On the same page? A qualitative study of shared mental models in an interprofessional, inter-organizational team implementing goal-oriented care

ABSTRACT

Goal-oriented care is an approach to care delivery that uses patient-identified goals to drive care planning. Implementing goal-oriented care requires team members to cognitively shift the focus from “what is the matter” to “what matters to patients,” and align their mental models of what it means to care for patients. Yet, no empirical studies of goal-oriented care apply evidence from the cognitive sciences, such as Shared Mental Model (SMM) theory. We conducted a qualitative case study of an interprofessional team that adopted goal-oriented care in Vermont, US (n = 18). Guided by SMM theory, we distinguished between task-related and team-related mental models. We used framework analysis and qualitative content analysis to determine mental model content and similarity. The most shared content areas were operationalizing goal-oriented care, engaging in formal and informal communication, taking a “whole-person” approach, taking a team approach, and building trusting relationships with patients and with other team members. Trust was the only construct that spanned both task and team mental model categories, highlighting the importance of both intra-team trust and provider-patient trust to the implementation of goal-oriented care. Team members developed SMMs through training, regular meetings, and interactions during care delivery. This study provides insight into the cognitive mechanisms that underlie team-based goal-oriented care delivery, which can be used to inform implementation, training content, and future research.

KEYWORDS:

• Shared mental models
• teamwork
• integrated care
• patient goals
• primary healthcare
• qualitative
• goal-oriented care

Introduction

Integrated community-based primary health care (ICBPHC) teams are being implemented to improve care for people with complex health and social needs related to poverty, social isolation, and mental health (Schaink et al., 2012). ICBPHC models are uniquely designed to address diverse and complex patient needs holistically. With an orientation toward improving the health of the population, ICBPHC teams build partnerships across a broad spectrum of health and social services agencies (Levesque et al., 2011). Thus, successful delivery of ICBPHC requires implementation of person-centered teams with members that span professional and organizational boundaries. These individuals share one or more common goals, interact socially, and exhibit interdependencies in task workflows, goals, and/or outcomes, as per the definition of a “team” (Kozlowski & Ilgen, 2006).

Primary healthcare organizations are redesigning care delivery to achieve team-based, person-centered care but have faced challenges in the process (Cronholm et al., 2013; Goldman et al., 2018; Pandhi et al., 2018; Rodriguez et al., 2013). Goal-oriented care (GOC) is an approach that can support ICBPHC teams to deliver person-centered care (Steele Gray et al., 2020). Under a GOC approach, health and social care providers work together and with patients to develop care plans based on patient-identified goals and priorities. GOC shifts the focus of care from “what is the matter?” to “what matters to you?” (Berntsen et al., 2019). This is an ideal approach for people with complex needs because of the focus on identifying outcomes important to them that often span different health and social issues (Steele Gray et al., 2020).

Part of the challenge in delivering ICBPHC using a team-based, goal-oriented care approach is that it requires leaders and providers to change how they practice, which fundamentally involves shifting how they think about their work. In other words, shifts are required in individuals’ mental models – their psychological representations – of how to execute their work (S. McComb & Simpson, 2013; Nutting et al., 2009). Developing similar mental models regarding work tasks and interactions ensures that team members are “on the same page” and can effectively execute team functions such as communicating, coordinating, decision-making, and managing uncertainty (Holtrop, Potworowski, Fitzpatrick, Kowalk & Green, 2015; Wagner, Austin, Toon, Barber & Green, 2019). While existing studies in team-based primary care discuss the concept of mental models (Cronholm et al., 2013; Nutting et al., 2009), they seldom draw from the rich literature on shared mental models.

Shared mental model (SMM) theory is a well-established theory in industrial psychology that links team cognition to superior team performance (Mohammed, Ferzandi & Hamilton, 2010). Most SMM studies examine the relationship between mental model similarity and team performance using quantitative methods (Mohammed et al., 2010). These studies confirm that SMMs predict team performance (Mohammed et al., 2010). However, the focus on performance and quantitative methods has left a gap in the literature regarding the content and influence of mental models in specific contexts and for specific sets of tasks (Mohammed et al., 2010).

