is still poor and interprofessional approaches to SDM is missing. 4 Patients and policy makers in Norway are calling for SDM. The Norwegian term for SDM, "Samvalg," was introduced in 2014 6 with strong support from patients and the Ministry of Health published a series of documents indicating the need for better use of SDM. [7][8][9][10] In addition, the Ministry of Health established a set of criteria for judging the quality of patient decision aids (interventions for facilitating SDM) to be published on the national platform helsenorge.no. 11 However, there is no proven effective SDM training for HCPs available in Norway. 5 Given full implementation of SDM in the health system requires use by all professions across the various settings, a comprehensive training curriculum is desired capable to adjust to varying needs and conditions as time, setting, or profession. The findings from the environmental scan on SDM trainings suggested the need to embed any newly developed training within The Medical Research Council Complex Interventions Framework. 4,12 This framework provides recommendations for structuring the development process of a complex intervention by five steps of evaluation on guiding the continuum of increasing evidence. 13 The first step is to make use of existing empiric and theoretic evidence. Further steps combine qualitative evidence from modelling studies with quantitative evidence from experimental studies to build an understanding of the mediating mechanisms, which makes efficacy traceable under implementation conditions.
There were a few interesting programmes to choose from, but the doktormitSDM training programme (see Table 5) had been rigorously evaluated and appeared to be a good programme to be used as a starting point of the development of the Norwegian curriculum (Norwegian: "klar for samvalg", English "ready for SDM").
This approach had been proven effective to improve communication in terms of SDM. According to several studies, it is adaptable for various settings (eg, one-on-one and self-studies), media (eg, web based), and time frames (eg, from 15 minutes to several hours). 5,[14][15][16][17] Rather than rebuilding the entire training, the Norwegian development was supposed to adopt elements of dokt-ormitSDM on the level of the didactic principles (see Table 5).
Unlike doktormitSDM, which hitherto has been used for physicians only, the new development should take into account the needs of a much broader professional target group by providing a variety of modules.
The overall aim of this study was to pretest the two first group training modules of "ready for SDM." Within the context of the entire curriculum portfolio for Norwegian HCPs, these modules are intended to introduce SDM and related basics as a door opener. Specific objectives were to (a) evaluate the modules' feasibility of applying to various target groups including interprofessional groups; (b) determine the training module's comprehensibility and acceptability; and (c) as a preparation for further evaluation, steps in terms of the Complex Interventions Framework, pretested the modules' ability to achieve a gain of knowledge.

| Research design
Using mixed methods, the descriptive study pretested two applications of the training module "ready for SDM/ group": AB (introduction to SDM for basic didactic training) and ABC (introduction to SDM for basic didactic training plus interactive training) ( Table 1). The study design was informed by the Kirkpatrick's model which structures evaluation using four levels: reaction, learning, behaviour, and results. Our study focused on the first two levels. More specifically, the evaluation comprised (a) a post intervention questionnaire to explore comprehensibility and acceptance, (b) a post intervention focus group to explore need for revision to the training programme, and (c) a questionnaire based pre-and post-knowledge test. The study was by no means designed to compare the two modules or compare participant experience across different settings. On the contrary, the study was designed to test the training modules under as much as possible usual continuing education type conditions. The study was approved by the ethics committee at the University Hospital in Northern Norway (UNN) (2017/1461). All participant signed a written informed consent form prior to their inclusion.

