Implementation of the three good questions—A feasibility study in Dutch hospital departments

Abstract Objectives To determine the feasibility of pragmatic implementation strategies for three good questions (in Dutch: Drie Goede Vragen; 3GV. What are my options; what are the risks and benefits related to these options; and what does this mean for my situation?) to increase shared decision‐making (SDM) efforts in Dutch secondary care, and identify barriers and facilitators of implementation. Methods Convergent mixed‐method design: pre‐post surveys with patients attending one of six clinical departments in a Dutch Hospital, post‐intervention interviews with patients and health‐care professionals. Primary outcomes: feasibility (reach, use of 3GV). Secondary outcomes: SDM, experiences with 3GV and decision making. Interviews focused on barriers and facilitators of 3GV use. Interviews were content coded and categorized into determinants of behaviour change. Results 35% of the respondents who had heard of 3GV (52%) used all three questions. 3GV use did not lead to more SDM (SDMQ9 M = Δ0.3;SE = 2.2) but patients felt empowered to decide (88%) and to SDM (86%). Barriers were as follows: time investment, other SDM projects and perception that the need to use 3GV differs per patient/consultation. Respondents preferred to use 3GV as they saw fit for the consultation, instead of literally asking them. Facilitators: easy, accessible information materials that can be flexibly used. Conclusion Implementation of 3GV seemed feasible, although influenced by contextual characteristics (eg type of decisions, patients, on‐going interventions). 3GV contributed to important elements of SDM, and respondents were willing to apply them in a way that suited their situation. Practice implications We recommend continuation of current and new implementation strategies to enable 3GV implementation in secondary care.


| BACKG ROU N D
Shared decision making (SDM) for health-care decisions implies that patients, their informal caregivers and health-care professionals share medical information and information about personal preferences, in order to make a value-based and informed decision. [1][2][3][4] Several studies have shown the benefits of SDM with regard to reducing patients' decisional regret and conflict, improving quality of life and leading to better treatment adherence and more conservative care. 5 SDM is now a sine qua non of patient-centred care, but it has not been widely implemented yet in health care. [6][7][8][9] Multiple barriers and facilitators have been found related to implementation of SDM. [6][7][8][9] Important barriers for health professionals included time constraints, lack of applicability due to patient characteristics and lack of applicability due to the clinical situation. 10 Important facilitators included provider motivation to use SDM, its positive impact on the clinical process and its positive impact on patient outcomes. 10 For patients, important barriers were related to a lack of knowledge and the power imbalance in the doctor-patient relationship. 7 Several organizational characteristics have been identified as well as potential barriers, including organizational leadership, culture, resources, priorities, and teams and workflows. 9 Numerous interventions exist to improve the adoption of SDM  [14][15][16][17] Although no evidence exists that these three questions impact all aspects of SDM, some preliminary data on the benefits of the questions for certain components of SDM do exist. Moreover, the questions increase patients' awareness about options and about their possible role in decision making about options on the patient's side, as well as greater information provision and behaviour supporting patient involvement on the health professionals' side. 10,11,13 Also, research has shown that both patients and health-care professionals are generally positive about their use. 15,16,18 However, additional SDM interventions (eg training, feedback) may be needed for health-care professionals, to actually improve SDM during the consultation. 15,16 In 2015, the three questions as formulated in the MAGIC project 17 have been translated to Dutch (known as 'three good questions'/'drie goede vragen'/'3GV'), in a shared, consensus-based initiative by the Dutch Federation of Patient Organizations (PFN) and the Dutch Federation for Medical Specialists (FMS), and in close collaboration with a translation bureau specialized in plain language translations, after which they were published on the PFN website for patients to find and use. In addition to this implicit mass media strategy, additional implementation strategies seemed necessary to actually achieve a positive impact on SDM. Therefore, to implement 3GV in secondary care, we adopted implementation strategies adapted to local contexts of six different hospital departments of the Radboud University Medical Center. This study aims to (a) determine the feasibility of implementation of 3GV in order to increase SDM efforts in Dutch secondary care and (b) to identify barriers and facilitators of implementation.

