Interprofessional Collaboration and Patient Health Outcomes in Urban Disadvantaged Settings: A Grounded Theory Study

INTRODUCTION Interprofessional interventions improve the ability of health professionals to work in teams, communicate effectively, respect and appreciate each other, and develop shared patient-centric values. While these interventions can change attitudes and beliefs about interprofessional collaboration, the relationship between interprofessional collaboration and patient health outcomes remains poorly understood, particularly for socioeconomically disadvantaged populations. This study sought to explore the relationship between interprofessional collaboration and patient health outcomes in urban disadvantaged settings. METHODS Constructivist grounded theory methodology was used to perform 4 focus groups and 19 individual interviews with health professionals working in these settings in the United States. Emergent themes were developed into a conceptual model that captures their views on the link between interprofessional collaboration and patient health outcomes in these settings. RESULTS 114 qualitative themes were identified and collapsed into 10 theoretical categories (interprofessionalism, building trust, coordination, facilitating sharing, patient care, enhancing reciprocity, common goals, effecting change, healthcare system disparities, and patient individual, group or population disparities), all of which were then merged into two theoretical concepts that explain all of the data (social capital and disparities). CONCLUSION Interprofessional collaboration works via better coordination and optimization of patient care, which explains how better patient health outcomes may be achieved. However, it is social capital and its cognitive elements of trust, sharing and reciprocity that underlie this phenomenon and explain why better health outcomes may be possible via interprofessional collaboration. Additional research studies exploring patient perspectives and the structural elements of social capital in this context are warranted. Received: 05/19/2019 Accepted: 10/01/2019 © 2019 Jadotte, et al. This open access article is distributed under a Creative Commons Attribution License, which allows unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. HIP E & Interprofessional Collaboration and Patient Health Outcomes ORIGINAL RESEARCH 3(4):eP1185 | 2 HIP E &


Introduction
Effective collaboration among health professionals is key to optimizing patient health outcomes. The increased complexity of healthcare systems (Institute of Medicine, 2001), the high degree of specialization within the health professions (Irvine, Kerridge, McPhee, & Freeman, 2002), the worsening non-communicable disease burden (World Health Organization, 2012) and the increasing scarcity of resources (Institute of Medicine, 2012) require healthcare delivery that is well coordinated, team-based and patient-centered to achieve optimal patient health outcomes within resource constraints.
Interprofessionalism, interprofessional collaboration, or interprofessional collaborative practice (IPCP), embody the philosophy that when two or more individuals from two or more health professions learn with, from and about one another and work together, there can be substantial improvements in the quality and efficiency of care, and its delivery becomes more cohesive and patient-centered (Herbert, 2005;World Health Organization, 2010). Thus, IPCP is the act of collaboration by continuously learning with, from and about each other, preferably with involvement of the patient, family and community, in the process of improving healthcare quality and optimizing patient health outcomes (Australian Interprofessional Practice and Education Network, 2012;World Health Organization, 2010). It is the construct that interprofessional learning (IPL) interventions are seeking to influence.
Interprofessional learning interventions that are designed to stimulate IPCP include interprofessional education (IPE), practice (IPP) and organization (IPO) (Reeves et al., 2011). IPE consists of all educational initiatives where students or professionals from at least two different health professions learn with, from and about each other to facilitate effective collaboration and improve the quality and efficiency of care as well as patient health outcomes (Centre for the Advancement of Interprofessional Education, 2002;World Health Organization, 2010). Examples includes lectures, case simulations, and other didactics engaging students from more than one health profession. IPP, on the other hand, consists of all endeavors that are interprofessional in nature but move beyond the educational realm, where health professionals are collaborating actively with each other, including learning with, from and about each other in the practice setting, to accomplish the common goals of improving healthcare and patient health outcomes (Australian Interprofessional Practice and Education Network, 2012;World Health Organization, 2010). IPO is a special type of IPP that expands the roles of health professionals to include health policy development and other organizational-level methods of collaborative engagements (Reeves et al., 2011). Exam-

Implications for Interprofessional Practice
• Interprofessional collaboration builds trust among health professionals, patients, their families and communities.
• Improved trust facilitates the sharing of information, decision-making roles and care delivery tasks, which are all components of good care coordination.
• Better care coordination leads to improved reciprocity in patient care from all stakeholders, including greater continuity of and adherence to care, provision of holistic care, and implementation of supportive institutional policies.
• Patient health outcomes can be improved when all stakeholders are responsive to all the elements of the relationship between interprofessional collaboration and patient care.
• Social capital, and its corollaries of trust, sharing and reciprocity, are the foundations of the relationship between interprofessional collaboration and patient health outcomes.
ples of IPP include interprofessional meetings, briefings or rounds among licensed health professionals working on the wards or other clinical settings, while examples of IPO include team-based quality improvement initiatives and the collaborative formulation of organizational staffing policies (Reeves et al., 2011;). All of these IPL interventions are thought to ultimately lead to better patient care and health outcomes and reduce healthcare costs (Interprofessional Education Collaborative Expert Panel, 2011).