The aim of this paper is to advance our understanding of how GOC is implemented by ICBPHC teams using SMM theory. Using in-depth qualitative methods, we identify the mental model contents that team members exhibited in implementing GOC, determine the extent to which these mental models were shared, and explore how the team developed SMMs.

Shared mental model theory

Mental models are individuals’ cognitive representations of external reality (Cannon-Bowers, Salas & Converse, 1993). Individuals have many mental models which help them interpret situations and guide decision-making (Cannon-Bowers et al., 1993). According to SMM theory, when team members have similar mental models, their collective performance is maximized (Klimoski & Mohammed, 1994). Mental models can become similar over time as team members interact and experiment with different ways of working together (Mohammed et al., 2010). There are two types of SMMs: task-related and team-related (Mohammed & Dumville, 2001). Task-related mental models refer to knowledge and beliefs about goals, activities involved in performing a task and their sequence, and contingency plans (Mohammed & Dumville, 2001). Team-related mental models refer to knowledge and beliefs about the roles and characteristics of team members, and the team’s communication and interaction patterns (Klimoski & Mohammed, 1994; Mohammed & Dumville, 2001). Both types of mental models are comprised of knowledge, what one knows or believes to be true, and beliefs, what one prefers or expects.

Decades of empirical evidence suggest that SMMs are an important predictor of team functioning and performance, particularly for teams operating in dynamic and complex contexts in which their circumstances are changing, and direct dialogue is not always possible (Dechurch, Mesmer-magnus, Walton & Patton, 2010; Mohammed et al., 2010). SMMs allow team members to think “on the fly” and act in ways that are consistent and coordinated in performing interdependent tasks (Mathieu et al., 2000). SMMs facilitate adaptation under changing conditions by enabling individuals to predict the actions of their team members (Mathieu et al., 2000). Conversely, divergent mental models create confusion among team members and can hamper the implementation of team-based interventions (Evans, Grudniewicz & Tsasis, 2018; Hysong, Best, Pugh & Moore, 2001). To enhance teamwork research and practice, research on SMM contents in specific contexts and task domains is needed.

Methods

Study design

A case study of how an ICBPHC team implemented GOC in Vermont, United States was conducted. While performing interviews, the research team noticed striking similarities in participants’ understanding and perspectives of GOC, despite differences in professional role and organizational context, and no prior awareness of the interview questions. To examine this phenomenon more closely we applied SMM theory. The post-hoc application of SMM theory is appropriate because a primary goal of inductive qualitative research is to learn through and from participants about what is important in the context under study (Pratt, Kaplan & Whittington, 2020). The Vermont Agency of Human Services Institutional Review Board, and the Toronto Academic Health Sciences Network research ethics board approved this study prior to data collection. All participants provided informed consent for their participation prior to data collection.

Setting

The ICBPHC team was 18 months into implementing GOC at the time of data collection. Members of the interprofessional and inter-organizational team included organizational leaders, primary care providers, case managers, home care coordinators, mental health counselors, and quality improvement specialists. The team included representatives from agencies across primary care, acute care, mental health, and social services (e.g., aging, child and family services, housing). The team was supported by a state-funded practice facilitator who supported the implementation including leading a Learning Collaborative, a regular meeting where team members discussed program goals, implementation successes and challenges, and complex cases.

All team members participated in an intensive 2-day training program on GOC which involved learning about effective goal-setting processes, defining roles, practicing GOC tools and techniques (Table 1), developing and reviewing mock shared care plans, and role playing. Ongoing training opportunities were available as needed. Team members attended bi-weekly Learning Collaborative meetings.

Table 1. Goal-oriented care tools and techniques.

