I am ready to see you now, Doctor! A mixed‐method study of the Let's Discuss Health website implementation in Primary Care

Abstract Background Let's Discuss Health (LDH) is a website that encourages patients to prepare their health‐care encounters by providing communication training, review of topics and questions that are important to them. Objective To describe LDH implementation during primary care (PC) visits for chronic illnesses. Methods Design: Descriptive mixed‐method study. Setting: 6 PC clinics. Participants: 156 patients and 51 health‐care providers (HCP). Intervention: LDH website implementation. Outcome Measures: Perceived quality and usefulness of LDH; perceived quality of HCP‐patient communication; patient activation; LDH integration in routine PC practices and barriers to its use. Results Patients reported a positive perception of the website in that it helped them to adopt an active role in the encounters; recall their visit agenda and reduce encounter‐related stress; feel more confident to ask questions, feel more motivated to prepare their future medical visits and improve their chronic illness management. However, a certain disconnect emerged between HCP and patient perceptions as to the value of LDH in promoting a sense of partnership and collaboration. The main barriers to the use of LDH are HCP lack of interest, limited access to technology, lack of time and language barriers. Conclusion Our findings indicate that it is advantageous for patients to prepare their medical encounters. However, the study needs to be replicated in other medical environments using larger and more diverse samples. Patient and Public Contribution Patient partners were involved in the conduct of this study.


High-quality communication between patients and health providers
is essential for better management of chronic illness. In Canada, as in many other countries, much of such care occurs in primary care (PC). It is estimated that 40% of health-care visits in PC include a follow-up for chronic health problems. 1 The prevalence of multimorbidity in this context is high. [2][3][4] Both of these characteristics undeniably increase the complexity of encounters. However, the average duration of a health-care visit in this context is short, approximately 15 minutes, 5 and the majority of patients do not know how medical appointments are structured and so have difficulty preparing for them. [5][6][7][8][9] Helping patients prepare for their PC health-care encounters thus provides an interesting avenue to improve the effectiveness of chronic illness health-care visits.
Indeed, the quality of communication between patients with chronic illnesses and their health-care providers (HCP) is recognized as playing an important role in the management of their illness. [10][11][12][13][14][15] Yet, compared with epidemiological studies, relatively few studies have specifically focused on the interactions between patients and HCP during their encounters. 15 Data from research on communication reveal how important it is for the HCP to play the role of partner or facilitator 16 in order to arrive at a mutual understanding of the nature of the illness and to engage in a shared decision process on the best course of treatment. [17][18][19][20] In recent years, the idea of focusing directly on patients' communication skills instead of those of HCP to improve the quality of the encounter has emerged. 7,21 Patient activation (i.e. encouraging patients to take an active role in the management of their chronic illness) is one of the best ways to reach this goal. [21][22][23][24][25] The evidence indicates that patient activation improves participation during health-care encounters, recall of the information discussed, treatment adherence and other health-care outcomes. 7,15,21,22,26,27 Thus, activation can increase patient commitment to the self-management of their chronic illnesses. 23,24,28 However, the majority of studies on patient activation have required the intervention of a trainer before the patient's planned HCP appointment. This approach is complicated to organize, imposes a rigid schedule on the patient and can lead to considerable costs, all significant challenges to its adoption. Online interventions present a potential solution to overcoming these barriers. 29 Let's Discuss Health (LDH) is a French Canadian website that encourages patients to take an active role in managing their health and supports their collaboration with their caregivers. It is akin to a 'stand-alone' electronic personal health record (PHR) with which it shares the aim of allowing patients to hold their own health record including active diagnoses or problem list, medication list and personal and family medical history. A major distinction between PHR and LDH is that it is not, as yet, connected to the HCPs' electronic medical record (EMR). LDH does not allow information transfer from the EMR to the PHR nor does it provide an electronic communication channel between patients and providers in between clinic visits as in many patient portals. 30 However, it fills an important gap not covered in almost all PHR, including the Carnet santé Québec 31 in that it guides patients in the preparation of their visit agenda for time-constrained medical visits and in setting their priorities. 32 16,20,36 Patients with chronic illnesses and their HCP were consulted at every crucial step of LDH's development process and are described elsewhere. 37 Even though LDH was judged extremely useful and pertinent by a group of test users during a validation study, 37 this does not guarantee it will be adopted into PC clinical practice. The potential for integration of this tool in regular medical visits for patients with chronic illnesses needs to be assessed.
The goal of this study is to describe the implementation of LDH in PC clinics and its adoption during routine family medicine follow-up visits of patients with chronic illnesses. More specifically, it aims to: (a) describe patient experience of LDH website use; (b) identify, from both patient and HCP perspectives, the association between patient preparation with LDH and patient activation, as well as with the unfolding of the visit with the HCP; and (c) identify, from patients, HCP and clinic project coordinators' perspectives, the facilitators and barriers to the use of LDH tools in this clinical context.