| The intervention
To meet varying needs of the Norwegian target group, the intervention was presented as two modules (AB vs ABC) which differed in duration (1 vs 2 hours), objectives, and extent of interactivity (Table 1). The typical target group had a variable composition of professionals, including nurses, doctors and other HCPs, depending on the clinical environment, the professionals available to attend, and/or the invitation received for providing training. Decisions about which of the two applications to apply were made based on the particular request and available timeframe. Although there was some variation due to local contexts, each module was provided mostly unchanged and by the same trainer (SK). The trainer was a registered nurse with a master's degree in Health and Empowerment and a Ph.D. student focused on SDM.
The 1-hour module (AB) is targeted at improving attitudes and knowledge about SDM. The first component (A) provides an introduction into SDM using a didactic lecture, the second (B), is a basic SDM lecture on the concept, related measures, and steps to realize SDM.
The 2-hour module (ABC) additionally focuses on improving SDM skills, particularly by the third component (C) involving interactive skills training such as, for example, role play.
Level and objectives were adjusted to the actual stage of implementation in Norway 18 and were considered responsive to HCPs current needs and a meaningful first step to approach a higher level of training using complementary modules such as doktormitSDM.
The "ready for SDM" training module follows a detailed curriculum, which is developed in line with doktormitSDM. 15 "Ready for SDM" uses didactic concept and knowledge of the German programme. In particular, all essential didactic strategies underpinning doktormitSDM (

| Setting and participants
Aiming at maximizing ecologic validity, the sampling strategy allowed for gathering experiences with application of the two training modules in a variety of settings and groups, under conditions, which were, as far as possible, representative for usual education in Norwegian health-care institutions and educational institutions. Initially, a matrix structuring the scope of health professional training in Norway, comprising educational settings: undergraduate (Bachelor), graduate (Master), continuing education (University/College) and in-service education (Hospital) was created (Table 3). Additionally, groups of participants who contacted The South-Eastern Norway Regional Health Authority with any request for an SDM training were consecutively sampled into the given matrix, if eligible. Decisions regarding their eligibility were made considering the following criteria: (a) the requested lecture was supposed to address SDM and allow for conduct of the entire intervention, either module AB or ABC; (b) the organizers agreed to use the training for evaluation purpose, implying data collection before, during and after the meeting;

| Data collection
Data were collected at baseline (attitudes towards SDM, previous SDM skills, knowledge) and post-training (demographics, comprehensibility, acceptance, needs for revision, knowledge).
Attitudes towards patient involvement, operationalized as a health care professional's willingness to apply SDM in clinical practice, and a subjective estimate of the participants' training level were surveyed in the context of an online quiz (5-point Likert scaled, from "to a very little extent" to "to a very great extent").
The same online quiz included questions assessing SDM knowledge with a multiple choice questions (5-items). To stimulate interest in the training module at the beginning, feedback on group level statistics was directly provided on a big screen. After the training, knowledge was reassessed using the same five items included in the paper-pencil questionnaire. These items are a test, previously piloted, and in use for certifying participants of an SDM e-tutorial. 14  reliable sources of information about effects of medical interventions.
The original set of questions were provided as a supplementary file.
The post intervention questionnaire also included four items assessing demographic characteristics: years of age, sex, profession, and years of professional practice.
Comprehensibility (1-item) and relevance of the course module to clinical practice (1-item) were assessed as subjective ratings of given statements on a Likert scale ranging from "very little extent" to "very great extent"; the item was to be answered with regard to each of the three components (A, B, and C) as part of the post intervention questionnaire.
Using the same answering format, acceptance was assessed by two items: one asking for willingness to recommend the training to others and the other providing a statement with an overall impression regarding the mod-

| Analysis of knowledge test
Answers from the pre-and postknowledge test were dichotomized to either "correct" or "incorrect." Paired t-tests were conducted for each of the five knowledge items to test knowledge gain during the training for significance (alpha <.05) Adjustment of alpha due to multiple testing was considered unnecessary since the items are theoretically independent.

| Analysis of quantitative data from the survey
Data from the 10-item post-training paper-pencil questionnaire were calculated using frequencies and either reported as percentages (pro-

| Analysis of the qualitative data
Data collected using free text formats and data from the focus group meeting were analysed based on principles of qualitative content analysis as described by Hsieh and Shannon. 19 Data extraction and analysis were undertaken by two independent researchers using the The participants were registered nurses (46%), nursing students (33%), other students (5%), physiotherapists (3%), occupational therapists (1%), physicians in or with specialization (6%), and 6% other professions (psychologists, social workers, health care assistants) (see Table 3). Age was in mean 35 (SD = 13) years, 91% were female and reported duration of clinical practice was in mean 11 (SD = 11) years.
The focus group included a physician, a nurse, a social worker, and an occupational therapist (age range 44-63 years).