Tool/Technique	Objective
Chart review	To understand the patient’s medical history and healthcare needs
Motivational interviewing	To lead strength-based conversations that help to gain patients’ trust
Eco-mapping	To identify all formal and informal team members who are involved in the patient’s care including family members
Camden domain cards(Camden Coalition, 2016)	To elicit patient-identified goals to understand what is important to the patient
Backwards planning boards (Camden Coalition, 2016)	To prioritize the patient-identified goals and create a list of actionable goals to work toward
Shared care plan	To co-create a plan of care that works for team members and patients

Participants

Working with the case site lead, a maximum-variation sampling approach (Palinkas et al., 2015) was used to identify 18 team members – both providers and leaders – from 7 organizations. The research team worked with a local site champion to identify and recruit clinical and managerial key informants. All participants had experience delivering GOC at the frontline or were involved with the development and implementation of the GOC approach.

Data collection

Data were collected over a three-day site visit during which the first and senior authors conducted 13 one-to-one, 1-hour semi-structured interviews, and one team observation. The authors are trained qualitative researchers in health services research and took a constructive approach to emphasize the lived experiences of the team. The core analytic team (JI, CSG, JM) are not clinicians, and therefore, did not approach data collection or analysis through a particular clinical lens. The observation was of a 75-minute Learning Collaborative meeting with 14 team members from five organizations. The purpose of observation was to capture the structure of the meeting, the contents discussed, and team interactions. The observation allowed researchers to see teamwork unfold and how team members discussed and operationalized the GOC process. Four additional interviews were conducted by phone. In total, 17 interviews were conducted with 18 team members (one interview was conducted as a dyad). Interview guides included questions such as, “What does GOC mean to you?” and “How did you learn about GOC?” (Additional File 1). The interview questions paired with probing generated rich data on participants’ knowledge and beliefs about team tasks and interactions related to GOC. Semi-structured interviews and probing are a common method for eliciting mental model content (Gisick et al., 2018; Mohammed et al., 2010). The first five interviews were conducted jointly (CSG, JI) to ensure a similar approach to data collection. After conducting each of the initial five interviews, the interviewers debriefed and made reflections. The researchers found that the interview guide captured data well and adapted the interview guide as interviews unfolded based on responses. Qualitative memos were written to capture reflections after each interview. All participants were provided an opportunity to review their interview transcript. Multiple conversations about the findings from the Vermont case study occurred with the site lead and a summary of findings was provided to the team after data collection. All interviews were recorded and transcribed verbatim by a professional transcriptionist. Data were collected from June – July 2017.

Data analysis

The data were analyzed iteratively to explicate the content and sharedness of mental models using framework analysis and qualitative content analysis (Gale, Heath, Cameron, Rashid & Redwood, 2013; Hsieh & Shannon, 2005). Initially, we attempted to distinguish between knowledge and beliefs in our analysis by examining the word choice of participants in describing the implementation of GOC. We coded chunks of text as task-related or team-related “knowledge” and task-related or team-related “beliefs.” However, knowledge and beliefs were too deeply intertwined to be systematically disentangled. Distinguishing between knowledge and beliefs may require more in-depth cognitive interviewing. In our final analysis, we therefore only distinguished between task-related and team-related mental model content (Table 2) which were more clearly articulated by participants and, as such, more conducive to classification.

Table 2. Definitions of shared mental models of goal-oriented care adapted from (Cannon-Bowers Et Al., 1993; Mohammed & Dumville, 2001).

