A Web-Based Social Network Tool (GENIE) for Supporting Self-management Among High Users of the Health Care System: Feasibility and Usability Study

Introduction

Background

It has been reported that globally, 1 in 3 adults have multiple chronic conditions (MCCs) [1]. There was an increase from 45.7% in 1988 to 59.6% in 2018 with regard to adults in the United States with 2 or more MCCs, and the weighted prevalence of 2 or more MCCs was higher in those aged ≥65 years [2]. Fostering self-management support for health conditions is particularly important, given the rising numbers and projected rise in complex multimorbidity. Self-management support builds problem-solving skills to enhance self-efficacy to carry out behaviors toward a desired goal and can support positive health outcomes, reduce the burden of long-term conditions for the patient and the health system, and decrease health system costs [3,4]. A qualitative systematic review identified challenges that patients experienced with self-management, including dealing with physical and emotional symptoms; living with pain, depression, and fatigue; and having a lack of understanding of self-management strategies related to conflicting information from providers [5]. Kang et al [6] found that quality of life scores were higher among patients with good versus low self-management strategy scores regardless of the number of comorbidities.

Improving access to health and social services (HSSs) to address self-management can be supported through information, referrals, facilitation, and system navigation by primary care providers [7-9]. Results from a longitudinal study of 300 randomly selected patients with diabetes or chronic heart disease found that connecting people to social support resources, including a variety of people and groups, supported self-management and physical and mental health [10]. Patients with multiple and complex health and social conditions are likely to derive the maximum benefit from linkages to HSSs [11].

In recent years, researchers have established that social networks can influence positive health behaviors and practices, and this is also true in populations that are managing long-term conditions [12-15]. Social connectedness has been shown to be particularly beneficial for vulnerable groups, such as those living in poverty and with chronic illnesses [16]. Reeves et al [10] established associations between connections with and the use of local networks, resulting in improved physical and mental well-being and better coping with their conditions. Personal and social networks and relationships in community settings can act as a conduit for accessing resources and provide support for managing long-term conditions, which can complement what is provided by formal service provision.

The implementation of a self-management support intervention in 31 primary care settings in England had poor uptake because of a perceived lack of relevance and fit to accessible sources of support and because primary care health care professionals did not prioritize self-management support [17]. Primary care providers in Canada have been tasked with fostering self-management through support and coaching, referral management, and linking to relevant community-based resources and services [4,18,19]. However, like their UK peers, they have struggled with limited time for coaching; a lack of knowledge of what community-based HSSs are available and how they can address health and social needs; and a lack of time to keep up with changing community services, including concerns about their quality [20,21]. Despite these challenges, it is argued that there is a need for primary care providers to implement effective self-management support interventions that incorporate connections to community resources for those living with long-term conditions [22] and, particularly, for those who are known to be isolated, requiring more encouragement to make connections [23-25].

The aim of the GENIE (Generating Engagement in Network Involvement) tool was to encourage the expansion of a patient’s social networks to reduce the negative health impacts of long-term conditions and to reduce the concomitant social effects, such as social isolation and loneliness [23]. GENIE is a web-based tool that aims to support self-management by leveraging adults’ engagement with their personal social network to facilitate the uptake of relevant community-based activities and HSSs. Studies have shown that when GENIE was delivered by trained facilitators to adults with chronic health and social conditions in the community settings, there was an increase in the diversity of participants’ networks and greater engagement with community activities [26,27]. Given these positive results and the challenges faced by primary care providers in implementing self-management strategies [17], research is needed to understand the feasibility, usability, and perceived impacts of implementing GENIE within the primary context, with the use of trained facilitators as an extension of the primary care team. This knowledge will be useful to inform future implementation of the GENIE tool in primary care and can serve as a basis for the development of outcome measures to be used in future controlled studies.

Research Questions

This study examined the feasibility, usability, and perceived patient outcomes of the implementation of GENIE with adults enrolled in Ontario’s Health Links Program. This program alerts health providers to individuals with high rates of health service utilization to target care and thereby reduce health care costs [28]. The research questions were as follows:

1. What is the usability and feasibility of implementing GENIE, facilitated by lay volunteers and primary care providers, with 55- to 69-year-old adults enrolled in the Health Links Program?
2. What are patients’, providers’, and volunteers’ perceptions of the impact of the use of GENIE?

