Ethnography in Health Services Research: Oscillation Between Theory and Practice

Abstract

The well-known divergence between what policy and protocol look like on paper, and what happens in the actual practice of daily life remains a central challenge in health services provision and research. This disparity is usually referred to as the theory–practice gap and contributes to concerns that scientific evidence fails to make substantial impacts on the processes of service delivery. In this article, we present an argument for the inclusion of ethnographic methods in health services research and show that this approach enables researchers to address this divergence by working within it. We trace how ethnography, through generative processes of oscillation, can take us beyond lamenting the gap and capture the relational dynamics of people working together in complex systemic arrangements. By moving from example to methodological reflection, to principle of research, we demonstrate how the oscillation of ethnographic research between theory and practice can productively contribute to the field of health service research.

Introduction

The ethnographic origin story begins in 1914 with Bronislaw Malinowski and his detailed study of life in the Trobriand Islands: Consequent to the first World War, Malinowski was forced to stay in the Trobriand Islands much longer than planned. This extended temporal displacement ultimately led to the production of the famous monograph, Argonauts of the Western Pacific (Malinowski, 1922/2014). During his time in the field, Malinowski learned about the life of the Trobriand islanders by both observing and participating in the practices of everyday life, thereby laying the foundation for the methodology of participant observation that continues to characterize ethnographic work today. His experience proved a revelation in the possibilities for ethnographic knowledge production and initiated a paradigmatic shift from “armchair ethnology” (Stocking, 1994), which drew on secondhand empirical material from travelers and missionaries, to field research, in which researchers live and work in the field site, actively collecting data themselves (Clifford, 1983).

We begin this article with the story of Malinowski’s monograph because it is a foundational anthropological narrative and regarded as the origin of contemporary ethnographic research. It provides an important backdrop for our proposition that ethnography must be regarded as a method in its own right and that it draws its value primarily from long-term immersion within a research field. It is this extended temporal presence that enables a continuous oscillation between theory and practice and which characterizes a profoundly inductive and relational scientific endeavor. Although this historical anecdote may be familiar to many ethnographers, in this article, we also hope to address colleagues from other disciplinary fields, in particular health services research as well as clinical and implementation sciences. In writing for various audiences, we recognize there may be certain redundancies for one or the other, but our goal is to generate an interdisciplinary conversation in which the insights offered here might be useful not just to one cadre or another but rather to various disciplines in a gesture of exchange and collaboration.

In what follows, we begin with an overview of health service research, as this field integrates diverse approaches and research logics, and the authors are currently all working in this area. In our work, we have observed that there is continued uncertainty as to what kind of insights ethnographic material and studies may offer, and what can and cannot be expected of its findings. Drawing on this experience, we try to demonstrate how ethnographic findings can be connected to other forms of data and inquiry and to outline a more productive way of integrating findings from ethnographic research with other forms of evaluation and knowledge production. Essential to our proposal is that the integration of ethnographic methods must be planned from the beginning of a research study. In this way, the ethnographic process can direct and develop the kinds of questions informed by long-term investigations embedded in practice. These questions, in turn, shape processes of data collection and analysis and offer novel insights into the gap that often exists between theory and practice in health services research.

For many years, the inclusion of ethnographic methods in health services research and related research fields has enabled researchers to engage in an important methodological oscillation. This oscillation, between theory and practice, has helped address the well-known divergence between what policy and protocol look like on paper, and what happens in the actual practice of daily life and implementation (Hopper, 1991; McAteer et al., 2018; Metzl & Hansen, 2014). This disparity is often referred to as the theory–practice gap and contributes to concerns that scientific evidence fails to make substantial impacts on the processes of service delivery (Brazil et al., 2005; Campillo-Artero, 2012; Greenway et al., 2019). While the specifics of the term “theory–practice gap” are only loosely defined, and the underlying concepts somewhat controversial, we take up this idea because it is frequently used within health services research to index a meaningful frustration: the persistent distance between theoretical knowledge and its practical application, which yields seemingly negative consequences (e.g., Greenway et al., 2019). We want to reframe the orientation to this gap and to position ethnography as a means of maintaining movement between theoretical frames on one hand and practical application on the other. It is not a one-off solution or an easy fix, but rather a methodological orientation that is more responsive and dynamic in its attention to both theoretical frame and everyday practice (Boyer et al., 2015). By oscillating between these two dimensions of research, we argue ethnographic methods offer a means of using the theory–practice gap productively (e.g., Greenhalgh & Swinglehurst, 2011; Jack, 2006).