| Conceptual frameworks
This study rested on two theoretical frameworks. The Reach, Effectiveness, Adoption, Implementation, and Maintenance framework (RE-AIM) [38][39][40] refers to the degree to which the intervention is integrated into current practices and established organizational policies, whereas the development of training tools for HCP and the conditions favourable to their implementation in clinics was based on the implementation of the organizational change model developed by Rondeau et al (2008). 41 This model presents the issues associated with change according to the organized action perspective.
Thus, Rondeau's model was used in conjunction with the RE-AIM framework for the study of the implementation and adoption of the LDH.

| Study design and participants
This was a mixed-method descriptive study to explore the use of LDH by PC patients with chronic illnesses. We enrolled 6 PC clinics who had demonstrated an interest in integrating LDH into their routine practices. These clinics were located in Québec (n = 3) and New Brunswick (n = 3), two provinces in Canada with French-speaking populations. Participating clinics varied in terms of geographic area (urban, semi-urban, rural), organizational structures, team and practice size and characteristics of the population served.
Most participating clinics, either academic or community-based, involved interdisciplinary teams, namely practicing family physicians, family medicine residents and nurses involved in the management of these patients. The invitation to participate in the study was thus sent to all these types of HCP. Anyone who was absent for a 6-to 12-month period during the implementation and data collection phases were excluded from the study. All six clinics identified a 'project coordinator', a well-established clinic secretary or nurse who was to be the local contact person during the course of the project. All patients aged 18 years or older, with at least one chronic illness and a scheduled follow-up appointment with a participating HCP during the study period (June 2015 to September 2016), were invited to participate. Other inclusion criteria were understanding, speaking and reading French and having basic Internet skills or having access to someone who did. Patients considered incapable of giving informed consent were excluded from the study.

| Study conduct
The study was conducted in four phases as described in Figure 1.
Here, we are reporting on the first three phases. First, a clinic 'pro- Patients were invited to participate in the study by clinic staff either by letter, phone or both. The invitation was adapted to the two models for clinic appointment setting that were in vogue at the time of the study. First, in the clinics with a traditional access model (where the HCP's schedule is completely booked in advance), we mailed a personalized letter signed by the attending HCP to eligible patients who had a scheduled follow-up appointment in the following 4 weeks. Two weeks later, a call was made by the project coordinator to check whether the patient received the F I G U R E 1 Study phase description letter and briefly discuss the advantages of preparing for an appointment. Second, in the clinics with Advanced Access (appointments schedule to be available in a timely fashion), the project coordinator called eligible patients to invite them to participate in the study, outline the nature of the engagement if they chose to participate and discussed the advantages of preparing their scheduled encounter. Participating patients were reminded to bring their 'visit preparation summary' to the visit. Patients signed their study consent form when they arrived at the clinic. In either model, the clinic's secretary called the study patients to confirm the upcoming appointment, remind them of their participation in the study and to bring their visit preparation summary. On arrival at the clinic, patients were referred to the project coordinator or the research assistant for all tasks related to data collection and study follow-up. Quantitative data were collected from patients, HCP and project coordinators.