Methods

Social Network Tool (GENIE)

Overview

GENIE is a web-based tool designed by a team of researchers from the United Kingdom [29]. GENIE has been previously implemented by trained lay or health care workers in various contexts in the United Kingdom and Europe to link patients to community-based HSSs to support them in reaching their life and health goals [22,24,26]. The GENIE tool has 3 core functions: (1) mapping a patient’s personal social network to better understand a patient’s support network and identify possible network members who can assist them; (2) selecting topics of interest that relate to patients’ interests under the categories of activities, health, learning, support, independent living, volunteering, and pets; and (3) geolocating local community programs, services, and resources related to the selected categories (Figure 1).

‎View this figure

The GENIE tool consists of 4 steps. First, patients enter an email address (or get help to obtain an email address) to log in and enable them to save their results, select from a list of common health conditions (eg, heart problems, stroke, diabetes, arthritis, and cancer), and enter their age and postal code.

Second, the patients generate a personal network map that lists individuals, groups, or organizations (eg, son or daughter-in-law, friend, and social club) that the patients consider important to them in relation to being healthy and living at home. Network members who are deemed to be most important to them are placed closest to the center of the circle, where the patients are placed, with others moving out into the outer circles. Each member is categorized by type (eg, family, friend, neighbor, group, or organization), which determines the typology of that patient’s network. Network typologies can consist of mostly friends and family members; mostly professionals; or a diverse mix of professionals, organizations, friends, and family (eg, My Network–Diverse; Figure 2). Diverse networks are the most robust social networks in the GENIE typology containing family, friends, and weak tie relationships, whereas very isolated and friend and family supported networks have fewer members and less diversity of relationships. Research has shown that people with long-term health conditions and diverse networks are associated with enhanced self-management skills [30]. Patients also indicate the frequency at which they meet with each network member. This information can help identify network members who may be more available to support a patient’s HSS use.

‎View this figure

Third, patients answer 12 questions about their interests organized under the following categories: (1) activities, (2) health, (3) learning, (4) support, (5) independent living, (6) volunteering, and (7) pets. Some questions have subquestions, for example, if a patient is interested in activities, they are prompted with subquestions to refine the topic (ie, reading and writing, drama and music, arts and crafts, or social clubs).

Fourth, once the patients complete the questionnaire, they move to a web page listing links to relevant community-based HSSs organized under the relevant categories. For example, if a patient indicates that they are interested in physical activity, they can find HSSs listed under the health tab related to this subactivity (Figure 3). HSSs are geomapped for selection based on the patient’s preferred distance from their postal code (1 km, 2 km, 5 km, 10 km, or 50 km). Relevant HSSs, including a brief description of their programs or services, are populated from the region’s community information database.

Patients review their results and mark their favorites (Figure 3), which can be saved, downloaded, and printed in a short report for easy access (Figure 4) [13]. Facilitators were to encourage patients to consider their social networks to help in overcoming barriers to access the desired HSSs.

‎View this figure

‎View this figure

Adaptation of GENIE for Use in Canada

GENIE was adapted for use in Hamilton, Canada. The UK open-source tool required minor word modifications in the questionnaire to address the linguistic differences between Canadian and British audiences (eg, changing the word befriending to friendly visiting). A second significant adaptation was linking GENIE to a back-end database of community-based HSSs. In the United Kingdom, community information services with databases, such as those in Canada, do not exist. The Hamilton database was maintained by the Region’s Community Information Service—Information Hamilton. Most programs and services included in the database are run by not-for-profit organizations or government organizations. For-profit services were included if there were no not-for-profit agencies available that offered the same or similar service (eg, home oxygen providers). All database entries were tagged with keywords using the AIRS/211 LA Taxonomy of Human Services which is the industry standard for the Alliance of Information and Referral Systems (AIRS) [31]. To link the database to GENIE, keywords related to all topics of interest listed in the GENIE questionnaire were identified. For example, if a patient had an interest in walking or outdoor activities, programs and services were tagged with the AIRS categories of walking programs, walking tours, and nature centers or walks. This enabled GENIE to filter information from the database to match the selected areas of interest. Information Hamilton staff pulled together a complete list of all database items with selected search terms for our review for relevance to this adult population and to mark services for exclusion, such as youth programs or programs outside of the city. The database was updated daily by Information Hamilton’s staff, who regularly reach out and work with local service organizations to keep the database up to date.