| Design
For this pilot study, we used a mixed-method triangulation design, consisting of pre-post surveys and post-intervention interviews in five outpatient clinics (Departments) of the Radboud University Medical Center (Departments of Cardiology, Radiotherapy, Breast Cancer, Nephrology and Psychiatry) and one inpatient clinic (General Internal Medicine). The qualitative interviews were used to better understand the quantitative data as obtained from the surveys, using a convergent, sequential model in which qualitative and quantitative findings are analysed and interpreted separately and then combined to complement each other. 19

| Participants
Eligible participants were patients attending one of the six departments within one university medical centre, their health-care K E Y W O R D S feasibility study, patient communication, secondary care, shared decision making professionals (physicians and nurses) and the project manager.
Departments were selected based on their shown interest in implementing the 3GV. Patients had to be 18 years or older and able to sign informed consent. Each department defined their own target patient population (ranging from very specific groups of patients to every patient that visited the clinic) and which professionals participated. Hence, we invited (a) patients who visited the outpatient clinic (department of psychiatry), (b) patients who were admitted to the inpatient ward (department of general internal medicine) or (c) patients who visited the outpatient clinic for the first time (departments of cardiology, radiotherapy and nephrology). Patients with breast cancer were only invited if they were facing a treatment decision after being diagnosed. Different groups of patients completed the preand the post-questionnaires and participated in the interviews.

| Strategies to implement 3GV
Implementation strategies were as much as possible embedded in standard care and current improvement initiatives, which differed per department; hence, implementation strategies differed too.

| Implementation strategies for patients
1. Information brochures ( Figure 1) were given by a nurse or resident during the consultation as part of an information package for all admitted patients or sent to the patient's (all patients) home before the consultation together with the confirmation letter for the appointment (Table 1).

| Implementation strategies for professionals
For each department, a tailored implementation strategy was designed (Table 1), which could include:  Note: Implementation strategies: (a) A general introduction meeting about SDM for all professionals (typically 30 min, during a regular staff meeting); (b) A 60-min workshop to explain and train how to use the 3GV for health-care professionals that were involved in the implementation (workshop 1); (c) An informative session to increase awareness of SDM in general and 3GV specifically and to learn from each other based on practice cases (workshop 2); (d) Shadowing or video-taping consultations two consultations per health-care professional in which the 3GV were used by an observer who had extensive experience in SDM training, physician-patient communication and person-centred care. The observer was present in the consultation room and used a structured rating list, based on the OPTION-5. After each shadowing session, the professional received personal feedback. The observations were also used in workshop 2, to initiate group discussions between the professionals on their experiences, how to deal with particular situations and how to improve SDM behaviours. Abbreviation: HCP, health-care professional.
a Data from interviews used.
the consultation room and used a structured rating list, based on the OPTION-5. After each shadowing session, the professional received personal feedback. The observations were also used in workshop 2, to initiate group discussions between the professionals on their experiences, how to deal with particular situations and how to improve SDM behaviours.
All strategies focused on practical aspects of implementation of 3GV and were easy to use in the actual practical context (hence 'pragmatic'). The exact strategy differed per department (Table 1).

| Primary and secondary outcomes
Our primary outcome was feasibility which was determined by qualitative and quantitative statements related to implementation of the 3GV,

| Data collection methods and measurement instruments
Data collection took place between November 2014 and July 2015, and consisted of quantitative surveys and qualitative interviews.
• A Quantitative pre-and post-implementation survey for pa-

| Sample size
We aimed for 25 completed surveys per department and per measurement moment in order to reach sufficient power for our measures of SDM (as per Ref. 21,23 ). In the breast cancer department, no post-intervention questionnaires were handed out because during recruitment for the pre-intervention questionnaires, it turned out to be too difficult to get patients with newly diagnosed breast cancer to complete a questionnaire due to the emotionally challenging period they were in. Instead, the most relevant items from the survey were included in the interview guide for this population.