| Qualitative feedback on the training modules
An overview of a priori categories and new subcategories is given in

| Limitations and Strengths
The study is limited by potential selection biases. For example, only 251 of the 357 who were trained (70%) participated in the study and completed data collection for at least one premeasure/postmeasure. This may have led to an overestimation of the preimprovement/postimprovement in knowledge, since it may be possible that participants who felt unsure did not respond to the post knowledge test. However, such a selection bias seems unlikely given most of the drop-outs were due to initial technical problems or poor technical insight when trying to access the online questionnaire.
Self-selection due to high motivation towards SDM might have caused ceiling effects regarding SDM-related attitudes and overall feedback on the training. 20 The group meetings were pre-organized by the clinical leaders and largely mandatory for the staff to attend.
Given that, a small number of participants refused to provide informed consent for the study, we do not consider the self-selection

Category Findings Example quotes
Presentation Linguistic Presentation The six steps structure of SDM is traceable The (six steps) approach was a structure that is easy to remember and the related memo-card were also very useful.

Need for a clear differentiation of types of user involvement
I need a better explanation of user involvement and SDM.
The structure of the presentation contributes to understanding.
The topic was described in a way I have not thought about before and I consider helpful for my own future practice. This was a structured presentation that has increased my understanding of the topic.

Presentation slides
The slides (shape, graphics, diagrams) contribute to better understanding.
Simply outlined-easy to understand Nice and comprehensible presentation. Very good presentation, technically/graphically. I liked the diagrams shown on the slides.
Size of the font is too small.
To me the font was sometimes too small and difficult to read because of the colour (not black).
Too high complexity. I would prefer less information by slide? Difficult to hear and read at the same time.

Completeness and acceptability
Need for more information on recourses, time, economy and positive effects.
It was maybe too little reflection on resources needed (time and economy) for SDM in a long term perspective. I think it would be important to present more about positive effects of SDM, a good thing that SDM is an ethical right.
The module provides increased awareness on the structure of SDM.
I became more aware of how structured SDM can be performed. The examples given were good, but I missed some simpler examples (not only big and crucial decisions).

Structure Variation of communication modes supports attention.
Natural variation (of the contents) is good. For someone like me who cannot be concentrated for a very long time -you made it work. Good with different elements in the content.
Composition and balance of didactic elements is appreciated.
Well-structured content, clear dissemination, questions and participation from the audience were allowed and encouraged.

Suggestion of small-group work to facilitate translation into clinical practice
It might be useful to have small conversation-groups during the presentation, to get it down to a more practical level. There were no 1 on 1 conversation groups.

Adaptive capacity Adaption to profession Suggestion of using domain specific video examples
The course would gain from presenting more subject-specific examples. E.g. use of videos, especially videos showing real SDM situations would make it more alive. Maybe video instead of the role play?-the optimal would have been a video in a recognizable environment, this would make us extra stimulated.
Adapt examples to professions/disciplines. It seemed mostly physician-focused. Little about nurse specific tasks. Could have been more examples from our everyday clinical settings. I would like, more adjusted to psychiatric health care. Can you adopt more relevant cases from neonatal medicine?
The content could be more customized to different professions.
The module is comprehensible regardless domain and profession.
I found that SDM was comprehensible to everyone despite that we were different health professions participating in the training.
The module needs more focus on interprofessional SDM.  Adaption to level of knowledge Differentiation to the level of knowledge Your vocabulary could be more customized to different professions. The vocabulary wasn´t enough customized to 2nd year students. I guess a group of physicians will probably understand all the terms, but maybe not all in mixed group with e.g. nurses and health economists.

Time/scope/setting
Specific proposals for amendments regarding time Suggestion to reduce information density It goes a little too fast, easy to fall off during the second half of the presentation, talk a little slower or maybe add another hour. Should have had some more time? A lot of material in a relatively short time.