Implementation of GENIE

GENIE facilitators were volunteers who attended a half-day training session to learn about the application of the GENIE tool during home visits. They visited with patients in 2 instances (the GENIE visit in Figure 1). At baseline, they would sign up patients and log on to GENIE, identify their social networks, and help patients to explore links to community support. The volunteers were trained to facilitate the use of the GENIE tool to engage in a discussion with the patients about their social networks, to discuss the access and use of community-based services, and to identify any additional services that they would like to access through the preparation of the GENIE report. The reports were sent to the patients’ electronic medical records for review by the primary care huddle team (a component of the Health TAPESTRY [Teams Advancing Patient Experience: Strengthening Quality] program) for planning and care coordination. The team told patients about any critical information regarding their plan of care and consulted with the relevant family physician in the clinic when needed. The report was printed and left with the patient after the home visit. After 3 months, the volunteer would return to the patient’s home, log on to GENIE, repeat the social network mapping, and revisit the GENIE report to determine if the patient had explored any HSSs. The volunteer role was intentionally limited to the role of a nonmedical volunteer facilitator. If any health care–based issues or concerns arose, patients would be encouraged to connect with their primary care provider [32].

Setting and Sample

The study was conducted between August 2017 and March 2018 at 2 sites of a family health team (composed of 2 interprofessional primary care team clinics) that serves 30,000 patients in Hamilton, Canada. It targeted patients who were enrolled in Ontario’s Health Links Program [28]. Modeled after accountable care organizations in the United States, England, Australia, and New Zealand, the Health Links Program was launched in 2012 in Ontario, Canada, to improve care coordination for patients with complex needs who are the highest users of the health care system. The program connects them to primary care providers and engages them in their health care via active care planning [28]. These patients are considered high-cost, high-need users of the health care system—the top 5% of the population who use two-thirds of the health care spending [33]. GENIE was implemented with a small sample of Health Links patients within the context of the Health TAPESTRY program. Health TAPESTRY provided a unique structure for GENIE’s implementation within a primary care setting, as the program includes the use of trained volunteers as an extension of the primary care team, who visit patients in their homes. Health TAPESTRY is a multicomponent primary care intervention that centers on supporting older adults’ life and health goals [34,35]. The Health TAPESTRY program components include (1) trained volunteers visiting in pairs to collect health information using web-based surveys related to health risks and life goals [36], (2) care coordination by an interprofessional primary care team, (3) the use of technology to share health information between volunteers and primary care providers, and (4) support for system navigation [37]. Trained community volunteers implemented the GENIE tool, in addition to the Health TAPESTRY surveys. GENIE results and survey data were compiled into a web-based report that was transmitted to the interprofessional primary care team to support the formulation of a patient care plan.

Ethics approval was received for this study from the Hamilton Integrated Research Ethics Board Project number 13-366. All participants provided written informed consent before data collection.

Data Collection

GENIE was field tested for feasibility and usability with a small group of volunteers and patients who were also receiving the Health TAPESTRY program. Data collection methods included (1) use statistics, (2) field observation in the community with patients and volunteers, (3) field observation notes taken during primary care team meetings, (4) focus groups with clinicians and volunteers, and (5) patient interviews. A summary of the data collected is provided in Table 1, and the focus group guides are provided in Multimedia Appendix 1.

Field observations focused on the usability captured via researcher observations during home visits with volunteers and patients at baseline. Field usability testing [38,39] was conducted via observation to assess the cognitive processes of users performing task completion of the GENIE tool, as trained volunteers facilitated the use of the web tool in a home visit. LC and RV observed volunteer and participant dyads to identify potential usability problems, with particular attention to goal-action sequences and interactions between volunteers and participants, dimensions of competencies (skills and knowledge required to complete the tool), barriers to the productive use of the tool, time to complete the tool, and ease of use for participants and volunteers [39]. Field observation notes captured procedures for each step or task to complete the tool (Multimedia Appendix 2). Tasks included setting up, completing the introductory and demographic page, completing network mapping and the questionnaire, discussing and tailoring the results, and printing the final results and the my network page that included the social network map. It also included comments and questions raised by the participants and volunteers as they worked through the task. Field notes were recorded by LC at interprofessional team meetings, in which patient GENIE reports were reviewed and discussed.

RV and LC conducted focus groups with volunteers and clinicians from 2 teams immediately following their huddle team meetings, which helped to gain participation in the research. Semistructured interviews were conducted by LC with study participants to explore the feasibility of GENIE using an interview guide that applied concepts from the Normalization Process Theory (NPT) [40] and perceived impacts. The NPT has been used for the feasibility study of a web-based program in primary care [41]. All focus groups and interviews with providers took place 6 months after the first use of GENIE by patients. The interview and focus group guide were tailored for each participant group (patients, volunteers, and the primary care huddle team) to explore the feasibility, usability, and perceived impacts of GENIE (refer to Multimedia Appendix 1 for the full guides).