Concept	Definition
Shared mental models of goal-oriented care	When multiple individuals develop overlapping knowledge and beliefs regarding the tasks and team involved in delivering goal-oriented care, we refer to this as a “shared mental model of goal-oriented care”
Task-related mental model of goal-oriented care	An individual’s knowledge and beliefs regarding the purpose of goal-oriented care as well as how it is delivered (e.g., procedures, tools, action sequences, environmental constraints, contingency plans, and equipment/technology related to the delivery of goal-oriented care)
Team-related mental model of goal-oriented care	An individual’s knowledge and beliefs regarding teamwork in general as well as team member characteristics (e.g., expertise, skills, preferences), respective roles, and team interactions required to deliver goal-oriented care (e.g., communication patterns, coordination strategies)

During the first round of coding, three authors (JI, JE, and CSG) independently coded the same three transcripts to ensure consistency in coding using definitions of SMMs of GOC. The coders met multiple times to compare coding of the data. Discordance in coding was discussed until consensus was reached. Once consistency was established, two authors (JI and CSG) proceeded with coding the remaining transcripts. After hand-coding all transcripts, framework analysis was used to identify mental model contents of GOC. A matrix capturing data on each type of mental model was created resulting in summaries by category for each participant. In framework analysis this process is known as “charting” (Gale et al., 2013). Once charting was complete, the summary cells from the matrices were amalgamated into a framework matrix, containing the summary cells of each category by participant. Using the framework matrix, we synthesized across participants to identify an initial set of SMM contents based on commonalities that were found.

Then, we applied qualitative content analysis (Hsieh & Shannon, 2005) to determine the extent of sharedness. Content analysis is an established method for systematically tracking and comparing mental model sharedness among participants (Dechurch, Mesmer-magnus, Walton & Patton, 2010). The SMM contents were used to systematically recode the findings from the framework analysis. Through this step, the frequencies of SMM content areas were determined. Mental model contents that were mentioned by at least two participants were considered shared. For example, stating “I don’t think you can really understand their goals if you don’t try to get to know them as a whole person first” (VT-01) and “you’re finding out what’s working for them in their life, what’s not working for them in their life, and, you know, what are their hopes, their dreams, their concerns, what are their goals” (VT-12) were both coded under adopting a patient-centered care approach and would count as being shared by two participants. Although the focus of this study was on SMMs, we counted the frequency of divergences in participants’ mental models to provide additional insight into the overall extent of sharedness. In summary, the framework analysis generated the mental model content, and the content analysis determined the extent to which that content was shared across participants. Field notes taken during the observation and memos written after observation and interviews were used to reflect on the findings as analysis evolved.

Findings

Our analysis demonstrated that the ICBPHC team held mental models of GOC that were broadly shared. Specifically, four task-related and five team-related content areas were shared by over 50% of participants, as shown in Tables 3 and 4. We also identified mental model contents that were not shared by participants (see Additional Files 2 and 3). We found 5 divergent contents for task-related mental models and 2 divergent contents for team-related mental models. These mental model divergences did not exhibit any meaningful pattern. Below, the most common task-related SMMs followed by the most common team-related SMMs are discussed. Although we report discrete SMM contents, our results suggest that the contents are deeply interrelated within and between task-related and team-related mental models.

Table 3. Shared task-related mental model content and degree of sharedness.

Task-related mental model content	Frequency (percent shared)
Delivering goal-oriented care involves the following tasks:
Operationalizing goal-oriented care (e.g., identifying a lead care coordinator, utilizing goal-oriented care tools, developing shared care plans)	18/18 (100%)
Adopting a person-centered or whole-person approach to care (e.g., getting to know patients as individuals, prioritizing patient-identified goals over provider goals for their patients)	16/18 (89%)
Developing trusting relationships with patients (e.g., through motivational interviewing)	14/18 (78%)
Undergoing intensive training and education on goal-oriented care as a philosophy and practice on an on-going basis	10/18 (56%)
Establishing a supportive context for goal-oriented care (e.g., dedicating time and the necessary financial and human resources)	7/18 (39%)
Applying qi methods (e.g., rapid cycle testing, case finding, performance measurement)	4/18 (22%)

Table 4. Shared team-related mental model content and degree of sharedness.