Suggestion to increase time to facilitate reflection
Perhaps more time for reflection, feedback and small conversation-groups during the training. Spend some more time for the first and second part, to have more time for discussion. More time would have been useful, and more time for reflection and discussion.

Suggestion to increase frequency of short brakes
More time for reflection. A break every 40 minutes, which is the standard norm for how long a person´s attention can be kept.

Prioritizing of components Suggestion to increase time for interactive training
Interactive training-there could have been more of that.

Suggestion to increase time for live feedback
There could have been more time on interactive live feedback. Both for the analysis of the role-play performance and for the moderator so that the patient and the coach can share their thoughts without getting interrupted. Rather more role-play than examples.

Various comments regarding balance between component A, B and C
The basics module (A) was very useful, somewhat as frame before going more into depth, and the six-steps to SDM including the card to look at worked very well. Spend some more time for the first and second part, to discuss a little more.

Interactive online quiz Facilitation
Suggestion to increase time for quiz I experienced that the audience was in need of more time for the smartphone based interactive questionnaire and that this became a bit stressful.

Method
Appreciation of didactic using Smartphonebased quiz The Smartphone-based quiz was a nice and fun interactive learning method. Up-to-date and inspiring with use of smartphone-based methods in the training. The smartphone-based questionnaire gave a nice overview of what everyone thinks SDM is about.

Language
Need for simplification of the language The quiz had some difficult wording. Some of the questions were maybe unnecessary difficult to understand, a better description might help. Some concepts should be simpler and more thoroughly explained, ex. cohort studies, RCT etc..

Specific proposals for amendments Comments defining concrete need for revision
The question about SDM competence must be specified. Is it about my level of knowledge, or is it about how I actually choose to meet/treat and include my patient in decisions? Some questions seems ambiguous, for example, only one choice. Then I think about deciding between treatment or no treatment. Treatment is a choice?
Interactive role-play Facilitation Increase time for interactive training I found the evaluation of the interactive training too little critical. The role-play was insensitive and unrecognizable, and I missed that this was discussed/ problematized. There could have been spent more time on interactive live feedback. Both for the analysis of the role-play performance and for the moderator so that the patient and the coach can share their thoughts and without getting interrupted.

Method
A pre-rehearsed role-play with teachers might be better. It is difficult for nursing students to take this role offhand. I did not like that people almost were demanded up on stage for the role-play.
Role-play is difficult when knowledge about the topic is limited. A video about a real situation might be better. Good to be challenged to participate in active learning SDM was well illustrated in the role-play towards the end of the presentation.
Note: This table presents the qualitative findings structured in main themes/categories, findings and example quotes from the free text answers provided in the evaluation questionnaire and from the focus group. bias to be very strong. The recruitment strategies were also used in an attempt to achieve a realistic setting, by taking into account the natural process used when organizing training.
The underrepresentation of physicians (n = 12, 6% participants) might be seen as problematic given physicians' key role in medical decisions. Due to the educational contexts, physicians were not invited to all the trainings and -because of other clinical dutieswhen participating more likely to attend only part of the time. However, physicians' participation in our study was consistent with the nature proportion of physicians in clinical interprofessional teams.
Furthermore, the study is limited by lacking systematic variation of possible training settings and contexts. For example, the 2-hour training applied to an intra-professional physician group was not represented in the study design (Table 3). Therefore, we might have missed some important information on feasibility. However, orienting the study within requests for training, we have likely addressed the most important settings and contexts for evaluation of this SDM training in Norway, and thus, we did not find it necessary to conduct the training within all theoretically possible settings (Table 3).
At this stage of evaluation, we did not control for trainer effects.
A recent study suggested that a trainer's charisma is associated with a trainee's intention to apply the skills they learn in training. 21 In the current study, we received large amounts of positive feedback regarding the trainer (SK), for example, giving an "enthusiastic," "engaging," "motivational," and "trustworthy" impression. We understand this feedback as encouragement to seek to identify important key behaviours of the trainer that were successful in teaching others. 22 A key strength of this study was the ability to provide the training modules to a range of health-care professionals and students. The study does not claim generalizability to any other than the target group, evaluation on generalizability will be addressed by further studies within the "Ready for SDM" framework.