| Measures and data collection
Patients were asked to complete two short self-administered questionnaires. The first, developed specifically for this study, asked for basic sociodemographic and clinical data, their impressions of the LDH website and its perceived usefulness for preparing their medical visit. The second questionnaire, the Patient Activation Measure (PAM 13), 28 is a validated instrument that assesses patient self-reported knowledge, skills and confidence in regard to the self-management of their health or chronic condition. It consists of 13 statements with the following answer categories: Disagree strongly, Disagree, Agree, Agree strongly or Does not apply. It was administered to patients after they visited the website and prepared for their appointments.
PAM scores were calculated according to guidelines provided by the PAM survey licence proprietor Insigna Health Inc. 42 PAM scores range from 0 to 100, with higher scores indicating greater activation.
Scores are categorized into four levels. Level 1 patients (score ≤ 47) are disengaged and overwhelmed; Level 2 patients (score 47.1-55.1) lack confidence and struggle to take action; Level 3 patients (score 55.2-66.9) begin to take action; and Level 4 patients (score ≥ 67) are maintaining healthy behaviours. Furthermore, project coordinators kept a study logbook that allowed us to document recruitment and provided information about patient participation rates. HCP filled the first self-administered questionnaire on their sociodemographic and professional characteristics. Immediately after the encounter with a study patient, they also completed a self-administered questionnaire to gather their impressions of the unfolding of the encounter, the value of the LDH summary sheet and its impact on the visit. This questionnaire was developed for the present study based on a similar questionnaire used in the Talking Health Together study, 15,43 a randomized trial evaluating the impact of a patient communication intervention delivered in two formats (e-learning or e-learning followed by a workshop) on patient participation in PC encounters, patient recall and patient health outcomes.
Project coordinators provided bookmarks, a form of LDH promotional material, to the HCP to distribute to eligible patients. We are not reporting on this phase of the study.
Qualitative data were collected from patients' and HCP's focus groups and semi-structured interviews with project coordinators.
The patient interview guide included questions designed to explore the user experience of the website and their experience of the follow-up appointment using the visit preparation summary. The HCP interview guide, based on the organizational change model, 41 included questions designed to explore their experience of consultations with prepared patients. All focus groups took place once the period of scheduled appointments for eligible patients had ended.

| Data analysis
Quantitative data: Descriptive statistical analyses were performed using SPSS software version 24. The descriptive data were obtained from the answers to the self-administered patient and HCP questionnaires completed following the appointments. An isolated analysis of each of the six clinics was carried out, and the findings were its implementation in routine clinical practices as well as its impact on the medical encounter (patient and HCP focus groups) as they emerged from the responses to the interview guide. The codes assigned were reviewed and discussed on a regular basis by this team of coders. In case of coding differences, a consensus was reached as to the thematic category that best applied. An individual analysis of each clinic was first carried out, and then, the six clinics (cases) were aggregated to identify facilitating factors and barriers to implementation. To ensure the best interpretation of the data, all the researchers, collaborators, partners (patients) and volunteer clinicians from participating sites were invited to review the analyses and thus play played a key role in the analysis process.
For data integration, we collated the data from quantitative and qualitative analyses in order to identify the similarities and the differences between the results.
Ethical approval was obtained from the Ethics Committee of the participating clinics in the two provinces. All participants (patients,

F I G U R E 2
Applying the RE-AIM model to the study HCP and project coordinators) gave written informed consent before taking part in the study.

| Clinics and participants
In total, 156 patients participated in the study (Figure 3), 106 were from Quebec and 50 from New Brunswick. Overall, patient participation rate was 16.5%, ranging from 7.9% to 37.3% per clinic depending on the onsite presence of dedicated research resources such as medical students or family medicine residents (data not shown). A majority of these patients were over 40 years old (84%), and 79% reported higher education levels (high school diploma). Just over half of the participants had 3 or more medical visits during the preceding year (54%).
Across all sites, 51 HCP participated in the study (range: 5 to 13 HCP per site). There was an equal representation of both novice (51%) and experienced HCP (49%). Patients and HCP sociodemographic characteristics are provided in Table 1.
By comparing the two provinces, we found that the sociodemographic characteristics of the patients were very similar with regard to age, sex and income. However, there was a slight difference in the level of education. Indeed, 29% vs 36% (high school or less) and 71% vs 64% (College/Technical training/University), in Quebec and New Brunswick, respectively.
In what follows, we report patients and HCP results, of the self-administered questionnaires followed by the focus groups discussions. Presentation of results follows the same pattern in Tables 2, 3 and 4 in which we first report results from the patients' self-administered questionnaires followed by the HCP's responses.
We have organized the presentation of results following the four major themes that emerged from the data analysis:

| Perceived value of visit preparation and summary sheet
Patients agreed that the LDH visit preparation summary generated by the system was helpful to manage their medical appointment.
They stated that the summary sheet was complete (92%), useful (90%) and that they discussed all items listed on it with their HCP (99%) ( Table 3). During patient focus group discussions, a consensus emerged that LDH was a helpful tool to avoid oversights, reduce the stress related to a medical visit and increase patients' sense of control as well as to help them structure their visit. Thus, findings from the patient focus groups confirmed questionnaire results.