Team-related mental model content	Frequency (percent shared)
Delivering goal-oriented care involves the following team processes and interactions:
Engaging in formal and informal communication (e.g., digital communication platforms, regular meetings, e-mail, phone calls, fax)	17/18 (94%)
Taking a team approach to care planning and delivery (e.g., working collaboratively, supporting each other, sharing the responsibility of care)	16/18 (89%)
Building knowledge of and trusting relationships with each other (e.g., learning about respective areas of expertise, experience, and backgrounds)	16/18 (89%)
Clarifying respective roles in care delivery	15/18 (83%)
Working with supportive leaders within and across partner organizations who enable teamwork (e.g., facilitate a shared vision, dedicate time for inter-organizational meetings, leverage champions)	10/18 (56%)
Engaging primary care physicians (e.g., including physicians in a consultative role, securing physician buy-in, addressing differences in care philosophy between physicians and other team members)	5/18 (28%)

Task-Related mental models of goal-oriented care

As Table 3 outlines, the most shared task-related mental model contents were operationalizing GOC processes (100%), adopting a person-centered approach (89%), and developing trusting relationships with patients (78%). All participants discussed utilizing the tools and techniques they learned during training to implement GOC (Table 1). However, most participants noted that the effective execution of those tools was only possible if one first adopts a person-centered care approach (89%) and establishes trusting relationships with patients (78%).

Given the medical and social complexities of the patient population, participants discussed needing to get to know the person behind the patient before assessing and addressing their care needs.

… we have the [GOC] tools … But I don’t tend to bring them to my first meeting with somebody. It’s basically just to get to know them … just sitting down talking to somebody … And you’re finding out … what are their goals. – VT-11 (Care coordinator, social services and housing agency, organization 4)

A person-centered approach to care, used synonymously with taking a whole-person approach, was deemed the “right approach” by most participants, some of whom described always delivering care in this way.

I wrote my essay on [the whole-person approach to care] when I was applying to medical school … So it’s been just my own way of looking at life really … The biopsychosocial aspects of one’s care is what … interested me and why I chose primary care … The whole person, the family system, you know, the community where they live, the services available to them … I don’t think you can really understand their goals if you don’t try to get to know them as a whole person first. – VT-01 (Family physician, academic health network, organization 1)

Others adopted a person-centered approach while working under the GOC model, which was “a huge challenge because it’s a whole new way of thinking” – VT-06 (Care coordinator, state-led initiative, organization 3).

Participants stated that central to person-centered GOC was empowering patients to identify their own goals, which may reflect both medical and non-medical needs. For this reason, participants discussed the importance of addressing social determinants of health as part of providing person-centered GOC. Sometimes, the social determinants of health were seen as more important than medical needs. Team members demonstrated this mental model at their meeting during which they collegially highlighted that a colleague was prioritizing their professional goals for their patient over the patient’s goals that were more related to housing and social needs. Team members asserted that patient goals, which often relate to social determinants, should be prioritized.

I mean every single person who comes through my door, I’d be lying if I didn’t say I have my own goals for them. Part of my ethics as a mental health counsellor and part of my personal ethics [is] I try to work for the autonomy and good of the individual … if I focus on what I want to do, I’m not really helping them – VT–03 (Behavioural therapist, state-led initiative, organization 3)

I think the biggest thing for me is that the plan of care always needs to be meaningful to the patient and meet the patient where they’re at. It’s not always what’s going to be best clinical practice or what’s medically indicated by the provider. It needs to be more that bigger idea of what health encompasses – VT-13 (Quality specialist, primary care practice, organization 7)

Team members also viewed developing trusting relationships with patients as the foundation of GOC hand in-hand with a person-centered approach. They explained that trust-building should not be rushed as it establishes a foundation for patients to share meaningful goals.