| Results in context
This training has been developed as part of a more comprehensive strategy to implement SDM in the Norwegian health system. The strategy is meant to be adaptive to the current status of implementation of SDM in Norway; therefore, the study focused on feasibility rather than efficacy of the SDM training. 18 As recently shown in a German study, 23 a programme might prove efficient, but due to lacking feasibility nevertheless fail to demonstrate effectiveness. To keep a realistic chance of making an impact on the health system, a training needs to be adaptable to various settings and timeframes. By its modular construction, the "ready for SDM" framework can meet these challenges and appears to provide a promising approach for training HCPs.
Several studies recommend training HCPs in SDM based on an interprofessional approach. 3,4 This is remarkably important as other professionals, in addition to physicians (eg, nurses), play an essential, and so far undervalued role in using SDM in clinical practice. 24 Also, patients point out the value of other HCPs' (non-physicians) participation in SDM. 25,26 However, participation of non-physician HCPs in tasks related to making health decisions, implies the need for restructuring proceedings on the patients pathway and more clearly describe the interprofessional team working with the patient. 27,28 Therefore, an interprofessional team-based SDM training is suggested to support HCPs in legitimizing and using SDM in their practice. 29 Training interprofessional groups or clinical teams in interprofessional SDM can involve mixed interactive didactics such as role-play or communication exercises. 30 Learning can also be facilitated by using domain-tailored video examples showing SDM performed by different HCPs. However, successful teamwork on the common aim to facilitate informed decisions might require additional efforts in understanding local barriers and cultures that need to be addressed. 31,32 The findings in our study are consistent with recent reviews on programmes developed to train SDM skills in HCPs, which conclude in claiming more systematic evaluation and adaptation to interprofessional practice. 4,5 The "ready for SDM" approach complies with both recommendations. The modules investigated in the current study build upon strong evidence from studies evaluating previous corresponding modules. Beyond comprehensible, acceptable and feasible, 15 the doktormitSDM module has even proven effective with regard to communication quality. 17 Although presented as various applications and having been used in multifold medical domains, dokt-ormitSDM is still focusing on physicians. The current module opens up for interprofessional settings, facilitating an interprofessional approach to SDM. Our results will inform further revisions of the training modules to better meet this aim.
Another recommendation for implementation of SDM is relating to the broader perspective of making use of different kinds of interventions. 3 The best results are gained by using both patient directed interventions such as patient decision aids and those addressing HCPs in combination. While political guidance in the national health-care strategy in Norway is quite supportive with regard to implementation of SDM, [7][8][9]18 there might still be a pronounced need to better balance emphasis between decision aids and HCP SDM trainings. The increasing number of patient decision aids both internationally and in Norway [33][34][35] should be complemented by a national effort to systematically implement SDM training. 18 As we continue to develop the "Ready for SDM" training, we will consider designing further studies in accordance with the complex intervention framework and the Kirkpatrick's model. 12 This study has only evaluated the first two levels of the Kirkpatrick's model, but the intention is to make changes based on these findings and evaluate the other levels.
In addition to evaluating efficacy of the SDM training modules, our findings reinforce the need to be more aware of relevant barriers towards using SDM in clinical practice, 31 develop interventions to address the barriers, and continue to monitor barriers in Norwegian health care settings.

| CONCLUSIONS
The two first training modules of the comprehensive Norwegian SDM curriculum are approaching implementation via mixed groups of HCPs rather than mainly or exclusively addressing the decision makers amongst the clinicians.
The training in general, both modules and each component, were easy to understand, acceptable and relevant. Participants achieved improved knowledge of SDM. However, the training modules need further adaption to achieve an interprofessional approach to SDM. Our findings will inform revision of the two modules before they will be tested for efficacy in a randomized controlled trial.