| Quantitative surveys
All data were entered in SPSS. We performed descriptive statistics on all data (calculation of n, %, means, SD, medians, ranges). No statistical tests were performed other than bivariate correlations between asking the 3GV, SDMQ9 and CollaboRATE. Due to fewer participants in 4 out of 6 departments, data were analysed and presented for all departments together, instead of per department.
For the breast cancer department, we used data from interviews instead of post-questionnaire results with regard to their experience with 3GV.

| Handling missing values
Missing values were coded as such and hence automatically ex-

| Qualitative interviews
Interviews were audio-recorded and transcribed verbatim. Any identifying information was removed from the interview transcripts. Two independent researchers performed a framework analysis 29 starting with open content coding and subsequent thematic categorization of the codes into one of six determinants of behaviour change (ie beliefs and motivations, attitude, subjective norm, perceived behavioural control, intention, behaviour). 27,29 This enabled us to systematically assess the process of implementation, its barriers and facilitators. The analyses were supported by Atlas.ti software for qualitative analysis, to support coding and structuring the data. The quotes in the results section were translated from Dutch.

| Characteristics of the survey participants and consultations
Participating patients were on average 58 years old, and sexes were equally distributed ( Table 2). The majority of respondents (74%) lived together with a partner and/or with children.
Most respondents visited their specialist for the first time, except for patients who visited the department of psychiatry (

| Feasibility
More than half of the participants prepared (any) questions before the consultation (pre: 65% and post: 57%), and more than 80% actually asked them (pre: 87% and post: 83%). Additionally, almost all participants were encouraged to ask questions (pre: 96% and post: 95%) and had the feeling their questions were adequately answered   However, there were no differences between pre-and post-measures of SDM (not on single items either), nor were there correlations between asking the three questions and perceived SDM.

| Barriers towards use and implementation of 3GV
Participants Thirty-one patients and health-care professionals participated in the interviews, of which 12 patients, 10 physicians, 2 nurses and 7 project leaders. The majority of the participants (71%) were female (Table 3).

| Intention and behaviour (Use of 3GV)
The most important reasons for patients not to use the 3GV were because they felt no need for it, as the health-care professional already structured the information in a similar order as the questions or because enough opportunity was created to ask questions.
No, it was already clear from the structure that the doctor used in the conversation. She started discuss-  Other reasons for not asking the 3GV according to patients and health-care professionals were that patients did not want more responsibility in deciding, that they were too ill, or too pre-occupied with their diagnosis or with the overload of information they already received (so they did not notice the 3GV materials).

Beliefs and motivations
All participants believed the 3GV would lead to more patient involvement in medical decision making (Table 3). Health-care professionals reported to use the questions to structure the consultations; patients thought that the questions could create awareness under patients that they are allowed to ask questions and to be involved in decision making.
Yes, and I also think it [3GV] will lower the threshold to do so. To ask those questions. As if they [patients] are like "hey, […] it is allowed. It is allowed to ask questions. [Patient]

Attitudes towards the materials
Most respondents were positive about the information materials. They thought that the 3GV were easy to understand, useful and simple but powerful. They thought the layout of the materials was attractive.

Perceived behavioural control
Organization of care: outpatient clinics versus inpatients wards. In outpatient clinics, patients have time to prepare themselves before the consultation and can bring a loved one to the consultation, who can support the use of 3GV. As the consultation time per patient in outpatient clinics is short, the 3GV can be used in preparation for the consultation which saves time in the consultation and increases efficiency. Additionally, in outpatient clinics many decisions are made, many types of questions can be asked, and patients often see the same health-care professional at each visit.
On inpatient wards, patients may have more decision moments and more frequent opportunities to use (and practice with) the 3GV, compared with the outpatient clinic where an appointment has to be made first. Also, nurses can help patients prepare to ask their physician the 3GV.