In what follows, four ethnographic examples will be presented, each demonstrating a dimension of how long-term ethnographic inquiries can provide nuanced insights in the context of health services research and related fields of studies and deriving from the work of the authors at the intersection of these two disciplines. We organize the cases to illustrate two methodological principles and four basic heuristics of ethnographic research. In our efforts to address scholars from both the fields of ethnography and health service research, we begin with a disciplinary introduction, contextualizing ethnography as an anthropological approach and positioning it in relation to mental health service research more specifically. The second part details are reflexive efforts in developing this case as well as our empirical and conceptual findings. In the discussion, we work these threads together, demonstrating what a reflexive oscillatory ethnographic thinking can look like and, at the same time, making the case that interdisciplinary efforts to date have too often failed to realize the value of full-fledged ethnographic processes, and explaining why this matters. Finally, we offer some concrete suggestions in support of collaborative work.

Moving Toward Disciplinary Considerations

What is Ethnography?

Ethnography refers to at least two different dimensions in cultural anthropological research. First, it designates fieldwork in a specific social setting: “The ethnographer participates in the daily routines of this setting, develops ongoing relations with the people in it and observes all the while what is going on” (van Maanen, 2011b, p. 1). This method, “participant observation,” is one of the main methodological instruments by which to do this work. Ethnographic research can also include other empirical methods, such as interviews, mental maps, focus groups, and qualitative network analysis. Second, ethnography describes the textual and/or other forms of representation used to document field research. The ethnographer keeps written field notes, jottings, and cinematic, photographic, or other digital or material accounts of the field (van Maanen, 2011b). Central to ethnographic research is the interconnected and dialectical realization of both of these dimensions. The ethnographer writes both during and after fieldwork, rewrites and annotates field notes, returns to the field, and, ultimately, after months to years of this process, derives case studies and theoretical accounts. This process generates an iterative epistemic stance that draws from the continuous movement between the theoretical frame and empirical data (Faubion & Marcus, 2009; van Maanen, 2011a).

By observing and participating in everyday activities, ethnographic researchers are able to generate a specific kind of knowledge deeply tethered to lived processes (Clifford, 1983). At the same time, they move along a spectrum between taking part and stepping back from the field, negotiating proximity and distance and the possible variations in this relationship over time (Emerson et al., 2011; Lindner, 1981; Sillitoe, 2012). The continuous oscillation of the researcher between “inner” and “outer” perspectives is the epistemic momentum of participant observation, allowing for events to be experienced immediately on one hand while also situated within the larger cultural context on the other. Through this ongoing process, the proximate details of the field experience can be abstracted and understood on a larger scale (Clifford, 1983, p. 127).

The ethnographic process is emergent and embedded in the field yet not divorced from theory. When we talk about theory in this article, we are referencing both the frameworks that researchers bring with them to their work, shaping sites of interrogation and lines of inquiry, and how, in ethnographic practice, theoretical orientations must be held open to changing processes (Biehl, 2013; Stewart, 2008). Theory in ethnographic research is not only a necessary instrument for demonstrating the validity and relevance of inquiry, or for compiling methodological tools that will best address the questions at hand, but rather throughout the course of the research, “theory” remains subject to revision and bound up in an iterative relationship with practice. Theory in this sense is neither deductive nor separate from practice; it does not predetermine the data that is encountered in the field. Rather, it offers a backdrop against which insights can be situated and compared, a process that renders the construction of further theory more open to surprise and necessary change than the reaffirmation of previously assumed truths.

Ethnography and Health Services Research

Health service research itself is a highly interdisciplinary field, relating and partly overlapping with other fields, such as implementation and clinical sciences (Flook & Sanazaro, 1993). In comparison, however, health services research is a relatively new field that developed through a combination of public health and social science perspectives with more clinically oriented approaches (Pfaff et al., 2017). Rather than being a well-defined entity, it is the resultant combination of various interrelated disciplines, approaches, and methods that contribute to both its specific value and internal tensions between different research logics, paradigms, and discursive backgrounds (Institute of Medicine 1979). Most commonly, multimethod and multistep approaches are used to examine how people achieve access to healthcare services, how they experience those services, and to assess the level of service effectiveness, efficiency, and quality provided (Marshall, 1985; Steinwachs, 1991). Both quantitative and qualitative methods are used to evaluate the structures, processes, and organization of healthcare services and their relation to contemporary health policy, financing systems, and medical technology (Pfaff et al., 2017).