To me it’s about relationship. They [the patients] have to build a relationship with their treatment providers. They have to feel safe … it can vary how long it takes to build that relationship … But as you see trust build, I think that’s when you begin to see people open up more and be more clear about what their goals are and what the barriers to those goals are. – VT–07 (Director, mental health agency, organization 5)

For me, it’s about relationship and connecting first … if I don’t develop a relationship with this person … it’s not going to work, they’re not going to want to talk to me … they’re not going to trust me … I make it about building a relationship – VT-10 (Care coordinator, primary care practice, organization 7)

The role of training in fostering SMMs of GOC was notable, emerging as the fourth most common task-related mental model content area (56%). Training was provided at the onset as well as throughout the implementation of GOC with the support of a practice facilitator and regular Learning Collaborative meetings.

The trainings that we went through … were so inspirational and had really good nuggets of information … there’s no magic bean for this, there’s no magic one way of doing it. It’s a philosophy. – VT-02 (Director, hospital, organization 2)

Once the [learning] collaborative ended, we decided to continue on with our meetings … we still have work to do in getting people trained in using some of the tools. We’re still in that learning process of making that mind shift of our own and empowering people. – VT-14 (Program manager, state-led initiative, organization 3)

Team-Related mental models of goal-oriented care

The most shared team-related mental model contents were engaging in formal and informal communication (94%), taking a team approach to care (89%), building knowledge of and trusting relationships with each other (89%), and clarifying respective roles (83%). Participants stated that GOC cannot be accomplished without taking a team approach. However, to work as a team and to coordinate tasks across professional and organizational boundaries, they needed to establish ways of communicating effectively, build relationships with one another, and clarify roles. These four mental model content areas were deeply intertwined.

As an inter-organizational team, they faced the challenge of working with multiple communication platforms and thus emphasized the need for technological solutions as well as establishing formal cross-boundary communication mechanisms to enable their work. At the time of data collection, the team was developing a communication platform that would enable contact between providers and between the patient and their entire care team, regardless of organizational affiliation. In the meantime, to prevent gaps in communication team members relied on e-mails, phone calls, biweekly face-to-face meetings, and case conferences to discuss how to meet patient goals and coordinate tasks. During team meetings, members frequently encouraged their colleagues to “email, call, whenever” and explicitly made themselves known as available resources. Structured meetings thus enabled the creation of informal communication channels.

… [learning collaborative meetings are] a chance for everyone to get to know each other really closely. And so we would start pulling each other aside after meetings and saying, “Hey, can we meet?” [and] “I’m here to talk about this patient but let’s talk about this one as well, and then we’ll meet about her next week.” And so we started having more casual interactions more frequently – VT-02 (Director, hospital, organization 2)

A lot of coordinating, a lot of meetings, a lot of phone calls, emails, updates … oftentimes it’s just picking up the phone and having that conversation. Or I will email the care coordinator and ask her if she can get this message to the doc … so the care coordinators are able to pass it on to the doctor and then they’re able to know what’s going on. – VT-08 (Case manager, agency on aging, organization 6)

In addition, team meetings proved to be important opportunities to clarify each other’s expertise and roles as well as develop trusting relationships. Through regular interaction, team members learned how their colleagues could contribute to different cases to address patients’ needs, which simplified the process of coordinating care. They also learned to lean on one another for professional advice and emotional support.

… they’ve always worked in their silo, and primary care has worked in their silo … but just getting people in the same room together is sometimes all it takes … And the more times you see these people and get to talk and figure out what they do and what you do, people start thinking … gosh, I could call on, you know, this organization to help me with this … especially in our learning collaborative meetings. Like this has just been invaluable, I know who to call at [mental health agency], at [home care agency] … [the hospital] … you can just pick up a phone and be like I have this issue, can you help me? And so building those relationships has been really important – VT-14 (Program manager, state-led initiative, organization 3)

So we have a lot of support. And that’s a huge … it’s great because we tend to talk to each other quite a bit or email, you know. We’re all in our own little worlds and our own little offices. But then you get these cases that just eat at you. And you know you’ve got to do something but you’re not sure what. And bringing it to the collaborative, it’s getting everybody else’s perspective – VT-12 (Care coordinator, social services and housing agency, organization 4)

Similar to the development of task-related mental models, leaders emphasized the important role of education and training sessions during which team members learned about how to work effectively as a team (i.e., achieve role clarity, communicate, manage conflict), thereby building team-related SMMs.