A systematic review (Lapkin, Levett-Jones, & Gilligan, 2011) has shown that IPL initiatives improve IPCP educational outcomes, such as student and health professional attitudes, beliefs, knowledge, skills and abilities (Canadian Interprofessional Health Collaborative, 2012). Several other systematic reviews have shown that IPL interventions also improve IPCP healthcare outcomes, such as behaviors, organizational practice, and patient and provider satisfaction (Canadian Interprofessional Health Collaborative, 2012). For example, IPL reduces length of stay and improves patient care management (Reeves, Perrier, Goldman, Freeth, & Zwarenstein, 2013;Reeves et al., 2008;).
However, several knowledge gaps persist. The lack of empirical evidence on the relationship between IPCP and patient health outcomes is pervasive. Health outcomes can be defined as those that can be measured objectively using modern medical diagnostic tools (ex. quality of life, survival from heart attacks, and the incidence or prevalence of diabetes mellitus). Patient health outcomes are directly related to population health outcomes, which accounts for the differential distribution of health outcomes in a population and its various social groups (Kindig, 2007). Many studies explicitly call for measuring the relationship between IPCP and patient health outcomes and exploring how IPCP effects changes in healthcare and patient health outcomes (B. F. Brandt, 2014;Reeves et al., 2013;Reeves et al., 2008;Zwarenstein et al., 2009), yet this call remains largely unanswered in the literature. This evidence gap is particularly poignant when IPCP is envisioned in the context of the health and healthcare disparities (Fiscella & Epstein, 2008;Institute of Medicine, 2002;Krist, Johnson, Callahan, Woolf, & Marsland, 2005;Perloff, Kletke, & Fossett, 1995) faced by socioeconomically disadvantaged communities (Vanderbilt et al., 2013;Xyrichis & Lowton, 2008).
Social capital, which consists of the sum of the cognitive and structural or interpersonal linkages that connect people together (Islam, Merlo, Kawachi, Lindstrom, & Gerdtham, 2006), has been proposed as a possible theoretical framework by which IPL acts to enhance IPCP (Gloede, Hammer, Ommen, Ernstmann, & Pfaff, 2013;Godley & Russell-Mayhew, 2010) and thereby potentially improve patient health outcomes. However, this proposed mechanism has not been demonstrated inductively or grounded in any empirical qualitative study to date. Furthermore, while social capital may explain how IPL improves the cognitive and structural functionality of collaborative relationships among health professionals (Godley, Barron, & Sharma, 2011;Godley & Russell-Mayhew, 2010), no explanatory mechanism on how IPCP works to change patient health outcomes, a more downstream but highly prized effect, has been proposed.
These knowledge gaps must be addressed to achieve interprofessional collaboration at all levels of the educational continuum (Institute of Medicine, 2013), establish a lasting and meaningful commitment to and a real partnership with patients, families and communities -particularly those facing substantial disparities in health and healthcare -and improve population health outcomes (Institute of Medicine, 2014). IPCP has substantial implications for health professional education, policy, and practice in the United States (US) (Reeves et al., 2011). Health professions students must increasingly be more interprofessionally competent to maximize patient-centered care, and health educators require empirical evidence on the association between IPCP and patient health outcomes to justify crafting appropriate and effective educational policies (Hammick, Freeth, Koppel, Reeves, & Barr, 2007) and validate the curricular utility of interprofessional rather than uni-professional education (Thistlethwaite, 2012). Hospital and healthcare systems administrators and clinicians need research evidence demonstrating how the implementation of IPL will improve patient health outcomes and reduce costs (B. F. Brandt, 2014;Interprofessional Education Collaborative Expert Panel, 2011). The purpose of this study is to inductively explore the relationship between IPCP and patient health outcomes in disadvantaged settings in the US from the perspective of health professionals working in this context.

Review of the Literature
The exact mechanisms linking IPL interventions with healthcare and patient health outcomes have long been unclear (Barnsteiner, Disch, Hall, Mayer, & Moore, 2007;Barr, Freeth, Hammick, Koppel, & Reeves, 2006;Barr, Hammick, Koppel, & Reeves, 1999). Several systematic reviews have explicitly called for the conduct of research studies that seek to explain how IPL interventions lead to changes in healthcare and patient health outcomes (Reeves et al., 2013;Reeves et al., 2008;Zwarenstein et al., 2009). A search of the literature via Google Scholar, Medline, CINAHL and Scopus, using the key words "interprofessional" and "health" or "patient" or "professional" yielded 9 relevant qualitative research studies ( Most of these studies used an unspecified qualitative descriptive methodology (Adams et al., 2014;Bradley Eilertsen et al., 2009;Chong et al., 2013;Eloranta et al., 2010;Fredheim et al., 2011;Goldman et al., 2010;Hjalmarson et al., 2013;Maneze et al., 2014), except one that used a grounded theory approach (Bajnok et al., 2012). None of the included studies took place in the US: 3 studies were done in Canada (Adams et al., 2014;Bajnok et al., 2012;Goldman et al., 2010), 2 studies were from Australia (Chong et al., 2013;Maneze et al., 2014), 2 studies took place in Norway (Bradley Eilertsen et al., 2009;Fredheim et al., 2011), 1 study was from Finland (Eloranta et al., 2010), and 1 study originated from Sweden (Hjalmarson et al., 2013). Participants included a range of health professionals as well as patients and their family members or care takers. The studies were conducted in a variety of clinical settings, including primary care, acute care hospitals, community health centers, or a mix of these and other settings.