Health service research may provide answers to a wide variety of research questions, including those ranging from normative, exploratory, descriptive, confirmatory, or explanatory in nature (Pfaff et al., 2017). In this context, ethnographic approaches are usually understood to offer a contextual understanding, evaluating practices according to situated forms of knowledge production. They are understood as strengthening the internal (connection between intervention and outcome) and external (understanding of the interrelation of context and outcome) validities as well as the translational impact of an intervention (Pfaff et al., 2017). Yet, to contribute in this way, it must be clear when ethnographic methods are used, how they are implemented, and to what ends. While work on health topics more broadly is not a new area of research for anthropologists (i.e., Kleinman, 1980; Martin, 2007; Rapp, 1999; Scheper-Hughes, 1993), integrated ethnographic approaches within health services research are still relatively few (Greenhalgh & Swinglehurst, 2011; Loblay et al., 2021; Savage, 2000). Even rarer are collaborations in which it becomes clear how ethnography can be helpful in this field, what kind of knowledge it specifically produces, and how these findings complement and enrich other forms of assessment. In short, the specific impact of ethnographic research in the context of health services research has gained too little attention.

One possibility for this incongruence is the tendency for ethnographic approaches to be used separately from other methodological frames. In the context of health service research, ethnographic methods are rarely integrated with other methods or parts of the study, leading to uncertainty regarding their compatibility, adaptability, or relevance (Hammersley, 1992). Another possibility is that when ethnographic methods are used in health service research, they are often limited in scope and time and seen as a subsidiary addition to the other methodologies. Whether due to tight budgets or project timelines, this structural condition impedes long-term participant observation and the results that derive from it (Cupit et al., 2018; Vindrola-Padros & Johnson, 2020).

These constraints are not inherently negative (Pink & Morgan, 2013), but they alter ethnographic processes in ways that offer imagined benefits at the risk of methodological integrity. For example, using short-term fieldwork or ethnographic interviews effectively is largely dependent on previous time spent in the field, meaning good short-term fieldwork is actually part of a much longer field presence and research history (Seligmann & Estes, 2020). Moreover, focused or short-term forms of ethnographic research are often designed to collect data-specific or predetermined questions (Bikker et al., 2017), and this undermines the inductive quality of long-term engagement and its essential ability to surface questions that were not obvious or available at the outset. These approaches actually reflect different research logics and impede triangulation, frequently relegating ethnography to a side project that is not recognized as useful in the production of generalizable findings (Carminati, 2018) or robust forms of evidence (van Wijngaarden et al., 2017).

We recognize that the inductive nature of ethnographic research often precludes easy integration of findings with other results, for example, from outcomes-focused or randomized controlled trial research, which serve as the basis for other studies (Morgan-Trimmer & Wood, 2016). These may be some of the reasons why ethnographic methods continue to inhabit an uneasy position in the field of health service research. In what follows, we outline four ethnographic examples to clarify the specific value of ethnographic analyses and surface the unique kinds of knowledge they can contribute to the field of health services research.

Moving Toward Application in Practice

How Did We Proceed?

Our reflections begin with an orientation to the field of health service research, as this field has a rich tradition of multidisciplinary and multidimensional approaches. Based on more than a decade of combined work at the intersection of anthropology and psychiatry, we present a theoretical and methodological account of the ways in which ethnographic engagement informs health services research with a particular focus on mental health. Building on the work of Annemarie Mol (2006), Jeannette Pols (2006), and Donald Schön (1983), we offer an interpretation of reflexive (and reflective) work that is specifically addressed to the reconfiguration of health services research as an iterative process and one that demands continuous negotiation of the objects of research and the positionality of the researcher in relation to these objects. We draw on these approaches because they advocate for a methodology that turns on insights learned in practice rather than on formulas derived in institutional settings. The largely unaccounted-for processes that characterize improvisation and responsiveness are thus foregrounded as essential methodological tools that yield valuable understandings. It is this “reflection-in-action” (Schön, 1983) that offers a constructive means for dealing with the uncertainty and instability that exist between theoretical frames and practical implementation.

We use our own ethnographic examples to simultaneously illustrate and explain aspects of ethnographic methodology that exceed single cases and are applicable to other fields of research and inquiry. While we draw primarily from work on psychiatry, we see ourselves as studying the so-called “chronic conditions”—which can be extended to many dimensions of health services research and practice. What is essential about “chronic conditions” is their multiplicity: People living with them are often managing diverse needs simultaneously. They may require psychiatric care and be struggling with homelessness, or they may be receiving health services but struggling with social isolation beyond their contact with treatment providers. They may be experiencing social hardship as a result of their condition, while this condition may be itself the consequence of social hardship. Their experience is always more than one thing (Mol, 2003).

The populations we work with are never one-dimensional and neither are the researchers themselves. We account for this relationality in our approach while still attaching it to a concrete object of study. Thus, the reflexive methodological work we outline is expressly situated. To make ethnographic work more practically available to health services research, we strive to outline a means of both situating health services as a concrete object of study and a process that must be considered in relation to context and change. What is crucial about this approach in our view is the necessity of time—the potential of ethnographic research to reveal novel insights relates to the long-term quality of the researchers’ presence in the field (e.g., Morse, 2016). This, in addition to facilitating the collection of detailed data from a variety of sources, allows situated reflexivity to take place and for the insights of ethnographic methods to be refined in relation to the complexity of everyday life.

