Viewing interprofessional collaboration through the lens of networked ecological systems theory

ABSTRACT

Interprofessional collaboration (IPC) is key to ensuring safe quality care for patients. However, IPC intervention outcomes are variable, leading to calls for systems theories to understand complex interactions in healthcare. Using networked ecological systems theory (NEST), we aimed to uncover facilitators and barriers impacting the interactions between nurses and physicians in a specialty healthcare center. A qualitative study involving 55 non-participant observations and 17 individual semi-structured interviews was conducted at the National Neuroscience Institute of Singapore from April 2019 to March 2021. Template analysis was used to analyze the data. The most important IPC facilitators were exosystemic institutional support and physicians’ willingness to engage in IPC in the microsystems that together enabled the establishment of disease-based outpatient programs fostering patient-centered interactions among different healthcare professionals (HCP). We also found that patient-, disease-, and systems-related knowledge played an important role in facilitating IPC. Macrosystemic entrenchments such as intraprofessional composition of ward rounds emerged as a significant barrier. However, microsystemic efforts such as chat groups connecting all HCP involved in the care of the patients in the wards have fostered IPC. Although still preliminary, these findings suggest NEST can be useful to inform systematic interventions to improve IPC.

KEYWORDS:

• Interprofessional collaboration
• nurse-physician relations
• systems theory

Introduction

Patient safety (Institute of Medicine, 1999) and care outcomes (Zwarenstein et al., 2005) are predicated on the collaboration among different healthcare professionals (HCP). Researchers have employed many different lenses to examine facilitators and barriers impacting interprofessional collaboration (IPC). These lenses range from the organizational (Xyrichis & Lowton, 2008), to medico-legal (Brown et al., 2011), to financial (Lelubre et al., 2018), to sociological (Baker et al., 2011).

However, IPC challenges remain in various contexts around the world (Brown et al., 2011; Franz et al., 2020; Thomson et al., 2015). In particular, tensions between nurses and physicians have been well documented and transcend geographic locations (Tang et al., 2018; Weller et al., 2011). Physician resistance is expected as medical doctors have historically been dominant in healthcare (Hall, 2005). Physicians continue to prefer conversing with other physicians or patients rather than with nurses during ward rounds (Stickrath et al., 2013). In Singapore, even when physicians were pro-IPC, other factors hindered collaboration. For instance, nurses were too bogged down by routine tasks and were unable to attend rounds. It has been suggested that healthcare leaders institute policies to provide protected time for nurses to attend rounds (Chew et al., 2019). However, this solution may not fully address the problem because a survey showed that nurses in Singapore felt fearful about voicing their opinions, or perceived their views as not being heard because of differing power relationships (Tang et al., 2018). Thus, giving nurses protected time to round with physicians may not lead to IPC. Studies have shown that implementing changes at only one level were generally ineffective in bringing about IPC. For instance, organizationally creating pro-IPC protocols alone could not change older relational non-collaborative practices (Sena & Liani, 2020).

These findings point to the complex and dynamic nature of interactions in healthcare (Braithwaite et al., 2018). Organizational and systems theories (OST) focusing on complex and dynamic relations between multiple players in large systems were deemed suitable for studying IPC (O’Leary & Boland, 2020; Suter et al., 2013). We sought to operationalize Networked Ecological Systems Theory (Neal & Neal, 2013) to identify and examine barriers and facilitators impacting IPC at a national and regional specialty center for the management and care of patients with neurological diseases in Singapore.

Theoretical framework: Networked Ecological Systems Theory (NEST)

An early attempt to examine collaboration from a multifaceted perspective was the Interprofessional Education for Collaborative Patient-centered Practice, which sought to connect interprofessional education (IPE) determinants and processes at micro, meso, and macro levels to IPC (D’Amour & Oandasan, 2005). However, that influential paper did not identify the theory from which these concepts were drawn. In recent years, the Ecological Systems Theory (EST; Bronfenbrenner, 1979, 1986) has gained traction for application within health professions education (HPE). EST was used to evaluate a longitudinal online IPE program, demonstrating that students progressing through their training became increasingly capable of considering factors impacting healthcare practice from the individual (micro) to group (meso), and eventually to higher organizational (exo), and cultural (macro) levels (Bluteau et al., 2017). Hamwey et al. (2019) also proposed using EST to help clinician-educators study the experiences of underperforming HPE learners.