Only one study done in Australia was conducted in a low-resource setting (Maneze et al., 2014).
The following 3 broad types of qualitative research questions were posed in these studies: What does it take to establish functional interprofessional teams from the perspective of health professionals (Adams et al., 2014;Bajnok et al., 2012)? What is the overall experience of health professionals (Bradley Eilertsen et al., 2009;Chong et al., 2013;Eloranta et al., 2010;Hjalmarson et al., 2013) and patients (Maneze et al., 2014) with interprofessional collaboration? What are the strengths and weaknesses of (Fredheim et al., 2011;Goldman et al., 2010) and barriers and facilitators to (Maneze et al., 2014) interprofessional collaboration? Although these studies addressed research questions that generated themes related to the relationship between IPCP and patient health outcomes, these themes were incidental findings as none of the studies sought out to explore this relationship explicitly, suggesting that these studies were unlikely to capture the types of rich data that emerge from having implemented an explicit and exhaustive qualitative methodology such as grounded theory (Charmaz, 2006).
Finally, a systematic review of the qualitative literature, using the meta-aggregation approach of the Joanna Briggs Institute (Lockwood, Munn, & Porritt, 2015), was conducted to synthesize the evidence on the phenomenon of interest (Y. Jadotte, Holly, Chase, Powell, & Passannante, 2016). This review demonstrated that there were no studies conducted from the perspective of health professionals working in low resource settings in the US and no studies that inductively explored the link between IPCP and patient health outcomes using grounded theory methodology (Y. Jadotte et al., 2016). However, while there is a lack of qualitative evidence on the relationship between IPCP and patient health outcomes, there is some qualitative evidence on the relationship between IPCP and healthcare outcomes.
Specifically, the IPCP healthcare outcomes of patient and professional satisfaction have been given some credence in the literature as potential links in the pathways to achieve the downstream goal of better patient health outcomes (Herbert, 2005). Prior research has shown that health professional satisfaction is an important measure of the quality of the work environment and a key determinant of sound healthcare delivery. For example, a systematic review of studies examining physician satisfaction with healthcare practice in the United States has found that physician satisfaction is impacted by job-related indicators, including collegial support and demands of the work environment. This review also reported a greater impact of these factors on primary care physicians (Scheurer, McKean, Miller, & Wetterneck, 2009). A large cross-sectional study among nurses had similar findings, showing that work factors such as supervisor support and work-group cohesion impacted professional satisfaction (Kovner, Brewer, Wu, Cheng, & Suzuki, 2006). However, the mechanism by which IPL leads to health professional satisfaction itself remains elusive. Social capital has previously been identified as a mechanism to explain this relationship (Soubhi, 2010). For example, one study suggests that social capital has implications for both the structure and content of social networks at work, moderates the general social climate at work and influences employee cooperation with managers and colleagues (Flap & Völker, 2001). Other studies have found that social capital in the workplace enhances the coordination of care among hospital staff (Gloede et al., 2013), minimizes clinician emotional exhaustion (Driller, Ommen, Kowalski, Ernstmann, & Pfaff, 2011), supports hospital-wide knowledge sharing, significantly influences patient safety outcomes (Chang, Huang, Chiang, Hsu, & Chang, 2012), promotes patient-oriented customer service behaviors that enhance patient and provider satisfaction (Hsu, Chang, Huang, & Chiang, 2011), and is a strong predictor of job satisfaction and quality of life at work (Felix, 2003). In addition, studies have found that interprofessional collaboration among health researchers functions through social network ties and is amenable to a social network analytic lens (Godley et al., 2011;Godley & Russell-Mayhew, 2010). Social capital may thus serve as a missing bridge between IPL, as an educational or practice intervention, and the optimization of healthcare and patient health outcomes, the most prized goal. While some of the links between IPL, social capital and some IPCP healthcare outcomes, such as patient or healthcare professional satisfaction, have been explored to date, many aspects remain unsolved. For example, there is little literature on the role of social capital as a potential resource that health professionals can tap to mitigate the social determinants of health, particularly in socioeconomically challenging settings in the US. There is also no evidence on how the health and healthcare disparities embedded in these settings in the US might influence or constrain the capacity of health professionals for collaborative practice.

Methods
This study explored the relationship between IPCP and patient health outcomes, from the perspective of health professionals working in disadvantaged settings in the US. A review of the literature shows that no prior conceptual models have been developed using an inductive research design to address this particular research question. Thus, a constructivist grounded theory design (Charmaz, 2006) was implemented to allow potential theoretical mechanisms that may explain this relationship to emerge from the inductively-derived data. The central research question was: How do health professionals who work in urban disadvantaged settings in the US perceive the relationship between IPCP and patient health outcomes within the context of healthcare and health disparities embedded in these settings?