Ethnographic Scenes and Insights

The following results—two methodological principles of ethnographic research and four heuristics along which to ground these principles in practice—demonstrate how ethnography can provide novel and necessary insights for health services research by attending to the everyday activities of health care. The notion of a “ methodological principle” here delineates basic epistemic prospects that are fundamental to ethnographic approaches, while the term “heuristic” describes practical strategies that enable us to understand the field, produce knowledge, and allow for decision-making during ethnographic research processes. While the methodological principles apply to ethnographic research broadly, the subsequent heuristics derive from and are central to our work at the intersection of anthropology and health services.

The two principles and the four heuristics are outlined first in theory and then illustrated by concrete examples from our fieldwork practice. With these three steps—methodological principle, heuristics, and example—we try to translate our ethnographic research into different degrees of specificity, moving from general principle to situated case. The principles and heuristics identified here are not exhaustive but rather identify unique aspects of ethnographic research that may be especially helpful in working across and through the theory–practice gap in the field of health services research. They apply beyond the single examples detailed here and can translate meaningfully into other contexts. In presenting them this way, we demonstrate this oscillation from theory to practice, and back again, finding synthesis in the process of reflection (Table 1).

Table 1.

An Overview of the Two Methodological Principles and Four Heuristics.

Methodological principle	1. Inductive character	2. Relational perspective
Heuristics	1. (a) Attending to informal epistemologies and tacit knowing 1. (b): Responding to surprise and the unanticipated	2. (a): Understanding the situatedness of local worlds 2. (b): Studying systemic embeddedness

Note. For a nuanced understanding, the reading of the cases and their interpretations is advised.

Methodological Principle 1: The Inductive Character of Ethnography

Ethnography is inherently based on an inductive research paradigm, meaning that more general conclusions are eventually drawn out of the specificity of ethnographic encounters. But to do this requires the accumulation of cases and particulars, which, at first, may seem daunting and disorienting. Over time, the inductive process allows for the grouping of cases, reflection on themes, and refinement of questions but all of this without the imposition of systematic, neat, or categorical conclusions (Spindler & Spindler, 1990). Participant observation focuses on how processes happen, why they are happening, and where they are happening. Researchers go into the field and attempt to take part in the everyday life of the actors on site. Thus, they are confronted with their own assumptions, their own experiences, and necessarily have to challenge any predetermined research questions, concepts, and ideas. It simply cannot be predicted exactly what will happen during a field stay before one has been there, meaning ethnography reacts to unanticipated occurrences in situ. Due to this inductive character—as we show in the examples below—ethnography can highlight the following two heuristics: (a) the work of informal epistemologies and tacit knowing and (b) the experience of moments of surprise and the unanticipated, which demand responsiveness and reflexivity on the part of the researcher.

Heuristic 1(a): Attending to Informal Epistemologies and Tacit Knowing

Attention to informal epistemologies—the organizational systems and thought structures by which people make sense of every day—often uncovers concerns that are not being addressed in more formalized modes of analysis. While epistemology generally references a mode of producing knowledge, we extend this concept here, both to ground our heuristic in the iterative practice essential to ethnography and to recognize that informants and participants in the field are building their own systems of knowledge all of the time (Toren & Pina-Cabral, 2011). People working in health services are under significant pressure, and informal settings (casual conversation, coffee breaks, after hours, and personal asides) may serve as spaces to vent and talk about things that are not directly related to work. Informal settings may also open up spaces for discussion of neglected or unacknowledged work issues or subjects that extend beyond work itself. Paying attention to informal epistemologies as they emerge in these settings expands the research focus beyond formal or “official” discussions, often catching conversations and exchanges that happen “in-between.” This allows for an integral view of the person; one who both performs a particular skill and is tethered to a life beyond professional and research contexts. These varied dimensions of the everyday influence one another in important ways that are not always immediately apparent:

Example: When working on a project to integrate peer professionals on crisis intervention teams, the ethnographer spent time with teams as they hired for the new peer positions. Some of the teams were having trouble filling these positions, and the project director did not understand why, as he knew that there were many interested applicants. By spending lengthy time in the office with one team leader and her clinical staff, the ethnographer learned that this team was made up primarily of people born and raised in The Bronx, and that their catchment area—meaning the area their service users came from—was also in The Bronx. The team leader was reluctant to hire a peer professional who did not share this affiliation with the neighborhood, because the team’s ability to relate to service users along lines of socioeconomic and ethnic identity was essential to how they understood themselves as able to do their work. This did not come up in formal discussion with the project director—in these conversations the team leader said she was still struggling to find someone who was “the right fit.” It was in the informal conversations when the team leader and her staff vented about the difficult hiring process that they articulated the importance of having someone who also knew the local neighborhood. “Knowing the neighborhood” is an index of tacit knowledge, and the unspoken condition for being “the right fit” that would be necessary to work with this team’s service users.