… we wrote a document called Effective Teaming. And it’s designed to be a resource … it talks about how to make a team meeting work, what are the values that underline teaming, what are the roles of people in the group … And we use that document as a … training tool. – VT-07 (Director, mental health agency, organization 5)

Discussion

In this study, we found shared task-related and team-related mental models of GOC in an ICBPHC team with members that span professional and organizational boundaries. We explicated the contents of their SMMs and explored how they developed. Overall, we found that task-related and team-related mental model contents were highly shared among participants. The most shared mental model content areas were operationalizing the GOC process (100%), engaging in formal and informal communication (94%), taking a “whole-person” approach (89%), taking a team approach (89%), and building trusting relationships both with patients (78%) and with other team members (89%). These content areas are complementary; the first two focus on the practical and tangible aspects of implementing GOC (“the what” of executing tasks and team interactions, respectively), while the remaining three are more intangible and emphasize the philosophical, affective, and social underpinnings of GOC (the “how” and “why”).

Trust was the only construct that spanned both task-related and team-related mental models, highlighting the importance of both intra-team trust and provider-patient trust to the implementation of GOC. Trust is frequently cited in the integrated care and team science literature as a requisite for collaboration (Amour et al., 2005; Aunger, Millar & Greenhalgh, 2021; Kirst et al., 2017; Li et al., 2018; Looman et al., 2021). In a study of integrated care networks in the English National Health Service, a hesitancy to trust team members across professional and organizational boundaries resulted in duplication of tasks (Li et al., 2018). In our study, the finding that the ICBPHC team valued trusting their teammates may explain how the team was able to delegate tasks and function cohesively across professional and organizational boundaries.

Patient-provider trust is also a long-known foundational mechanism for high-quality care (Brennan et al., 2013; Murray & McCrone, 2015). When it comes to working with patients who are living with medical and social complexities, building trusting relationships may be even more critical. For these patients, opening up about social needs, such as issues related to housing, food insecurity, and being underinsured, may be difficult and even unfeasible without first establishing trust (Andermann, 2016). Our study supports prior research arguing for provider and staff training on communication about sensitive and stigmatizing issues (Schoenthaler, Hassan & Fiscella, 2019). Participants’ SMMs regarding the role of trust in both intra-team and provider-patient interactions is notable, reinforcing the need to look beyond the tangible tools of GOC to the psychosocial dynamics of human relationships that shape the potency of GOC practices.

The literature is clear that SMMs are critical for team performance (Dechurch et al., 2010; Fiscella, Mauksch, Bodenheimer & Salas, 2017; Mohammed et al., 2010; Müller & Antoni, 2021). However, they may serve an even more important function when implementing interorganizational teams due to the lack of colocation and organizational boundedness (Morrison & Jaime, 2020; Schmidtke & Cummings, 2017). In this study, team members developed SMMs through training and meetings which leadership prioritized by dedicating time and resources. These relationship-building opportunities appeared to be the crux for fostering and maintaining SMMs as suggested by the value placed on trust by participants. Our finding regarding the importance of training, meetings, and frequent interactions in team relationship-building is in alignment with both the literature on primary care transformation (Giannitrapani et al., 2018; Helfrich et al., 2016) as well as the SMM literature (Marks, Sabella, Burke & Zaccaro, 2002; Stout, Cannon-bowers, Salas & Milanovich, 1999). Particularly as teams work in fluidity with changing membership, our findings underscore the importance of providing consistent training with collaborating organizations and opportunities to interact to foster SMMs.