Data Collection
Four semi-structured focus group (totaling 23 participants from a variety of health professions) and 19 individual interviews (with 8 participants who identified themselves as members of the "allied health professions", plus 6 physicians, 4 nurses, and 1 dentist) were conducted with licensed health professionals who had worked in the targeted setting for at least six months. All focus groups had at least 4 participants and an interprofessional make-up: at least 2 different health professions were represented in each of them. A semi-structured interview instrument that was piloted with a smaller sample and yielded good data relevant to the research question (Y. T. Jadotte, 2014) was used to guide the data production process. Focus groups allowed the capture of co-constructed data among participants (Charmaz, 2006), while individual interviews allowed themes related to the longstanding power differentials within the health professions to emerge (Hall, 2005). The theoretical purposive sampling approach of grounded theory was used: additional interviews were conducted until the constant comparative analytic method yielded theoretical saturation, wherein no new codes or iterations of codes emerged in several consecutive transcripts (Charmaz, 2006), at which point data collection ended. All interviews were audio-recorded using a professional high-quality portable microphone and digital device and transcribed by a professional transcriptionist. The transcripts and audio-recordings were reviewed together by the principal investigator to ensure full transcriptional data accuracy. The final sample size consisted of 42 participants and 23 transcripts.

Data Analysis
The study followed the constant comparative analytic method: the textual data from each transcript were analyzed by the principal investigator and a second coder, and all newly identified themes were integrated into an emerging theory grounded in the data, prior to conducting the next interview and gathering more data.
The results of this analysis guided the collection of further data for analysis (Charmaz, 2006). Open, focused and theoretical coding were used to code, categorize and relate the data into a plausible theory on the relationship between IPCP and patient health outcomes in disadvantaged settings. An axial coding strategy, which is more consistent with objectivist grounded theory (Charmaz, 2006), was not used in this study to allow a more data-grounded theory to emerge with less direct influence from the coders and investigators. The principal investigator and another coder not affiliated with the investigators evaluated the reliability of the coding strategy using an emergent codebook to code a random sample of 17%, or 4/23, of the transcripts. Independent coding of a 10-20% random subset of a qualitative study's transcripts is considered sufficient to perform the calculation of inter-rater reliability (Hruschka et al., 2004;Lombard, Snyder-Duch, & Bracken, 2004). The coding strategy had an 85% level of agreement between the two coders, which satisfies acceptable qualitative methodological quality control benchmarks (Hruschka et al., 2004).
An incident-by-incident coding approach, rather than a line-by-line approach, was implemented to allow the identified codes to emerge into themes and categories that meaningfully capture both discrete and overlapping ideas on the phenomenon under study (Charmaz, 2006). Directed content analysis was used for coding. This approach entails the a priori development of a codebook based on pilot data or theoretical knowledge of the literature, and the application of the codebook to the data produced in a study (Hsieh & Shannon, 2005). Any text that did not fit the existing codes was assigned a new code, which is congruent with constant comparative analysis and the need to achieve theoretical saturation. This coding strategy facilitated the inductive creation of a conceptual model grounded in the data that provided a mechanistic framework to explain the association between IPCP and patient health outcomes, within the context of disadvantaged settings in the US. The software program Dedoose was used to code and categorize the qualitative data, while the graphical software program OmniGraffle helped to create diagrams of the conceptual model.

Ethical Review and Approval
This project was approved as an expedited study with minimal risk by the institutional ethical review board of Rutgers Biomedical and Health Sciences, protocol number 20140001018.

Findings
The incident-by-incident open and focused coding approach in this study yielded 114 unique codes, resulting from data reduction for 1809 excerpts from the participants. By further reducing these 114 codes into 10 unique categories and two overarching theoretical concepts (as listed in table 1) that captured all categories and codes, theoretical coding facilitated the organization of all themes into a plausible mechanistic model that answers the research question. The identified theoretical concepts expanded the explanatory power of the model as they are established and well-researched constructs. Figure 1 diagrams this conceptual model to facilitate the illustration of its pathways. Key excerpts are provided in Appendix A to capture the participants' voices in support of the identified themes. Categories are listed in bold, while codes and sub-codes (which are codes that could be subsumed into other codes) are italicized. The relationship between the themes that emerged and current knowledge in the literature is covered in the discussion section, as part of the additional literature review at the conclusion of a grounded theory study (Charmaz, 2006). During the theoretical coding stage, a dialectical relationship emerged among pairs of the 10 categories which helped make sense of these data. As such, findings are presented together for those pairs: building trust and interprofessionalism; facilitating sharing and coordination; enhancing reciprocity and patient care; effecting change and common goals; and healthcare system disparities and patient individual, group or population disparities. Bringing efficiency to healthcare systems, improving patient outcomes Disparities Healthcare system disparities Differential system demeanor toward stakeholders, facility-provider disconnect, healthcare misconduct, information technology/electronic health record limitations, differential support systems for teams, healthcare facility differences, sociopolitical disparities can lead to poor care delivery or care fragmentation Patient, individual, group or population disparities Empathy gap, lack of trust, differential expectations, complex problems, low socioeconomic status Table 1. List of theoretical concepts, categories and codes.