In this example, the underlying criteria of what it meant to be “the right fit” could not have been asked directly in interview or survey method inquiries about hiring or team dynamics. The questions would have been too leading. Nor could it have been known beforehand that local familiarity, the tacit knowledge of being part of the neighborhood, was shaping peer professional hiring practices. Hiring for peer professional positions was already a challenging task: This role was new in city mental health services. While a tremendous amount of focus and direct attention was given to the complexities of hiring peer professionals, this other dimension of hiring, of being the right fit, was not understood until it was observed ethnographically.

These and other similar concerns are missed if researchers only ask questions limited to prescribed and predetermined parameters as imagined in formal settings. They usually require lengthy and patient processes of trust-building, especially in the case of uncovering local and hidden identity constructions that are usually not spoken of. The “in between” of conversations is essential to understanding informal epistemologies that capture implicit thinking and styles of thought that actively shape how service provision is taking place in the present. This is a crucial concern for ethnographic work, and short-term approaches risk missing these insights or not understanding their full influence or impact on the field. Participant observation opens up the possibility to attend to these informally communicated but crucial aspects of health services research and provision, in this case, effective staffing and team building.

Heuristic 1(b): Responding to Surprise and the Unanticipated

Ethnography is an open research methodology that builds on reactivity within field situations and is characterized by field-specific opportunism (Breidenstein et al., 2013). It is central to ethnographic work that it be led in part by both the field and the object of study, and thus it cannot be fully laid out in advance if one does not want to lose epistemic momentum. Within health services, the settings in which care is provided characteristically depend on a vast and varied array of interests, actors, stakeholders, infrastructural aspects, technologies, material questions and resources, time pressures, and ethical questions. This makes it more or less impossible to anticipate all possible situations in advance. Therefore, instead of embarking on research with a premade plan, ethnography remains sensitive and responsive to what is taking shape amid all of these aspects in the present moment:

Example: The ethnographer conducted research on two different financial and therapeutic models for psychiatric care across Germany: The first was a capitation model in which a fixed budget was allotted for all hospital-based psychiatric treatment according to the number of service users treated; the second was a fee-for-service model in which reimbursements were tied to specific services and interventions. In the majority of the clinics working with the capitation model, there was an increase in cross-sector therapeutic treatment and greater flexibility across different therapeutic settings. However, what the ethnographer discovered during participant observation was that one major advantage of the capitation model clinics was their exclusion from case inspections by the health insurance companies. Again and again, staff referred to the relief they felt at not having to prepare for these case inspections. These inspections necessitated significant preparatory work from clinics, as the clinics must prove retroactively that their treatment decisions were financially legitimate. This administrative labor limited the hospital staff’s ability to attend to other tasks. The inspection process caused additional administrative work not only because of the collection of proof and documentation, but also because this reporting system demanded a kind of translational work for the staff—from psychotherapeutic accounting to administrative and economic justification. Unlike their colleagues working under the capitation model, the clinicians working with fee-for-service model did not have the same time or resources available to work creatively across different therapeutic settings. The difference in administrative load was revealed to be just as important as any particular therapeutic or financial intervention.

In this example, the research began with a focus on two different financial models and their comparison. However, through the process of participant observation, it became clear that what was more important in determining practical outcomes was the administrative burden of the different forms, rather than the financial distinctions. With this insight, the focus of the research questions and the relevant units of comparison could be shifted. Thus, the focus of the research questions was adapted to address practical aspects discerned by actors in the field in real time. Researchers were consequently able to capture the lived distinctions that more concretely shaped practice and outcome, which were different from the factors imagined at the outset of the project.

Ethnography may change its course and its relevant questions throughout the entirety of the research process and can thus complement other perspectives or avenues of inquiry, in this case, the focus on different financial models. This can be especially useful if the research process remains inductive in nature and if there is sufficient time available for a continuous, iterative, and open-ended inquiry. Participant observation is especially well-positioned to support attention to the complexities of “in situ moments,” such as the daily administrative burden discovered here, which would not have been captured through attention to treatment or financial dimensions alone. This responsiveness allows for unanticipated insights and surprising interactions to refine the research questions and areas of focus.

