Some Complexities of Seeking Access for Ethnographic Research in Health-Care Institutions

Abstract

A growing body of scholarship reflects on the complexities and challenges of attaining access for ethnographic research. Some of these are particular to formal organizations, including the understudied gatekeeping role of institutional review boards (IRBs) in organizations that provide state services to vulnerable populations. This article examines access challenges encountered in a project to conduct observation and photography of the work routines of nurses in El Salvador’s public health-care system. An examination of the contrasting responses and outcomes of access negotiation with several different sets of authorities in the health-care system reveals that even in large bureaucratic research sites with formally structured gatekeeping roles, rapport developed over time with influential individuals can shape access negotiation outcomes, partly through informal social relationships. The findings also show that that without technically denying access, IRBs may set conditions that effectively make the “research bargain” too costly. Also suggested by the comparative analysis are organizational (hospital) and system (Health Ministry and public health-care system) factors that may make authorities at different levels more or less open, protective, or defensive in their stance toward cooperating with academic researchers. The article concludes by signaling the need for ongoing discussion on what social researchers can expect from IRBs, especially in developing countries.

Introduction

Over the past 20 years, scholarship on ethnographic research methods has seen a growing body of work across a range of disciplines and interdisciplinary fields reflecting on the complexities of seeking access to research participants and their environments. We know from these often “cautionary and confessional” descriptions (Van Maanen, 1991) that seeking and attaining access is rarely an orderly, unidirectional path through successive layers of gatekeepers to the participants. Instead, it is frequently described as a “messy” process (Billo & Hiemstra, 2013; Turner & Almack, 2017) in which the researcher must continuously renegotiate with people in gatekeeping roles. Moreover, seeking, attaining, and maintaining access takes place through the entire period of data collection, and some would say, even into the write-up and analysis of data (Riese, 2018).

One feature of many organization-linked field sites that remains under-explored in the literature are formalized research ethics bodies, which go by different names but are usually generically referred to as institutional review boards (IRBs). I would argue, somewhat tentatively, that the growing prevalence of IRBs will affect the negotiation of access to settings dealing with health-care, education, or other public welfare activities, by making it imperative to engage with top-level gatekeepers. This stands in contrast to many informal research sites, and also differs from what researchers of formal organizations encountered in the past, or what they encounter in settings where, for cultural or other reasons, there is less consideration for the privacy of people served by these organizations (Van der Geest and Finkler, 2004). This may be the case in less developed countries, where to date, there has been little research on access issues in general (Van der Geest and Finkler, 2004; Zaman, 2008).

In addressing these thematic and geographic lacunae in the literature, this article uses insights from ethnographic research conducted on nurses in El Salvador to contribute to a necessary conversation on the access challenges that arise in researching health-care and related organizations. I comparatively examine the process and outcomes of attempts to attain access to several different kinds of clinical settings within El Salvador’s public health-care system, research that involved mainly observation and photography of nurses in their daily and nightly work routines. Although the uniqueness of every ethnographic research initiative precludes general theories about access and gatekeepers, comparing differing responses to access endeavors within the same project, as other scholars have done (e.g., Wanat, 2008), can help to identify better practises that lie within the researchers’ control, as well as characteristics of the organizations and their broader environment that also shape access.

In assessing differing degrees of access in my research on the nurses, I make use of Grant’s (2017) typology of hierarchically ordered tiers of gatekeepers in organizational settings as micro, meso, and macro levels. The analysis of my “access journey” (Grant, 2017) confirms the common characterization of the process as messy, nonlinear, and not always successful. A key factor in the more fruitful attempts was having developed good relations over the previous 3 years with influential individuals at both the macro and meso levels of the health-care system who could influence other gatekeepers both directly and indirectly. Working against access in the less successful case, particularly at the meso level, were the lack of an influential advocate for my research within the organization, and the relative autonomy of doctor-led organizations (hospitals) vis-à-vis the top-level health-care system authorities. Finally, the amount of time allotted for research also affected access in terms of variety and depth of observation that could be conducted.