The findings for the themes of Building Trust and Interprofessionalism show that health professionals believe IPCP consists of having an appreciation for the different specialties, valuing diversity in professional skills, communicating effectively, and working well in teams. They also suggest that health professionals believe IPCP does not always happen without specific interventions to nurture it and that it must be grounded in common goals shared among all health professionals, including the need to bring greater efficiency to the healthcare system and improving patient outcomes.
The findings for the themes of Facilitating Sharing and Care Coordination suggest that health professionals view care coordination as the central element through which interprofessionally competent healthcare teams can improve patient health and system outcomes. Three important components of care coordination were unearthed: effective care delivery, purposeful information sharing, and reliance on team-based decision-making that truly values and includes the perspectives of all health stakeholders at all times.
The themes of Enhancing Reciprocity and Patient Care emerged as the next step in the pathway. Specifically, the concept of patient care is broken down into its subcomponents that can be impacted by IPCP-driven care coordination. According to health professionals, effec-tive patient care means ensuring the continuity of care, facilitating patient adherence to care plans, and establishing policies that support rather than inhibit care. It is this type of care that is most likely to yield the desired population health outcomes and healthcare system goals. Effective patient care is thought to lead to real change and improvements in the Common Goals of all health professionals, including subjective and objective patient and system outcomes.
The themes of Health and Healthcare Disparities emerged as significant barriers to the optimization of patient health outcomes using IPCP. The data shows that both systemically-embedded disparities and disparities individually embodied in some patients are thought to be a major impediment to the provision of team-based, coordinated, continuous, and patient-centered care. A number of healthcare systems disparities are perceived to be critical to this pathway, including differences in the types of support systems that are available to facilitate teamwork (such as the unavailability of potential collaborators or the long distances separating them), the need for health professionals to adapt to roles and tasks beyond their usual practice, and the fact that the healthcare system treats its stakeholders differently, which influences the capacity of these stakeholders to work together in teams to achieve common goals. Other perceived system disparities include: the inherent power differentials amongst health professionals and between health professionals and patients; the large gaps in compensation for health professionals who work in poor and underserved areas, as well as for those who work in preventive or primary care; the differences in the healthcare facilities themselves (including the availability of language services, large patient caseloads and prolonged waiting times); the unevenly distributed capabilities and limitations of information technology or electronic health record systems; and the numerous inequitable healthcare laws and policies that disparately affect some populations more than others, resulting in either challenges in care delivery or the fragmentation of care. Information sharing is another element that can be severely impeded by healthcare system disparities. More specifically, there is often a substantial disconnect in the sharing of information between health professionals and healthcare facilities, which is thought to be responsible for much of the healthcare misconduct that is associated with some locales. This includes patient misconduct, such as doctor shopping and misuse of opioid prescriptions, and health professional misconduct, such as accepting kickbacks for referrals.
Certain disparities were thought to apply specifically to disadvantaged populations. Low socioeconomic status was seen as the most substantial disparity for these patients, and it was linked to patients being essentially denied access to care and being forced to choose between life and their health (such as having to choose to pay for food over medications). These situations were perceived as frustrating to the health professionals who care for them. Because of these formidable challenges to care, many patients (particularly those who have been severely disadvantaged and historically mistreated by the healthcare system, such as poor Blacks and Latinos) have a lower level of trust in the healthcare system and the health professionals who attempt to care for them, have learned to be helpless and fail to demand that their care be collaborative in nature, and often give up on pursuing adequate care. The ultimate results of these disparities in the patient's support system are: greater failure to adhere to all forms of care, even when it may be available to them and is offered by health professionals working collaboratively, and the embodiment of the disparities in the healthcare system and the patient's environment in the form of presenta-tion at advanced stages of disease and emergence of chronic diseases at much younger ages than otherwise expected.

Discussion
A prior qualitative systematic review of the literature had already identified a number of key themes on the manifest pathway between IPCP and patient health outcomes that are congruent with the findings of this study (Y. Jadotte et al., 2016). These include: the attainment of the 4 IPCP core competencies as a fundamental initial goal; care coordination as the most proximal variable affected by IPCP; and patient care as the next most relevant step in the pathway. Other key ideas that were already known from this synthesis of the literature include: the importance of team leadership as a key element in care coordination; the value of shared decision-making and problem-solving in healthcare; the role that continuity of care (or lack thereof) as well as social, economic and cultural barriers can play in patient care; and finally the impact that the lack of mutual accessibility and commitment on the part of health professionals and healthcare systems can have on the successful achievement of IPCP (Y. Jadotte et al., 2016). While the manifest pathway (i.e. the how) of the relationship between IPCP and patient health outcomes was known, the literature to date had not yet addressed the latent pathway, which answers the question of why. For example, why does IPCP facilitate care coordination? Although it was generally understood that IPCP should affect care coordination, that better care coordination should lead to better patient care, and that improved patient care should optimize health outcomes (Y. Jadotte et al., 2016), the mechanisms via which these effects were accomplished remained unclear.