EST can be used to examine direct and indirect influences in a comprehensive and integrated way (Foo & Goy, 2021). Researchers can use it to study IPC by examining factors ranging from the individual to the organizational and cultural levels (Supplementary Material A).

Despite EST’s usefulness, its drawback is that the original representation of systems within concentric circles obscures the important relationships between each system (Neal & Neal, 2013). To address this limitation, we drew on Neal’s and Neal’s (2013) networked EST (NEST) to inform our study’s theoretical framework.

NEST (Neal & Neal, 2013) lends greater theoretical clarity (Ajjawi et al., 2017) as their conceptualization allows for the exploration of social interactional patterns found in the different contextual systems. The clarity is afforded by Neal’s and Neal’s (2013) reformulation of key EST concepts (Supplementary Material B). Figure 1 – adapted from Neal and Neal (2013) – is an example of how NEST could be used to study IPCs.

Figure 1. Hypothetical example illustrating a networked healthcare ecology [Adapted from Neal, J.W., & Neal, Z.P. (2013). Nested or networked? Future directions for ecological systems theory. Social Development, 22(4), 722–737. https://doi.org/10.1111/sode.12018].

1. The solid lines represent intra- and interprofessional interactions; the dotted lines represent system boundaries. 2. Dotted lines indicate that system boundaries are permeable. 3. The different shapes of ovals and rectangles used to represent different ecological systems do not have any special significance. They were chosen merely for esthetic reasons to keep the diagram compact.

Figure 1 demonstrates how NEST offers a framework to examine the multiple factors that directly or indirectly influence the relationships between individuals (Coleman, 1988; Granovetter, 1973), as well as between individuals and organizations (Neal & Neal, 2013). As can be seen, Nurse A interacts directly with physician (P) in the outpatient clinic and supervisor of nursing department (D) in the nursing department microsystems. However, D and P’s interactions in the mesosystem, which impacts Nurse A’s work, are also influenced by D’s relationship with the institution medical board (MB) situated in the exosystem. In turn, MB is influenced by the healthcare supervisory organization and the Ministry of Health in the exosystem. These interactions are also indirectly influenced by macrosystemic values and practices pertaining to particular professions, as well as chronosystemic patient safety concerns and the need to provide cost-efficient and effective healthcare.

As depicted in Figure 1, healthcare is a complex network of relationships involving numerous professions. NEST is thus useful for examining these relationships.

Study aims

We aimed to examine nurses’ and physicians’ experiences and perceptions of IPC barriers and facilitators at our study institution from a systems perspective. Specifically, we sought to use NEST to understand how direct/indirect influences of multi-level factors within healthcare environments influenced IPC.

Methods

We used a qualitative design combining observations and interviews to examine individuals’ behaviors, experiences, and perceptions in an in-depth and nuanced manner (Patton, 2015).

Research setting

The National Neuroscience Institute (NNI) is the national and regional specialty center for managing and caring for patients with neurological diseases in Singapore. NNI operates out of two main campuses/hospitals [Tan Tock Seng Hospital and Singapore General Hospital] and four partner hospitals. Specialist physicians, nurses, and allied health professionals (AHP) namely radiographers, psychologists, and medical technologists from the Departments of Neurology, Neurosurgery, Neuroradiology, Neurodiagnostic and Neuromuscular Laboratories integrate their expertise and work together to care for patients. NNI staff also work collaboratively with HCP from the main campuses/hospitals and partner hospitals to deliver care to patients with neurological diseases in both inpatient and outpatient settings in these hospitals.