Methodological Principle 2: The Relational Perspective of Ethnography

Ethnographic research situates questions in context and searches for relationships between different topics. While other empirical methods often consider single aspects and variables independently from one another, controlling for variation and influence, ethnography allows for a cross-cutting view of multiple themes. In the context of health services research, this offers the means to better understand the contingences as well as the complex relationships that exist within and between institutions, policies, ethical concerns, and surrounding structures. The grounding of a relational perspective in this context depends on recognizing (a) the situatedness of local worlds and (b) the systemic embeddedness of health services.

Heuristic 2(a): Understanding the Situatedness of Local Worlds

Health services take place as part of a concrete bodily and material world. The interactions between service providers, service users, and material resources, such as transportation, documentation, communication, and supervision actively, shape everyday working experiences and the possibilities for action by both staff and service users. This can include the concrete dimensions of buildings, floorplans, rooms, and policies regarding closed and open doors, such as on hospital wards. It also includes less-tangible aspects, such as digital infrastructures, systems of documentation and accounting, and interpersonal exchanges. Every structural dimension and interaction that are part of service provision exist within these multiple situated interrelationships:

Example: While working on the evaluation of different psychiatric models in Germany, the ethnographer spent time with different professionals: doctors, nurses, social workers, therapists, peers support workers, etc. The digital documentation of care services posed an array of crucial problems; some anticipated, others not. In one particular hospital, it took nearly ten minutes for the computer to load a single service user information sheet, while the doctor had only fifteen minutes officially to speak with that service user. In another hospital, digital service user files were introduced, but some of the doctors felt that it was impossible to have an admission interview with a person and write at the computer at the same time. They argued that they were not able to establish contact with the service user and write at the computer simultaneously. This led to a time-consuming doubling of the documentation process—doctors would make notes and jottings while they were talking with the service users, and afterwards have to document the same information in the computer system.

While at first glance these examples could be seen to communicate a resistance to change on the part of doctors, extended participant observation revealed that the incorporation of digital technologies (a) was not uniform across hospital contexts, and nonetheless, there were some similar themes across the contexts and (b) had concrete impacts on the relationships between doctors and service users that were not considered at the outset of implementation. These attitudes could not be captured in statistical analyses of service user data and the streamlining of computer-based documentation systems. The ethnographic presence here allowed for the discovery of local specificities that vary from hospital to hospital, regarding the practical engagement with digital record keeping. Thus, it became knowable that while usually regarded as more efficient and secure, the usefulness of digital records in hospital settings is dependent on the quality and speed of technical support as well as the ways medical professionals understand their responsibilities to service users. This allowed the research focus to reflexively adjust to capture the ways the technological conditions of the built environment shaped practitioners’ response to digital record-keeping rather than assuming the difficulty was in the imposition of a new task.

Heuristic 2(b): Studying Systemic Embeddedness

Health services are not just self-contained projects but rather are embedded within larger institutional and cultural systems in complex ways. This larger context includes policymaking, ethical concerns, the distribution of resources, and so on. The different surrounding structure of urban and rural mental health services hospitals, for example, concerns more than just the experience of service users in the hospital, working with mobile teams, or with community-based services. The provision of care is additionally contingent on the lives of professionals beyond these service roles and the resources available to their clients in everyday life, which depend on the cultural and geographical dynamics of the surrounding environment. Different local settings have different health concerns, and hospitals are very rarely positioned to address what goes on for service users beyond the boundaries of their institutions:

Example: Nursing students in their final year of qualification were asked to take on the responsibility of managing an entire psychiatric ward. A phase of participant observation during this process began by trying to understand how these nursing students decided to prioritize the various demands of everyday care. The researchers observed that a significant part of the nursing students’ efforts focused on the organizational and technical side of caregiving: documentation, administering infusions, taking care of the physical needs of service users. At first glance, it appeared that forms of psychosocial support as they related to service users’ experiences outside the clinic were often neglected or considered a secondary task that could only be accommodated after all of the basic medical needs of service users had been met. However, when discussing these results with the nursing students, the prioritization of medical protocol was revealed to be a consequence of various safety regulations derived by policy makers that had sedimented into everyday clinical routine. These safety regulations were derived at a state level, rather than a local level, but the consequent policy decisions concretely shaped the nursing students’ actions on the ward and directed their focus towards service user accountability and safety as it was defined within the boundaries of the institution. Even though these policies originated outside of the hospital and the nurses’ ward, they were able to alter day to day practices through routinization over time, while obscuring the more local experiences that might more profoundly shape the accountability and safety of service users at the local level.