This study provides an inductively developed and detailed model for how these complex relationships manifest. It also empirically identifies care coordination as one plausible mechanism by which IPCP acts to change patient health and system outcomes. While there is currently no literature on the relationship between IPCP and care coordination, a prior systematic review aimed to develop a working definition of care coordination, and to identify theoretical frameworks linking care coordination to healthcare factors and interventions and patient health outcomes (McDonald et al., 2007). The authors defined care coordination as: "the deliberate organization of patient care activities between two or more participants (including the patient) involved in a patient's care to facilitate the appropriate delivery of health care services. Organizing care involves the marshalling of personnel and other resources needed to carry out all required patient care activities, and is often managed by the exchange of information among participants responsible for different aspects of care" (Mc-Donald et al., 2007, p. 41). The themes of information sharing and care delivery modeling identified in this review were also found in this grounded theory study. However, this study clarifies that "the marshalling of personnel and other resources" may signify a need to establish clear mechanisms for decision-making.
The critical role of health information technology in information sharing is also an important contribution of this study to the interprofessional care literature. A prior systematic review examining the role of health information technology (HIT) -such as electronic health record systems -in improving the Triple Aim of cost, quality and outcomes, concluded that there is a link between the use of such technology and the use of evidence-based guidelines for patient care, enhanced surveillance and monitoring of patient conditions, and decreased medication errors (Chaudhry et al., 2006). This study suggests that perhaps this relationship occurs because HIT can facilitate information sharing, which is a key element of care coordination. Since care coordination is thought to play a key role in the prevention of errors and medical complications, it may be that effective information sharing is the pathway by which HIT helps reduce medication errors, as indicated in this prior review. No prior studies had inductively examined this particular phenomenon in the interprofessional care literature by making the link between these different variables (Y. Jadotte et al., 2016).
The theme of care delivery, which consists of the approaches chosen by health professionals and healthcare systems to implement patient care interventions, had not yet appeared as an important element in this pathway in the interprofessional care literature (Y. Jadotte et al., 2016). A key insight here is the idea that IPCP may not be successful without care delivery models that are supportive of its function. Many well-known care delivery models exist, including the patient centered medical model, nurse managed health clinic, and accountable care organization (American Nurses Asso-ciation, 2010). However prior studies on the phenomenon of interest had not yet identified that the care delivery model itself may determine how effective IPCP can be in fulfilling the promise of better health and system outcomes (Y. Jadotte et al., 2016).
This study suggests that in the context of IPCP, patient care should be conceptualized as the combination of continuity of care, patient adherence, and institutional policies. No prior studies had identified which specific aspects of patient care can theoretically be affected by IPCP (Y. Jadotte et al., 2016). A prior systematic review on the definition of continuity of care concluded that "[it] is achieved by bridging discrete elements in the care pathway -whether different episodes, interventions by different providers, or changes in illness status -as well as by supporting aspects that endure intrinsically over time, such as patients' values, sustained relationships, and care plans. Processes designed to improve continuity -for example, care pathways and case management -do not themselves equate to continuity. For continuity to exist, care must be experienced as connected and coherent" (Haggerty et al., 2003). This study suggests that the processes designed to improve continuity of care, as mentioned above, can be thought of as care coordination. For example, what is case management but the coordination of information sharing between health professionals, patients, families and healthcare systems? As such, this study advances the field's knowledge by identifying specifically what continuity of care should consist of with regards to IPCP.
The concept of "being part of the team" as an element of care coordination is an important contribution of this study that is supported by a prior systematic review of the literature which concluded that "involving patients has contributed to changes in the provision of services across a range of different settings. An evidence base for the effects on use of services, quality of care, satisfaction, or health of patients does not exist" (Crawford et al., 2002). This review is congruent with two thematic findings of this study: having all stakeholders be part of the team, including patients, is a fundamental underlying explanatory element in care coordination; and care coordination primarily works by improving the continuity of care. However, this insight had not yet been documented in the interprofessional care literature as a fundamental output of IPCP (Y. Jadotte et al., 2016). Furthermore, this study demonstrated inductively that social capital and its cognitive facets of trust, sharing, and reciprocity (Islam et al., 2006) may represent the underlying theoretical constructs that explain why IPCP works, forming the latent mechanism that facilitates IPCP's impact on patient health and system outcomes. Specifically, this phenomenon happens in 4 steps. Interprofessional interventions build trust among healthcare teams, by helping all those involved in them to look beyond traditional hierarchies and perspectives and overcome the entrenched attitudes that have for so long stalled the development of truly effective teams. Once trust has been built within a team of health professionals and the patient, everyone is part of the team and works diligently to improve the work environment. Consequently, information sharing, decisionmaking, problem-solving, and care delivery modeling become more streamlined. As this process continues, team members over time learn to work well together and become more likely to hold each other accountable for ensuring optimal patient care: this means providing care that is continuous (that is, non-fragmented and holistic), that patients adhere to, and that is well supported by institutional policies. Finally, when patient care takes on this optimized form where all stakeholders are receptive and reciprocating, improvements in patient health and system outcomes can take place in both the objective and subjective realms. Once change has been effected in these two realms of measurement, there is further motivation for health professionals to continue working collaboratively, and the cycle repeats itself.