Inpatient care is delivered by multidisciplinary teams of physicians, nurses, and AHP. These teams consist of a mixture of fixed members (physicians of varying seniority; four to six members) and fluid members (nurses and AHP of the wards). Outpatient care is mainly delivered by fixed teams of nurses and physicians (two to four members).

Participants and sampling

We had planned to apply maximum variation sampling technique (Patton, 2015) and recruit HCP from different professions, ranks, and departments [Neurology (NL), Neurosurgery (NS), and Neuroradiology (NR)]. Participants from NL and NS responded before the recruitment was suspended due to the COVID-19 pandemic. Increased patient load, heightened hospital safety measures, and a national lockdown in Singapore placed non-patient care activities on hold between January 2020 and February 2021. Upon Singapore’s phased re-opening, we resumed recruitment in March 2021. However, cluster outbreaks in April 2021 precluded further participation by HCP. Given the continuing and foreseeable uncertainties in Singapore, the study team decided to end data collection and analyze data collected from NL and NS.

Data collection

Data collection began in April 2019 and ended in March 2021. An interview originally scheduled for January 2020 with a physician that was postponed was eventually completed in March 2021. As no new issues were raised in that last interview, and we had enough data to start using NEST to describe the direct/indirect influences of multi-level factors pertaining to IPC within our study institution, data sufficiency (Dey, 1999) was deemed to be reached.

All nurses and physicians at the study institution working in inpatient wards and outpatient clinics were invited by e-mail to participate voluntarily. A total of 55 HCP participated in the study. We conducted 52 hours of observations involving 55 HCP during inpatient ward rounds and outpatient clinic sessions. Our observational tool focused on HCP’s roles, activities, and interactions, communication, and collaboration, and teamwork (Gum et al., 2019). Observation field notes were recorded by hand. To avoid affecting regular nurse-physician interactions, we made our observations at a distance that enabled us to hear the participants’ speech, but not so close as to be intrusive or cause disruption. Subsequently, 17 individual semi-structured interviews (for interview guide, see Supplementary Material C) lasting between 40 minutes and an hour were conducted. Two of the interviewees were not observed.

A qualitative researcher (YYF) and a Research Assistant (RA) trained by YYF conducted all observations and semi-structured interviews. YYF has a PhD in education, and the RA is a graduate and former medical social worker at a public hospital in Singapore. The researchers had no prior relationships with the HCP observed and interviewed. All interviews were audio-recorded, transcribed verbatim, de-identified by the RA, and checked for accuracy by YYF.

The observations and interviews were arranged to accommodate the HCP’s busy work schedules. The order of data collection did not present significant differences in the analysis of observation field notes and interview transcripts.

Analysis

We used template analysis (King, 2012) to analyze our data. As our study was informed by NEST, we identified five a priori themes based on this framework and study aims (barriers and facilitators in the micro-, meso-, exo-, macro-, and chronosystem). However, being mindful of “listening to the data” (Rubin & Rubin, 2012), we also inductively coded a subset of the data called Subset-1. The selection of Subset-1 was based on the participants’ rank in terms of seniority (for example, Medical Officer, Registrar, Consultant) and how amenable or supportive they were toward IPC. The purpose for inductive coding was to surface any additional themes that were relevant but had not been included amongst the a priori themes (Supplementary Material D).

During this process, we reflexively contextualized our coding decisions to our prior experience and questioned our biases and assumptions. We deliberately sought each other’s alternative perspectives made possible by the diversity of our research team, which comprised physicians (KT, NCKT, JR, WSL), a health services research manager (XX), a healthcare researcher and pharmacist (EL), a senior healthcare research associate (QC), and a qualitative researcher (YYF). Following discussions, we created our initial coding template.

YYF and QC then applied this initial template to another subset of our data (called Subset-2; selected based on profession). At this stage, we hierarchically ordered (King, 2012) the codes such as making “Primacy of institutional leadership and physicians in IPC adoption” an integrative theme by subsuming the a priori themes under it (King, 2012). These revisions were applied to Subset-1 to ascertain their applicability.