Health services, like the institutions within which they exist, operate within larger contexts. Everyday care practices are embedded in broader institutional and discursive backgrounds that concretely shape their procedure. By following a dialectic approach attending both to local practices and to contextual conditions, ethnography is well-positioned to examine the relationship between what happens in the local as well as wider contexts, thereby providing a means to oscillate between different levels of analysis. Instead of assuming fixed boundaries that delineate a particular care practice, ethnographic approaches recognize the interconnectedness of health services with their surrounding environments, attempting to analyzing a phenomenon in all its complexity. Such an analysis, however, requires time and resources to attend to various levels and facilitate a meaningfully entangled understanding.

The interconnectedness of institutional care leaves service provision vulnerable to various contingencies and the influence of unpredictable circumstances; in this case, the configuration of everyday practice according to policy decisions made well beyond the institution itself. In this context, ethnographic methods offer a means by which to attend to the permeability and variability of institutions and services, demonstrating that institutional routines are often a product of regulation and influences on a higher level and that local circumstances can be unintentionally obscured. The analysis of this dynamic can contribute to a better understanding of the complex relationships that exist within and between macro- and microlevels of service provision and potentially lead to substantial improvements in how service provision functions across such divisions.

Discussion

By necessity, ethnography has always been a type of epistemic mediation (Boyer & Howe, 2015). Today, this mediation is understood to be itself a method that reflexively examines the relationship between theoretical constructs and everyday practice. When engaging in interdisciplinary research, embracing this oscillatory perspective can yield answers to critical questions not formally asked but that came to the surface, iteratively, and dependent on the allotment of significant time and attention.

To this end, we have tried to show in this article how the four heuristics of ethnographic engagement situate research questions in wider political and ethical fields, offering better understandings of the systemic embeddedness of health services. We argue that ethnographic approaches aim to (a) capture the implicit thoughts and decisions that happen “in between” the official representations of service provision; (b) foreground reflexivity and responsiveness, thereby yielding a deeper understanding of the dynamic relationship between theory and practice; (c) assess the interdependencies of healthcare settings and those who work within them; and (d) attend to the permeability and variability of institutions and services.

What Oscillation Offers

This brings us to a consideration of the kinds of questions that can be answered with ethnographic field research. It is important to point out that while ethnographic research may seem more uncertain than other frames—in that the research questions are also an emergent part of the iterative process—this flexibility is a strength that allows for logical abduction. Abductive reasoning requires a different orientation to case and specificity, one that allows for observation, revision, more observation, and further revision. While it begins with specificity, this work engages an expansive and continual process of refining reasonable explanations based on subsequent encounters in the field. A researcher may begin with a specific question but end up revising this to better adapt to events experienced in situ. As a result, it is very rare that the ethnographic inquiry ends with the same set of research questions with which it began. This can put ethnography in conflict with primarily deductive or classical research designs—in which the interpretation of data is subject to the imposition of higher order categories. When data are messy, this imposition risks obscuring fresh insights (Hauser, 2012). However, as we have tried to demonstrate here, it does not have to be this way: The results of ethnographic research often end up with very different understandings of what is happening in the field (i.e., the daily practice of service provision) in comparison to what is imagined in the initial stages of a project and according to other methods of inquiry. This, we argue, has great utility.

Rather than seeking the answers to predetermined questions, the ethnographic endeavor has the potential to answer critical questions that were not formally put but rather emerge out of a constant iterative flow. In oscillating between theoretical frames and the daily practices from which they emerge, this process depends on the inherent instability of participant observation (Shattuck, 1997) and recognizes that the difficulty inherent in making necessary shifts in perspective is itself uniquely productive. Ethnography does not provide quick or easy solutions; it is itself a form of change via reflexivity and participation. It is a methodological movement that depends on participating, observing, and reflecting back and in the process bridges theory with practice and uncovers new connections and further lines of inquiry. It is this generative potential, necessarily distilled over time, that constitutes the distinctive yield of ethnography. ¹

To this end, ethnographic work pursues questions that are not only about what is done but also how something is done and the process by which actors and institutions find their way within theory and practice. In other research methods, the outcomes of a process are captured but not the mechanisms that contribute to that outcome. Moreover, ethnographic methods also surface the contextual and cultural factors that shape this process of navigation and make implementation sustainable. To access this information flow, researchers need to look not just at the information that has been identified as data but also at how these data are contextualized, how is it communicated, how it is dealt with, how it is used or operationalized, and by whom. This process can contribute to novel theoretical developments about institutions and complex processes of intervention, it can facilitate new interdisciplinary collaborations, and it can be translated into practical adjustments in the everyday provision of health services. Thus, ethnographic approaches are an instrument for both research and service provision.

Ethnography’s applicability in both realms helps to make sense of the “noise” that regularly plagues research study design and the subsequent attempts at change and revision of implementation projects (Hohmann & Shear, 2002). It exposes moments and interactions in which unforeseen details can actively impact the viability, sustainability, and effectiveness of health services provisioning, not only by evaluating the effectiveness of an intervention but also by contributing to an understanding of why and how that intervention took shape in context.