Lastly, this study also identified the influence of disparities in health and healthcare as significant barriers to the achievement of better health outcomes via IPCP. Although there is ample literature on the topic of social capital and its link to health disparities from outside the realm of interprofessional care (Berkman & Glass, 2000;Bourdieu, 1986;Islam et al., 2006;Macinko & Starfield, 2001), as well as editorials (Soubhi, 2010) and deductive empirical studies (Godley et al., 2011;Godley & Russell-Mayhew, 2010) that measured social capital in the context of interprofessional care, no prior inductive research studies specific to this phenomenon and to the context of the lives of disadvantaged populations in the US had been identified (Y. Jadotte et al., 2016).
There are a number of study limitations. Similar to all primary qualitative research, this theoretical model is intrinsically subjective, at risk of sampling error given the small sample size, and subject to poor external validity given that it used a convenience sample. However, the literature review shows that the study's findings are nevertheless consistent with a systematic review of the qualitative literature on this phenomenon. In addition, given that the sample originated solely from the urban setting, it is possible that the study's proposed mechanistic pathway only applies to this setting and is not transferable to the broader healthcare environment of the US. However, this seems unlikely since the constant comparative analytic approach yielded clear distinctions between the general mechanisms of action of IPCP and the impact of disparities on these mechanisms. Importantly, the profession of social work was underrepresented in the study sample: only one social worker was interviewed. This is because the protocol for this study modeled the sample after the IPEC 2011 expert panel report, which was produced by a committee of health professional organizations in the US that at the time did not include the profession of social work (Interprofessional Education Collaborative Expert Panel, 2011), even though now this collaborative recognizes and includes social work in all of its deliberations (Interprofessional Education Collaborative, 2016).

Conclusion
Interprofessional collaboration, including effective communication, teamwork, appreciation for different specialties and diversity in professional knowledge and skills, occurs when all health stakeholders learn to trust each other and therefore serves as a potentially effective tool to build cognitive social capital (Islam et al., 2006). The task of building trust requires overcoming entrenched attitudes and looking beyond traditional hierarchies in healthcare. Once IPCP is achieved, there are improvements care coordination, which consist of the facilitated sharing of information, decision-making and care delivery modeling, all of which become possible because with successful interprofessional collaboration, everyone is part of the team (including the patient and family) and the work environment is improved. Subsequently, patient care is optimized, including better continuity of care, adherence to care, and supportive institutional policies, all of which are based on an enhanced level of reciprocity, which itself depends on all team members working together over time and holding each other accountable for patient care. Only when patient care is optimized in this way does it become possible to truly effect change in objective outcomes (i.e. patient health outcomes and system efficiency) and subjective outcomes (i.e. improved employee and patient satisfaction), all of which are the common goals of health stakeholders. Numerous system disparities (i.e. facilityprovider disconnect, differential system demeanor toward stakeholders, healthcare misconduct, differential support systems for teams, information technology/ electronic health record limitations, sociopolitical disparities that can lead to poor care delivery or care fragmentation, healthcare facility differences) and population disparities (i.e. differential expectations, empathy gap, lack of trust, low socioeconomic status, complex problems) impede the IPCP pathways to better health outcomes for disadvantaged populations.
Thus, while it is important to ensure that health professionals become increasingly more interprofessionally competent (Interprofessional Education Collaborative, 2016;Interprofessional Education Collaborative Expert Panel, 2011), this study suggests that there is also a substantial need to provide additional support for those who work with disadvantaged patient populations, if they are to accomplish their task of providing holistic, non-fragmented, and patient-centered team-based care. By making a research-grounded link between interprofessionalism and the perennial problem of health and healthcare disparities that is endemic in the US, this study proposes that there is a significant need for further study on the impact of disparities on the function of healthcare teams. This includes the impact of disparities on the patients and families as integral members of the team, as well as on the health professionals who are tasked with the role of healing them in a context that is unsupportive and more cumbersome than it ought to be.
This study adds much to the interprofessional care literature by answering the calls made by many researchers (B. Brandt, Lutfiyya, King, & Chioreso, 2014;B. F. Brandt, 2014;Reeves et al., 2013;Reeves et al., 2008;Zwarenstein et al., 2009) and organizations (Institute of Medicine, 2015) to specifically study the role of context in the pathway between IPL and health and system outcomes. In particular, it expands the model developed by the National Academy of Medicine (NAM) on measuring the impact of IPE on collaborative practice and patient outcomes, by deciphering the steps in the pathway between IPL and patient health and system outcomes. In essence, this study found that the IPL outcomes are linked to better care coordination, which leads to improved patient care, and results in optimization of health and system outcomes. This is perhaps the most fundamental contribution of this study to the current body of knowledge on this phenomenon. The study also suggests that the enabling or interfering factors that may influence learning outcomes as well as health and system outcomes as stated in the NAM's model may consist of the systems and population disparities identified in this study. For example, the NAM categories of financing policy and workforce policy can be related to the themes of institutional policies and sociopolitical disparities in this study. More importantly, however, this study elaborates clear mechanisms via which these types of factors influence the phenomenon of interest, which had not yet been done in the literature.