These changes were then shared with the team and accepted after review and discussion. YYF and QC further applied the revised template to the remaining data, and as no further revisions were made, it became the final template. The themes were then examined and interpreted.

Trustworthiness

To ensure trustworthiness, we first coded the data independently, then critically compared our codes, and reflexively discussed revisions to the analysis template (King, 2012). Additionally, we conducted member checking and maintained an audit trail (King, 2012) consisting of research memos and meeting minutes.

Ethical considerations

Our study was approved by the SingHealth Centralized Institutional Review Board (CIRB Ref No: 2019/2188).

Results

Our analysis produced two interrelated main themes: (a) primacy of institutional leadership and physicians in IPC adoption, and (b) knowledge facilitates collaboration. Subjects of observation comprised 55 HCP (years of experience ranging from less than 1 year to more than 20 years) representing nurses (24%), AHP (7%), physicians (58%), and other professionals (11%). Interview participants comprised 17 HCP (years of experience ranging from 2 years to more than 20 years) representing nurses (35%) and physicians (65%).

Primacy of institutional leadership and physicians in IPC adoption

Working relationships between nurses and physicians at our study institution were cordial. However, the depth and breadth of nurse-physician interactions depended on two factors: exosystemic policies developed by institutional leaders directing work arrangements, and pro-IPC physicians in the microsystem.

Impact of institutional leaders and physicians on IPC in the outpatient setting

Some physicians at our study institution co-run disease-based outpatient clinics with Nurse Clinicians (NC) or Advanced Practice Nurses (APN). At NNI, NC are registered nurses providing direct care to patients and their families. APN are registered nurses with expertise for extended practice, and an APN license is awarded upon the completion of a 2-year, full-time Master’s Degree and a 1-year internship supervised by a physician.

The disease-based programs for dementia and stroke patients were jointly created by pro-IPC physicians in the microsystem and supported by institutional leaders in the exosystem. Although these programs were less structured than clinical departments, which were the primary microsystems that gave different HCP their professional identities (e.g., nurse, physician), they were nonetheless settings created with the support of institutional leaders and pro-IPC physicians. As such, disease-based programs formed a secondary microsystem with their own boundaries where HCP from different disciplines developed a shared identity within the program (i.e., stroke program nurse, stroke program director) as shown in Figure 2.

Figure 2. Disease-based outpatient clinics as clinical microsystems.

1. The solid lines represent intra- and interprofessional interactions; the dotted lines represent system boundaries. 2. Dotted lines indicate that system boundaries are permeable. 3. The different shapes of ovals and rectangles used to represent different ecological systems do not have any special significance. They were chosen merely for esthetic reasons to keep the diagram compact.

In this clinical microsystem, physicians’ support was critical because collaboration could be a logistically complex and time-consuming enterprise. In Singapore, APN who co-run outpatient clinics with physicians do not have prescribing rights and can only review stable patients if the patient has specific pre-defined conditions for which the APN are accredited. For example, a stroke APN would be able to review a stroke patient but not a dementia patient. The partnering physician would thus need to check the APN’s review, confirm her/his treatment plan, and prescribe medications

We see our very own sub-specialized cases who are very stable … the doctors need not see (them) … But the doctors will do the prescription … (the doctors will) read through our history, and if there’s a bit “not right” information or they want to find out more, they can ask. (N1-APN1)

For the partnering physicians, providing a “confirmatory” consultation meant switching between the APN’s patients and their own:

I do cross over to see N1’s (MD1’s partnering APN) cases as much as possible, but sometimes if I’m stuck, N1 might see a case and then send (the patient) home without me physically seeing the patient. [MD1-Senior Consultant (SC)]

In instances where the partnering physician could not see the APN’s patients, s/he might have to carve out time to discuss those cases with the APN later: “To catch up, at the end of our clinic, we will do a very quick update on what the plans are and what needs to be followed up on.” (MD1-SC)

The collaboration between nurses and physicians in the outpatient setting thus can be resource-intensive. As MD19, a partnering physician of an APN said, “actually it is faster for me to see all the patients myself, but the extra effort is worthwhile (as long as patients were offered better care).” This was because such collaborations enabled nurses to better cater to patients’ needs, especially when busy physicians could not:

Sometimes the consultant is very busy, they don’t have time to talk to the patient. When we see the patient, we have more time to talk to the patient and (do) some education … help them comply with medication and manage their condition better. (N8-NC)

Patients can call us directly when the doctors are on leave … if (there’s) anything that’s very important, I can always escalate to [the doctor], or escalate to any doctors available, so patients feel more reassured. (N1-APN)

Patient safety was another reason why physicians perceived collaboration to be critical. As noted by a physician participant, caring for patients is a complex endeavor. Thus, it was important to have “multiple perspectives (from nurses to help) to prevent (medical) errors” (MD24-SC).

Observations of microsystemic interactions in the outpatient setting indicated congruence between the views expressed during interviews and actual practice in the clinics:

Physical layout: N1 and MD1 co-run a disease-based outpatient clinic in two adjoining rooms connected by a door.

After seeing patient A, MD1 went over to N1’s room to see patient B. N1, who had taken patient B’s history, gave MD1 an update and proposed some recommendations for upcoming treatment plans, some of which MD1 endorsed.

MD1 addressed patient B’s concerns about the treatment plans and cost of treatment. Here, N1 joined in, actively demonstrating knowledge about the hospital system, insurance cover, and cost of treatment, and offered advice to help reach a conclusion. (Observation notes, 18 Apr 2019)

Impact of institutional leaders and physicians on IPC in the inpatient setting

Compared to the disease-based outpatient clinics that served as a pro-IPC clinical microsystem, the inpatient setting had a different organizational structure. As a result, interactions between nurses and physicians in this space remained episodic, and their behaviors were guided by their primary professional identities rather than a shared identity around the patients they were caring for collectively. In reflecting on a specific instance of an unsuccessful interprofessional encounter, a nurse participant highlighted the importance of disease-based program in reorienting healthcare professionals away from their primary identities and toward the patients they cared for:

If you’re already in that particular program, you’ll fight for that program patient. But what we’re having now is … just a general patient. To me if you’re in a program, you own a certain group, you’re responsible for their growth. (N10-Senior NC)

Inpatient care interactions were observed to be primarily intraprofessional in composition where medical teams conducted ward rounds each morning in a hierarchical manner. Macrosystemic entrenchments were dominant where consultants (specialist physicians) made important decisions regarding patient care that Medical Officers (MOs; junior physicians) would implement under the Registrars’ (specialists in training) supervision. This work arrangement was designed for “efficiency” (MD17-Reg; MD23-MO).

The implications of such workflows were that Consultants and Registrars in general interacted with ward nurses far less than the MOs. These self-reported workflows were confirmed by observations of morning ward rounds that noted the frequency of intraprofessional interactions within medical teams to be high. In contrast, interprofessional interactions between nurses and physicians were “infrequent” (Observation notes: 24 Apr, 26 Jun, 15 Jul, 21 Jul, 14 Aug, 18 Sep, 26 Sep 2019). The reported exception was senior physicians’ interactions with the study institutions’ APN (see “Disease-related knowledge”).

Previously, it was assumed that frequent contact among HCP of different professions would foster IPC, but for MOs who were the “first point of contact” with nurses, this sometimes created “conflict(s)”:

When you are on-call … you receive … “x” number of calls, mostly from nurses asking for input … documentation … intervention. And all of these requests are made blind to the basic principles of an on-call doctor being there to respond to emergencies, principally. These requests may not necessarily be strictly medical. The doctor is overburdened … and the doctor has to find a tactful way of saying, “No”. (MD10-Reg)

Implicit in the above account was a reference to on-call physicians’ heavy workload, an effect of the indirect influence of exosystemic staffing policies. However, it should be noted that physicians’ response to overwork did not always result in conflict but varied according to their IPC-orientation. A physician participant shared how s/he sought to facilitate the ease and speed of communication with ward nurses through technological affordances:

I created a group (chat) because it’s easier for them … Then they don’t have [to] go through the (official communication channels which) takes longer. Also, by being in the group chat, everyone … can help to take note (of patients’ health status)’. (MD18-MO)

The virtual chat group is an important ground-up initiative to forge IPC. It has both symbolic and actual functions. On the symbolic level, the chat group amounted to a virtual clinical microsystem akin to the disease-based programs. However, unlike those programs, chat groups were not institutionalized by leaders in the exosystem and were therefore less likely to develop into a full-fledged clinical microsystem to foster a shared identity among HCP practicing together. On the practical level, as everyone in a chat group was a member with equal status to express themselves regardless of their professional titles, this communication platform effectively downplayed professional boundaries and hierarchies between nurses and physicians, an important step to engendering IPC.

In this section, we demonstrated how NEST could be used to make visible systemic dimensions impacting IPC. Based on our analysis, encouraging different professions to work together was necessary but not sufficient to engender IPC. Systematic arrangements on multiple levels – which our study identified as support from institutional leaders and physicians – were needed to bring about or strengthen IPC. In the following section, we discuss how the quality of interprofessional encounters also mattered greatly, and knowledge seemed an essential determinant of that quality.

Knowledge facilitates collaboration

Knowledge served as an IPC facilitator in the experience and perceptions of many of our participants. Here, knowledge was conceptualized as comprising three dimensions: patient-related knowledge, disease-related knowledge, and systems-related knowledge pertaining to IPC.

Patient-related knowledge

This form of knowledge pertained to comprehensive insights about the unique circumstances of each patient that spanned the nursing, functional, and biopsychosocial domains. This patient-related knowledge appeared to foster trust, which engendered microsystemic nurse-physician collaboration:

When nurses give us their input, usually we take it quite seriously. They spend more time with each patient so what they say, we do trust. Let’s say, discharging patients, some patients may have financial difficulties, so they will know all these nuggets. [MD17- Reg]

Disease-related knowledge

Disease-related knowledge refers to integrated information and reasoning, which can be used to justify care inputs. It was perceived as being valuable as it helped to “convey the clear logic behind a course of action” (MD11-MO). The implications for nurse-physician interactions were instantiated by the following excerpt:

In the morning, I do doctors’ rounds. If I notice that something is not very right with the patient, I can just let the doctor know. Before I became an APN, I spoke up less … I learnt a lot from the (APN) course because it helps me to reason … Now I can justify my reason for doing certain things more confidently. (N4-APN)

N4’s excerpt shows how disease-related knowledge facilitated the voice of nursing staff in the ward round microsystem by giving them the means and confidence with which to share care inputs, and in so doing, deepened collaboration with physicians. The acquisition of disease-related knowledge depended on the exosystemic support of institutional leaders because, as mentioned previously, an APN license entailed the completion of a “2-year, full-time Master’s Degree” (N1-APN). Physicians’ support was also critical as a high degree of commitment was required to supervise APN training during the “one-year physician-supervised internship” (MD19-SC) following the Master’s Degree required for full registration. This again underscored the primacy of institutional leaders and physicians on IPC in this study setting.

Systems-related knowledge pertaining to IPC

Systems-related knowledge pertaining to IPC was defined as the awareness of the interplay between various factors from multiple systems. The corollary of this awareness was the appreciation for IPC:

Part of the push towards IPC is cost … because there’s quite good data showing that our Nurse Clinicians and our Advanced Practice Nurses can do just as good a job and at far lower cost (than doctors) … the other driver has been the discourse about patient safety, to prevent errors. And the third thing is improving routine care to better care. (MD24-SC)

The excerpt suggested that MD24’s awareness of the chronosystemic influences – namely the patient safety discourse and the need for cost-efficient, good quality care – shaped her/his pro-IPC-orientation. The opposite was equally true: the absence of systems-related knowledge pertaining to IPC can adversely impact microsystemic collaboration. N11 – a nurse clinician who cared for patients in several inpatient wards – shared how some ward nurses may not appreciate how stretched on-call physicians might be:

The doctor could be busy with other patients in other wards but as a ward nurse, my patient is my priority. You have to see my patient first. But the doctor is busy, maybe another patient is more sick, high risk. That’s why there’s actually a lot of miscommunication between the doctors and nurses. (N11-NC)

N11 also talked about the need for HCP to replace what s/he called “microculture” (that is a narrower outlook) with “macroculture” (that is, broader perspective). It is interesting to note N11’s use of the prefixes “micro” and “macro” before the word “culture.” This suggests that the conceptualization of interprofessional interactions in healthcare as being influenced by multi-level factors was congruent with HCP’s actual work experience.

Discussion

This paper demonstrates that NEST can serve as a framework to elucidate how systems in complex healthcare settings created IPC barriers and facilitators. Specifically, it described the mechanisms by which multi-level systems interacted to act as a facilitator to diminish barriers presented by another system in the complex healthcare setting. Our themes fit largely into our a priori framework: we identified IPC barriers and facilitators located in the micro-, exo-, macro-, and chronosystems. The mesosystem factors were seemingly absent because while interviewees readily spoke about their personal IPC perceptions and experience in general, they were reserved about discussing issues related to particular colleagues and supervisors. In the main, however, we have demonstrated how NEST can be used to examine IPC complexities in an interrelated manner, which is a strength of systems theories.

A noteworthy example is the impact of disease-based programs on IPC in the outpatient setting. These programs enable specific groups of nurses and physicians within NNI to work together to look after specific cohorts of patients. The patient-centered nature of these programs facilitated the development of a shared identity among HCP, fostering IPC. This is congruent with prior IPC literature, where shared identity was deemed essential for engendering collaborative practice (Reeves et al., 2010; Xyrichis et al., 2018).

Using NEST, we showed how the creation of disease-based programs required both the backing of institutional leaders and the concurrent support of pro-IPC physicians willing to navigate logistical hurdles to co-run clinics with APN. Collaborative physicians were influenced by their awareness of chronosystemic issues such as patient safety and the need to provide cost-efficient, good quality care for aging populations. As indicated in a recent scoping review, the inverse of chronosystemic awareness is the lack of understanding of the goals of IPC – a barrier to collaborative practice (McNaughton et al., 2021). Thus, for IPC to flourish, the goals of multiple systems within a healthcare setting need to be aligned. In light of this, NEST-informed IPE programs such as the one reported by Bluteau et al. (2017) may be useful in helping HCP develop an appreciation of IPC goals.

Like prior studies, we also found hierarchy to be a barrier. Hierarchy characterized by medical dominance (Chong et al., 2013; Hall, 2005) aptly described our study’s inpatient setting, where ward rounds were still largely intraprofessional in composition and led by physicians. Unlike the outpatient disease-based programs, there was no mechanism in the wards to break down this barrier that has been reported both in older and more recent literature (Baker et al., 2011; Etherington et al., 2021; Waggie & Arends, 2021). It appears that without the counter influence of pro-IPC forces in the exo- and microsystems, the barrier of hierarchy forged in the macrosystem through previously siloed training would continue to persist. Under such circumstances, HCP would stick to profession-specific roles and interact less frequently (Hatton et al., 2021), an impediment similarly noted in our findings.

Besides creating disease-based programs, IPC support in exo- and microsystem was also crucial in helping APN’s gain confidence in sharing their care plans by acquiring disease-specific knowledge through the pursuit of higher education. This knowledge gained was correlated with the APN’s confidence in speaking up (Hendricks et al., 2017; Pfaff et al., 2014). This correlation may be explained by the fact that knowledgeable nurses were perceived as being competent (Baggs & Schmitt, 1997) because they could transmit information precisely (Schmiedhofer et al., 2021), which concurred with our findings.