| Let's Discuss Health users' experience and its impact on the visit
In view of the questionnaire data, patients and HCP were generally satisfied with LDH and that LDH contributed to improved communication during the medical encounter (    basic implementation assistance that included clinic PHR champions, HCP and staff training, patient PHR adoption rate reached 25.6%. 47 The practices in our study that involved medical students and residents in the implementation process thus provided a form of 'implementation assistance' probably accounting for the increased uptake of LDH by patients in two of the practices. As expected, fewer patients with lower education status responded to our invitation to prepare for their encounters using LDH, a form of health IT. 46 Poor computer skills, lack of Internet access, low health literacy and limited physical and cognitive abilities all contribute to what has been coined the 'digital divide'. 48  reported that patients presented with clear reasons to consult and that the LDH visit preparation summary was easy to integrate into routine visits. However, a certain disconnect emerged between HCP and patient perceptions as to the value of LDH in promoting a sense of partnership and collaboration. Some HCP did not feel that the visit preparation summary provided new information or that it helped organize the visit. It was as if their expectations of the role of the summary were not met. A few HCP commented for example that the visit preparation summary did not provide information on important 'red flags' in relation to reported symptoms. This could explain the absence of 'perceived added value', of the visit summary expressed in the HCP focus groups. In this regard, we would like to point out that the summary sheet was conceptualized not as a 'mini' case history useful for the HCP but as a tool to help the patient clarify their own agenda for the encounter and organize the information, he/she wished to share with their HCP. Moreover, the concept of 'agenda setting' at the beginning of any HCP-patient encounter is recognized as the keystone of an effective medical encounter and well-aligned consultations. 51,52 This is true for initial and follow-up visits in the long-term management of these patients as recognized in Wagner's Chronic Care Model. [52][53][54][55] The LDH was created to help patients prepare, organize and prioritize topics for discussion for their upcoming visit thus enabling them to engage in a collaborative visit agenda setting with their HCP. 56,57 These aspects do not seem to have been considered important in the opinions expressed by the HCP.

| D ISCUSS I ON
During the focus group discussions, HCP were divided with regard to the impact of patients use of LDH on the encounter length.
According to some HCP, LDH seemed to lengthen encounters, while others reported that they did not note any changes. A previous study which evaluated the efficacy of web-based educational approaches on patient-provider communication did not find any change of the encounter length. 15 Some HCP expressed some doubts about LDH's usefulness and their reluctance to integrate it in their practice because of their concern of being overwhelmed by questions and needing to devote more time to appointments. Because of this, some of them were uncertain of their intention to continue to encourage LDH use. In their view, encouraging patients to use LDH required an additional effort on their part which they were not necessarily ready to commit to. Similar to our findings, many studies have shown that HCP express many concerns about use of patient portals. 58,59 However, Wittink et al (2018) showed that using a health information technology (HIT) designed to help complex patients with multimorbidity disclose their stressors to their HCP was associated with increased psychosocial concerns and stressors disclosure without increasing visit length. 60 These disclosures also arrived earlier in the encounter. Yet, successful adoption of patient portals and visit preparation web-based tools depends on their integration into regular care programs and the recognition that active patient participation is an integral part of effective patient-HCP relationships. 45,58 Patients who participated in the study were highly activated as indicated by high PAM scores (93% of patients had 3 or 4 level PAM score), which is consistent with Hibbard and Green's study 61 as well as Henselmans' study. 62 However, in the absence of a patient baseline PAM score, it is not possible to attribute the level of activation observed to the LDH implementation. Our research design, a real-world implementation study, did not allow us to evaluate a change in patient PAM scores before and after the use of LDH as questionnaires were completed by patients after they had experienced the use of LDH. It is unlikely that the use of LDH could have had a huge impact on the activation score considering the elements that are measured in this instrument. However, it makes sense that already activated patients could have been more attracted to the type of self-management support LDH offered than less activated patients.
Patients view LDH as an opportunity to take responsibility for their own health-care and play a greater role in it. Its use contributes to building a better understanding of their health problems. Both data from questionnaires and focus groups showed that patients intended to continue using LDH and even, for some of them, encourage their relatives and acquaintances to use it. This is interest-  68 there is a need to better teach HCP on how these tools may support patient engagement, and self-care and to downplay the problems they anticipate due to use of such tools by their patients.
Many ideas to facilitate the use of the LDH were suggested by study participants. Among others, they suggested having a video demonstrating LDH functionalities projected in clinic waiting rooms, implementing automated reminders to visit LDH before the scheduled appointment, including a hyperlink to LDH website in the clinic's own websites, allowing patients to electronically share their visit preparation summary with the HCP before the scheduled visit. However, unlike some patient health records (PHR) used in Anglo-Saxon countries, 33 LDH currently does not offer any capacity to interface with the clinic's electronic medical record (EMR). Such a secure bidirectional interface would allow patients to access health information from their medical records and insert it directly into LDH and allow HCP to access in a timely fashion the contents of the patient's visit preparation summary. It was believed necessary that the various stakeholders will need to intervene in a coordinated manner to promote and facilitate the use of LDH and that a simple invitation to visit LDH will never be sufficient to promote its use in a meaningful way.
Of interest, HCP did not mention the role they could play to encourage LDH use by their patients, possibly because they did not recognize its added value for patients. Getting patients to prepare for the medical encounter is probably one of the most powerful ways to address the inherent asymmetry of the doctor-patient encounter. [69][70][71][72] Thus, not perceiving the importance of the role they play in encouraging patient participation represents a significant barrier to the adoption of LDH. Patients value their trusted HCP's opinion and would probably consider this suggestion to prepare as an explicit invitation to partner with them. For example, HCP checking explicitly if the patient has prepared for the visit using LDH during the encounter could go much further than a letter or phone call.