A number of areas of research remain to be investigated. Additional qualitative studies are warranted, especially those that explore the perspectives of patients. Studies using less subjective methods are also needed to truly study this phenomenon, for example using nonparticipant observations documenting the interactive behaviors of health professionals and patients. This will minimize the Hawthorne effect and allow more valid measurements of how interpersonal interactions grounded in IPCP truly impact care coordination, patient care, and population health and system outcomes. Studies with comparator groups to which health professionals are randomized are also desirable.
In addition, more inductive and deductive research is needed that verifies whether social capital is indeed the underlying theoretical mechanism of action in IPCP and examines the structural components of social capital, including the types of social networks that IPCP helps build and the types of social ties (i.e. strong or weak) by which it connects all health stakeholders. Some aspects of this phenomenon have already been investigated in the published literature. For example, one study looked at the social networks of patients and how they impact health outcomes (Christakis & Fowler, 2009), while others have examined how the social networks of health professionals facilitate collaborative work (Godley et al., 2011;Godley & Russell-Mayhew, 2010). Although this study has begun the work of bridging these two distinct areas of investigation, it re-mains unclear whether the social capital of health professionals, including the types of social networks and social ties, influences patient health outcomes. This study proposes one theoretical pathway by which that critical phenomenon could take place.
By identifying explicit measurable variables in this pathway, this study empirically clarifies what prior studies had proposed using consensus methods (Institute of Medicine, 2015), and it provides a useful evidence base to guide practice and future research. For example, because social capital may be the canonical theory underlying IPCP, health professionals should now be more cognizant that their perceptions of IPCP itself could be a key element in optimizing population health and system outcomes and that their social ties with colleagues and patients may play a key role in the achievement of these desired outcomes. It also renews the call for health professionals and policy makers to contend with the abysmal effects of health and healthcare disparities as fundamental barriers to the optimization of health and system outcomes via interprofessional collaboration.

Themes/Findings
Key Excerpts Building trust "I think it's changing. I think it's changing because a lot of other professions are coming to the fields. Now suddenly a lot of nurse practitioners are around, physician assistants, they're also taking care of patients, the primary care providers. I think now it's changing, but it will take a lot of time until it will be real professional collaboration. We don't have trust in each other. This is the major barrier." Overcoming entrenched attitudes "And I just feel like there's a lot of kind of entrenched attitudes about how things are done that are really difficult to overcome. And I don't know, it might take like a whole generation before some of these folks have been doing this for so long are finally kind of out of the system. If they're really not willing to change, you know, that might just be for the best. So hopefully the people we're training now are developing more of the skills." Looking beyond traditional hierarchies "I think what the bigger challenge is what I call the social dynamic of teams. If you look at the history of health care in this country, it was a very doctor centric, doctor driven model. The doctor was God, walked on water, and everybody bowed down and did what he or she said. And most often it was a he, and everyone worked to serve the doctor. We are now evolving to a place where everyone should be working to serve the patient." Interprofessionalism "The question for me as we're having this discussion is what exactly is collaboration. Is collaboration more so everybody working together in harmony? Is collaboration everybody doing their part when their number is called? Or is collaboration something different altogether? And I think that's what's not been quite defined in health care right now, exactly what cooperation is." Appreciation for different specialties "God, it's like you get more done when you acknowledge what each profession brings to the table."

Diversity in professional skills
"Yeah I think it's pretty clear, you know, that it's using various talents of the professions around you."

Communication
"Sometimes those answers may not be answers that we agree on, and how do we develop communication strategies that will allow us to do that. And I think at the same time, it really is also coming to some common understandings around how we can communicate." Teamwork "I consider it team work, you know if you're taking care of let's say patients, be it a nurse or a doctor, one person can't do it so you need to kinda collaborate as a team." Interrelatedness of interprofessional competencies "I actually reject the premise of the question to be honest with you because I think that part of the reason that these competencies were developed in the way they are is that communication without valuing teamwork and without understanding the need for respect between different professionals doesn't work. That's the problem. Health professions education 101 is all about communication. Clearly, the problem has been that it's all about communication without layering on these other very important competencies that inform what that communication should look like. So I actually don't think about it that way at all. I don't think about communication standing alone ever." Interprofessional interventions "So obviously I'm involved in this leap of faith that we're doing that says if we train people to work together, then they will know how to work together better when they get out and that will make them more skilled as part of teams." Tackling disparities "I mean I can see how interprofessional collaboration should be able to help address disparities." Facilitating sharing "Yes. That's why I help them ask the right questions. That's why we help them here. That's why we do that. That's why we have coordination of care letters. That's why we're permitted to call the medical doctor. That's why we teach them or refer them how to go and get dental care. That's why when the doctors work with them here, they'll refer them to a sleep disorder clinic or a nutritionist or audiologist or whatever they may need. The psychiatrists here do that. And then as a therapist, I help and work with them to ask the right questions because you can't get the right services if you don't ask the right questions." Being part of the team "They're very good at it up there. They really know how to do it. They're very much into everybody is a part of the team." Improving the work environment "And plus I think it just makes life easier. If there's something that I'm doing that I don't need the nurse or the medical assistant to do, then it frees their time to do something else that would be important because there's always something more to do. It's never ending."