Intention and behaviour (Use of 3GV)
Sometimes, the 3GV were explicitly used in the consultation, but more often they served to structure the conversation.

Suggestions for improvement of the implementation strategies for 3GV
Not all professionals were clear about their role in applying 3GV. The [physician about shadowing] The moment that the 3GV materials were spread was positively evaluated by both patients and health-care professionals.
The moment allowed all patients to use 3GV in preparation of their consultation, either when materials were sent to their homes or when they were available in waiting rooms/at the receptionist desk. However, when 3GV brochures were sent to patient homes together with their appointment letter or the information package, some patients did not notice the 3GV materials between all other information. It was suggested to make the materials available at the hospital's information desk and other places in the hospital, as well as online.
Patients and health-care professionals mentioned to seek for possibilities about how to improve the reach and actual use of 3GV. It was also mentioned that in supporting people to prepare for the consultation with 3GV, informal caregivers should be better involved.
So, I can recommend it [3GV] to all, without question.
Even though some healthcare professionals will always have objections such as "as long as it does not bring more work" or "as long as I do not have to make more notes". That is the overall concern that you hear everywhere when something new is being implemented. But if I was a patient, I would benefit from it, so I would say "yes".

| D ISCUSS I ON
In this project, we sought to determine the feasibility of pragmatic strategies to implement 3GV to increase shared decision-making (SDM) efforts in Dutch secondary care. This also includes the identification of factors that act as barrier or facilitator in the implementation process. We found that implementation of 3GV created awareness for health professionals and patients about the possibility for patients to ask questions, but that only few patients used the 3GV, and that use of the 3GV did not lead to more SDM in the Second, in line with other literature 32 , the availability of 3GV materials did not make all patients use the 3GV, partly because the availability of materials did not always lead to actual awareness of 3GV (many patients still had not heard of 3GV after implementation), but also because not all patients considered the questions necessary to get more information or a more active role in decision making.
For example, it was often mentioned that health-care professionals already provided structured information (thereby addressing the answers to the 3GV), which may be a result of the implementation strategies targeting the health-care professionals. Additionally, decisions were often made in earlier consultations, so the questions came too late in the care process. Such organizational issues should be well thought-through when implementing the 3GV.
Third, use of 3GV did not lead to more SDM as measured with validated tools. A likely explanation for this is that the 3GV focus primarily on information provision which is essential for decision making, and a prerequisite for other steps in SDM, but on its own it does not immediately lead to more SDM, as only few steps in the SDM process are addressed with these questions. Although we did not find differences on single items of the SDM-Q-9 either, which would indicate whether the physician paid more attention to specific steps of SDM, it is still possible that the biggest improvement after 3GV happened on a patient level. Indeed, one of the barriers for patients in SDM 7 is the fear to ask questions, which is well addressed by the 3GV, because the 3GV create awareness about the possibility for patients to ask questions and provide example questions. Hence, 3GV may still be an important step to increase SDM use in clinical settings.
Indeed, our findings show that 3GV helped patients to ask questions, to get informed and to be aware of their options. As well, 3GV made patients feel empowered to make a decision, get the feeling to be allowed to participate in the discussion about treatment options and make a shared decision. Indeed, post-implementation patients reported somewhat more active roles in decision making too.

| CON CLUS I ON AND PR AC TI CE IMPLIC ATIONS
In conclusion, pragmatic implementation of 3GV seemed feasi-

ACK N OWLED G EM ENTS
We would like to thank all participants, as well as the participating outpatient and inpatient clinics for their enthusiasm, and for all efforts that were done in light of this study.

CO N FLI C T O F I NTE R E S T
This project was initiated by the Radboud University Medical Center and the Dutch Federation of Patients' Organisations. All authors declare to have no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.