| Study limitations
This study has several limitations. First, because we did not reach the planned sample size, of 150 patients per province, we could not compare the data according to the two provinces (Quebec vs New Brunswick) to identify differences, especially in regard to the perceived value of LDH, and its perceived impact on both patient participation in the visit as well as its impact on its organization.
Second, since we did not audio record the encounters, we do not know how often or to what extent the patients and HCP referred to the patient summary sheet and how it influenced, for example, visit priority agenda setting. Lastly, this study is subjected to a patient selection bias as discussed in the previous section; it is possible that patients who accepted to participate were already highly activated patients. 73 In this self-selected patient group, however, the usefulness of LDH was confirmed.

| CON CLUS ION
LDH has fulfilled its promise for patients that have used it but the picture is less clear for the HCP. The findings of our study contribute in filling a knowledge gap on how to implement the use of electronic tools which assist patients in the preparation of their medical visits.
Our study suggests a strong need for training and support to assist less activated patients with portal registration and use, particularly those with limited health literacy. Efforts to encourage electronic tool use among these patients should directly address health literacy and support access for caregivers. Choosing to use LDH moves patients from a 'passive' stance in the medical encounter to an 'active' one.
This change of behaviour is important and as any change of behaviour it will not happen overnight. It will take several requests from staff and HCP before we observe the transition from a passive to an active attitude. This attitude change is likely to help the HCP to organize visit priorities and to increase the success of the proposed treatments.
The study needs to be replicated in other medical environments using larger and more diverse samples. In future studies, it would be relevant to analyse communication patterns during encounters prepared with LDH. In this context, it would be interesting, for example, to measure results on visit agenda setting, patient initiated discussions, amount of dialogue 15 and patient recall and adherence to the recommendations and interventions discussed during the encounter.
In terms of next steps in our own context, we are exploring the possibility to include LDH into the Québec Health Booklet, because of the complementary functionalities it provides to those already present in this provincial web platform such as access to laboratory test results, imaging reports and making medical appointments. The Québec Health Booklet would thus become even more patient-centred than it is at the moment.

ACK N OWLED G EM ENTS
The authors thank the health-care providers, managers and patients of the six primary care clinics for their participation in the project.

CO N FLI C T O F I NTE R E S T
Both BD and CR, as project coodinators, received honoraria for their work in this project. MTL and CR are married. The authors have no other potential conflict of interest or financial relationship relevant to this article to disclose.

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.