Challenges for Integration

The two methodological principles of ethnographic research offered here—its inductive character and the relational perspective—are essential if health services research is to benefit from the insights of ethnographic methods. At the same time, we recognize that some characteristics of this methodology may seem at odds with scientific realities, and this requires careful consideration of both the incorporation and the output of ethnographic research to be more compatible with the disciplinary needs of health services research and practice (Savage, 2000).

For example, ethnography is often misconstrued as being limited by its specificity. The work of the case study—detailed accounts of long-term research—is often criticized for failing to offer universal or generalizable knowledge. We want to correct this misunderstanding of the ethnographic case as well as offer a meditation on form. Cases, as presented in ethnographic work, are the product of extensive research and data collected over an extended period of time and derived from the intensive everyday work of participant observation (Evens & Handelman, 2005). The case may demonstrate specificity rather than generalizability, but what the case also demonstrates is density. A “case” will not be used in an ethnographic representation if that case does not itself represent a larger trend, theme, or process from the broader field, and, in this way, it is representative of more than a single specific instance. Bringing differently positioned researchers together in the process of building cases is essential to reframing this process and positioning cases as the synthesis of broader realities in the field. To this end, we offer a two-part suggestion for integrating the context and outcomes of ethnographic research.

First, the necessary conditions for collaborative work need to be positioned earlier in the research process. One of the most important adjustments for integrating ethnographic research is building participant observation into study design from the very beginning and supporting the endeavor through to the end (Pope et al., 2016). Having ethnographers on the ground from day one supports the active engagement with findings and fieldnotes for the entire research team. Taken over the entire process of data collection, this creates space for situated reflexivity across diverse research teams and for engagement with iterative work among differently positioned stakeholders from different methodological backgrounds. The sharing of field reports, the facilitation of member checking across the team (Brear, 2020), and the circulation of process memos prepared by ethnographers and others can build a more reliable and robust collective investment in iterative processes. By working together from the early stages of research, field notes can be distilled in briefs and interim reports that are both interdisciplinarily legible and retain the unique perspective offered by ethnographic engagement.

Second, while ethnographic work is often encased in lengthy manuscripts, positioning collaboration earlier in research and over the long term can facilitate cooperative writing processes and the production of diverse forms of publication and dissemination. On one hand, this means engaging with interdisciplinary journals and presentations, building bridges between disciplinary-specific language, and engaging as much as possible other researchers in the daily work of the ethnographic process such that all components of the research or implementation project are working in tandem with the collection of valuable contextual data. On the contrary, it means bringing researchers with different skill sets and perspectives into the reflexive dialogue foregrounded in ethnographic work. This could mean reformatting clinical reports as ethnographic vignettes or opening institutional protocol to interrogation. Allowing space to voice questions and concerns regarding service provision in theory and practice embeds reflexivity in project outcomes as they are produced. Meaning, it is not only the reconfiguration of end products in regard to publication but also the internal processes of research and analysis that can be adjusted to facilitate attention to crucial ethnographic concerns.

All of this demands a synthetic approach to research and the willingness to experiment with new forms of communication and reflection. This is not to say that all researchers need to adopt ethnographic methods; after all, ethnographers are specifically trained to take on this kind of work. Rather, our proposal is that health services researchers learn to work alongside and in coordination with ethnographers—and vice versa—from the developmental stages of research, thereby embedding attention to emergence and change in the structure of projects from the outset. In doing so, the goals and outcomes of ethnographic research might appear less disparate, less incongruous, and instead stand to deepen the complex work of qualitative research, furthering its capacity for novel insights through the negotiation of different perspectives.

Author Biographies

Lauren Cubellis, PhD, MPH, is Max Kade Postdoctoral Fellow in the Berlin Program for Advanced German and European Studies at Freie Universität Berlin. Her research examines notions of risk, crisis, and care at the intersection of cultural anthropology, psychiatry, and public health.

Christine Schmid, PhD, is a research assistant in the Department of Psychology and Ergonomics at Technische Universität Berlin and Medizinische Hochschule Brandenburg Theodor Fontane. Her research engages medical anthropology, the anthropology of science and technology, and the anthropology of knowledge and expertise.

Sebastian von Peter, MD, MA, is Professor for mental health services research at the Medizinische Hochschule Brandenburg. He holds a medical degree and a master’s degree in social anthropology. Sebastian von Peter is leading the home treatment team at the University Hospital Rüdersdorf/Brandenburg and is in charge of a research team for mental health services research that aims to evaluate psychiatric services using transdisciplinary and participatory approaches.