Interestingly, although studies from non-Asian settings emphasized the importance of role understanding and using language appropriate for communicating with other professions (Suter et al., 2009; Waggie & Arends, 2021), non-medical HCP in Singapore appeared to adopt an accommodating stance by acquiring medical language to communicate within interprofessional teams. The APN in our study seemed to readily receive support from pro-IPC physicians to supervise them in year-long internships to acquire disease-specific knowledge to communicate with their medical counterparts. This contrasts with studies set in North America (Baker et al., 2011; Reeves et al., 2015) and Britain (Xyrichis et al., 2017) where contestations over professional jurisdiction reportedly undermined IPC to varying degrees. NEST can thus be useful in examining how macrosystemic ethnic cultural factors may interact with other systems (namely at the meso, exo, and chrono levels) in a healthcare setting to effect different interprofessional interactions in the microsystem.

Strengths and limitations

Strengths of our study are: first, the novel use of NEST to surface important systemic dimensions affecting IPC. Although EST itself is not new, we believe that this is the first study to apply NEST in investigating IPC. Second, the combination of interviews and observations not only allows us to understand and examine participant perceptions and experiences in an in-depth and nuanced way, it also serves as a means to corroborate findings from the interviews with observed behaviors and team dynamics on the ground. Third, our investigation sampled both inpatient and outpatient settings with their respective fixed and/or fluid teams, which provided a good cross-sectional view of IPC (or lack thereof) in action.

Our study did not include AHP and that limited our understanding of how their inclusion might impact IPC in our study setting. However, the COVID-19 pandemic was a disruption we could not anticipate or overcome. A second limitation is selection bias. Many of the HCP who agreed to be interviewed seemed to embrace IPC. Their support for IPC likely induced them to participate in the interviews even though speaking to the researchers was time-consuming.

As observation was used as a data collection method, questions concerning participant reactivity may arise (Paradis & Sutkin, 2017). To avoid affecting regular nurse-physician interactions, observations were made at a distance to enable the participants’ speech to be heard clearly, but not so close as to be intrusive or cause disruption. While some HCP appeared self-conscious at the beginning, such behavior did not persist as each observation session lasted minimally 4 hours. Furthermore, the interview data allowed for a considerable degree of triangulation.

Suggestions for further research

NEST could serve as a comprehensive framework to help policymakers such as health ministry officials and hospital leaders identify direct or hidden IPC barriers and facilitators in their specific healthcare settings. This would enable them to formulate policies to address the multi-level factors that impact IPC. More research involving AHP is also needed to test the applicability of NEST beyond nursing and medicine. In addition, as communication technologies such as chat groups are increasingly used in healthcare teams, it may be useful to investigate how technological affordances impact IPC.

Conclusion

NEST can be used to unveil barriers and facilitators impacting how nurses and physicians perceive and experience collaborative practice. The identification of these factors could help policymakers devise and implement IPC plans effective for their particular settings.

Acknowledgments

We would like to thank Dr Tina Martimianakis (University of Toronto and The Wilson Centre) for her valuable comments on our draft; Ms Shihui Tang, the Research Assistant; and Ms Jeannie Lum, and Amy Cheng for their administrative support.

Disclosure statement

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

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website

Additional information

Funding

This work was supported by the Lee Foundation.

Notes on contributors

Yang Yann Foo

Yang Yann Foo is an Assistant Professor with the Duke-NUS Medical School.

Kevin Tan

Kevin Tan is a Senior Consultant Neurologist with the National Neuroscience Institute.

Jai Rao

Jai Rao is a Senior Consultant Neurosurgeon with the National Neuroscience Institute.

Wee Shiong Lim

Wee Shiong Lim is Senior Consultant Geriatrician with Tan Tock Seng Hospital.

Xiaohui Xin

Xiaohui Xin is a Senior Research Manager with the Singapore General Hospital.

Qianhui Cheng

Qianhui Cheng is a Senior Executive with the National Neuroscience Institute.

Elaine Lum

Elaine Lum is an Assistant Professor with the Duke-NUS Medical School.

Nigel CK Tan

Kevin Tan is a Senior Consultant Neurologist with the National Neuroscience Institute.

Nigel CK Tan is a Senior Consultant Neurologist with the National Neuroscience Institute.