We faced the challenge of determining how to measure “sharedness.” We decided against introducing a strict definition for “sharedness” or imposing an arbitrary cutoff. This ambiguity reflects a trend in the literature on SMMs where measuring sharedness is not clearly defined in studies. Several reviews have noted the elusiveness of defining and measuring SMMs (Floren et al., 2018; Gisick et al., 2018; Mohammed et al., 2010). Floren and colleagues suggest eliciting mental model contents at the individual level, classifying contents, and then assessing the sharedness at the aggregate team level, as we have (Floren et al., 2018). While this is a helpful starting point, questions remain around what constitutes sharedness and the extent to which sharedness is required for effective teamwork. We found that certain mental model contents were more commonly held among team members which raises the question of whether it is sufficient for only select mental model content to be widely shared for optimal team functioning.

Another challenge we faced was regarding distinguishing between knowledge and beliefs, which is consistent with previous SMM studies (Mohammed & Dumville, 2001; Mohammed et al., 2010). As noted in the methods, we opted not to make this distinction. Doing so would require in-depth cognitive inquiry to unpack how and why individuals came to the mental models they held and any assumptions (explicit or implicit) that may influence their thinking. Future studies are needed to address underexplored methodological and practical questions regarding SMMs both theoretically and empirically, such as those we raise above.

Strengths and limitations

The first limitation of this study is that SMM theory was identified post hoc; we did not probe participants using an interview guide informed by SMM theory. However, this may also be a strength of the study. SMMs emerged as important in the data spontaneously, which is in line with the primary aim of inductive qualitative research (Pratt et al., 2020). Second, the cross-sectional nature of the data means that findings do not reflect how mental models evolved over stages of implementation. Finally, the data is based on one case study site during which health system reform toward population health management were political priorities, limiting the transferability of findings (Grembowski & Marcus-Smith, 2018).

The main strength of this study is the use of a theory-driven and systematic approach to identify SMMs of GOC among an interprofessional and inter-organizational team. The identified SMM contents offer practical guidance for primary healthcare teams that are interested in implementing GOC, particularly for interorganizational teams. Explicating SMM contents of GOC is unique, as literature on team science and primary care evaluations tends to focus on tangible factors influencing performance such as team composition, rather than the powerful invisible dynamics that shape teamwork, such as team cognition. Additionally, we contribute to SMM theory in two key ways: first, we examined SMMs in a team that crosses organizational boundaries, and not co-located teams with a common employer and vision; second, this study heeds the call in the literature to shift from a focus on generic SMM content to examine domain-specific content, namely, GOC implementation (S. McComb & Simpson, 2013). This study thus provides deeper insight into what SMMs look like in practice.

Conclusion

This study used theory-driven qualitative methods to systematically explicate the contents of SMMs in an ICBPHC team that adopted GOC. Study results provide insight into the mechanisms that underlie taskwork and teamwork in efforts to implement GOC for people with complex health and social needs. The findings of this study can guide future implementation of primary healthcare teams that wish to adopt a goal-oriented approach and inform theoretical and empirical advances in SMM theory. Further research is needed to validate which of the identified SMM contents are necessary or sufficient for implementing GOC in different health system contexts. Research is also needed to advance the qualitative measurement and analysis of SMMs.

List of abbreviations

GOC goal-oriented care

ICBPHC integrated community-based primary healthcare

SMM shared mental model

Ethics approval and consent to participate

The Vermont Agency of Human Services Institutional Review Board, and the Toronto Academic Health Sciences Network (#16-0321-E) research ethics boards approved this study prior to data collection. All participants provided informed consent for their participation prior to data collection.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13561820.2022.2113048

Data availability statement

The data that support the findings of this study may be available on request from the senior author, CSG, and approval from research ethics boards. The data are not publicly available because they contain information that could compromise the privacy of research participants.

Additional information

Funding

This work was supported by a start-up fund held by the senior author (CSG) at Lunenfeld-Tanenbaum Research Institute, Sinai Health System. Additionally, this research was undertaken, in part, thanks to funding from the Canada Research Chairs Program.