The participants were recruited via convenience sampling. Lists of Health Links patients were distributed to their physicians from the 2 clinics that participated in the Health TAPESTRY program. A total of 25 Health Links Program patients were invited to participate in the study via a letter from their primary care physician. Physicians selected these patients based on age (55-69 years), enrollment in Health Links, and a clinical assessment indicating that they could benefit from improved care coordination offered by the GENIE and Health TAPESTRY. The target number of participants was 10, with diverse demographic characteristics (gender and age), which was deemed sufficient for usability testing [42,43]. A research coordinator received signed consent forms from 11 potential participants who were contacted by telephone to schedule the first of 2 home visits. One participant could not be contacted, another participant died before the first planned visit, and a third patient participated in the first home visit but withdrew from the study because of mental health distress. A total of 8 participants completed the study. All clinicians involved in the Health TAPESTRY program were invited to participate in the focus groups, which included questions about the GENIE tool. All huddle clinicians (a selected small interdisciplinary core team who met regularly to plan patient care) participated in a focus group held at each of the 2 sites (n=16). The remaining clinicians who worked in the clinic and were members of the huddle team were also invited to participate in one-on-one interviews. Of all the clinicians, 17% (7/41) agreed to participate. All clinic managers and volunteer coordinators agreed to participate in the interviews (n=3). Three volunteers who were trained to facilitate the GENIE tool participated in the focus group.

Participants

We recruited 5 male and 3 female patients (4 patients from each clinic) with an average age of 63 years (SD 4.6; range 57-69 years). Half of the patients (4/8, 50%) had no computer access and had never used computers, whereas the other half (4/8, 50%) had used computers regularly. A total of 6 patients were married, 1 was divorced, and 1 had unknown marital status. Participants had a mean of 3.9 (SD 1.8) chronic diseases, ranging from 2 to 8 chronic diseases, including depression (n=6), anxiety (n=5), diabetes (n=5), cancer (n=4), arthritis (n=4), chronic obstructive pulmonary disease (n=2), heart disease (n=2), and pancreatitis (n=1).

Of the 3 volunteers, 1 was a male university student and 2 were retired females. The primary care team participants were members of the interprofessional teams at 2 clinics. One team had 7 members and included a dietitian, an occupational therapist, a physiotherapist, a pharmacist, a system navigator, 2 physicians, and a registered practical nurse. The second team had 9 members and consisted of an occupational therapist, a physiotherapist, a pharmacist, a system navigator, a physician, a psychologist, a registered practical nurse, and 2 nurse practitioners.

Analysis

Qualitative data (interview, focus group, huddle notes, and field observation notes) were uploaded and organized using NVivo (QSR International) software, version 10 [44]. Two authors (LC and RV) reviewed all the data sources. Field observation notes were coded in NVivo according to their organizing criteria using a qualitative descriptive approach (Multimedia Appendix 2) by LC, and they were reviewed by RV. LC coded the interview and focus groups by inductively organizing the coding using the interview guide that was guided by NPT [40]. RV reviewed all the coding, and the research team reviewed the final coding structure with themes to increase rigor in the results.

Participants’ network maps were categorized using criteria developed by the GENIE founders and coauthors AR and IV, to identify the network typologies (based on the size of the network and type of network members). All authors reviewed and discussed the preliminary and final findings and interpretations.

Results

Areas of Interest and Services Chosen for Potential Follow-up in GENIE

All patients completed a network map and a questionnaire that asked about a person’s interests in various types of activities. Table 2 shows the categories and areas of interest chosen by participants. Patients were most interested in getting more physically fit (8/8, 100%), managing their weight (6/8, 75%), and learning about their health condition (6/8, 75%). All other topics garnered some interest, except knowing more about supports for a pet. Participants chose a range of 4-9 topics of interest, with an average of 6.

Patients favorited programs or services from the HSS database for potential follow-up that were related to the following: social clubs (5/8, 62%), home support (2/8, 25%), swimming (2/8, 25%), drama and music (1/8, 13%), and fitness and exercise (1/8, 13%). A total of 4 patients followed up with their favorited community services, including community or seniors’ social clubs (n=3) and an aquafit class (n=1). One patient was disinterested in the community services. Patients could also find relevant HSSs to address the health conditions of interest. Patients selected diabetes (1/8, 13%), cancer (1/8, 13%), and other health conditions (1/8, 13%) but did not follow up on any of the relevant HSSs.

Feasibility and Usability

Overview

Most patients and volunteers shared positive perceptions of the tool overall, such as the perception that the tool was easy to follow and understand. In addition, most patients noted that they would recommend it to others. However, several feasibility and usability issues were identified, as listed in Textbox 1 and described later. Quotations from participants are indicated by the participant type and ID number, such as patient 1, volunteer 3, and clinician 5. As we did not always capture the clinician’s names in the huddle meeting notes, participant IDs are missing for some quotes.