Access as “Messy” and Nonlinear

In social research, access is the necessary process of contacting, communicating with, and building relationships with people whose interpretations, customs, social roles, interactions, and actions we seek to understand (Feldman, Bell, & Berger, 2003). A considerable body of scholarship in the past two decades closely examines access-seeking and negotiation with gatekeepers as a crucial and often painstaking component of ethnographic research (Crowhurst, 2013; Feldman et al., 2003; Turner & Almack, 2017; Wanat, 2008). Much of this work responds to the dissatisfying tendency in prior (1970s, 1980s) writings to gloss over the access process, or to describe only the successful attempts, portraying them as “seamless and effortless. . . [whereby] the researcher simply decided what site or sites she or he wished to study, asked for permission, and received it with nary a rejection” (Smith, 2001, p. 5). Attention to the difficulties and failures in obtaining access can help to identify common pratfalls and can prepare researchers to respond flexibly to both hurdles and unexpected opportunities (Carmel, 2011; Cipollone & Stitch, 2012; Scourfield, 2012). As well, delays, backtracking, and denial of access on the part of gatekeepers can yield information about the research setting, such as how authority flows (Mannay & Morgan, 2015).

What researchers seek in engaging gatekeepers has been likened to a “research bargain”: In exchange for access, the researcher agrees to something that can range from adhering to conditions around their presence, to taking a more active role that contributes in some way to the gatekeeper’s goals and tasks (Horwood & Moon, 2003; Sanders, 2006). Often the “bargain” that is attained must be continually renegotiated (John, 1990; Turner & Almack, 2017). This is partly because multiple gatekeepers situated at different levels or positions of the research setting may not all have the same understanding of the bargain (Horwood & Moon, 2003). But even for any single gatekeeper, especially at higher levels, an expression of approval may not translate into access at lower levels (Grant, 2017).

Access Complexities in Formal Organizations

Seeking access to formal organizations entails distinctive challenges (Cipollone & Stitch, 2012; Delamont, 1992; Troman, 1996). While some speculate that these research settings are easier to access than informal settings because lines of power and influence are clearer (John, 1990), more commonly they are described as presenting a greater number of challenges and impediments. Some of these are seen as stemming from the nature of bureaucracies. For example, organizations adhere to routines that, in the minds of gatekeepers, are likely to be thrown off by a researchers’ presence, and their raison d’etre is to pursue goals to which a researcher generally has nothing concrete to contribute. As well, organizational gatekeepers have a repertoire of methods to withhold cooperation, such as “controlling communication” while giving the impression that they are granting access (Wanat, 2008). Alternatively, they can set the terms of the research bargain impossibly high (Wolf, 2004, cited in Scourfield, 2012). In health-care systems, hospital authorities can deny access “either formally as through a research review committee, or covertly through denial of access to particular areas, activities, or potential informants” (Matthews, 1987, cited in Zaman, 2008, p. 137). Indeed, whereas institutionalized ethics protocol entities could be described 15 years ago as an “extreme case” of an organization’s means of impeding access (Wolf, 2004, cited in Scourfield, 2012), IRBs are now much more commonplace. This will be discussed in greater depth below.

A further challenge in organizational settings is that multiple tiers of formally structured authority make it hard to know which level of gatekeeper to approach first, or whether to engage more than one level simultaneously (Feldman et al., 2003; Scourfield, 2012). Grant (2017) offers a useful vocabulary regarding access negotiation in formal organizations: The macro level refers to access negotiation with top-level directors who are in a position to grant official access. At the meso level are mid-level managers who grant access to particular settings. At the micro level are the research participants whose cooperation is sought in order to observe “everyday activities.” This typology does not preclude further divisions within each level. As well, in larger systems of multiple interlinked and hierarchical organizations, what is macro, meso, or micro is relative; each organization may be considered meso in the broader system, but its directors are the macro level for that field site. An additional qualifier to categorizing gatekeepers based on formal roles and titles pertains to organizations that are large workplaces, like hospitals: These are often “fractured fields” with inter-occupational tension and rivalry, and differing occupational subcultures (Carmel, 2011). This may mean that gatekeepers in different units contradict one another in regard to the merit of the research, regardless of hierarchy.

Researchers must nevertheless confront the reality of the organization’s hierarchically positioned gatekeepers. Here an interesting debate can be identified in the literature. In older writings on the topic, some ethnographers posit that negotiating macro-level access can be dispensed with (Buchanan, Boddy & McCalman, 1988). In more recent work, others have argued more moderately that meso level access should be pursued first as a way to open access at the macro level; the latter will naturally flow from the former (Chambliss, 1996). Supporting that general argument, many have found that obtaining macro-level access is not enough, in part, because, as Wanat (2008) observes, access is different from cooperation; the access that senior gatekeepers grant (which might be more appropriately called authorization) does not guarantee cooperation at meso and micro levels (see also Grant, 2017; Reeves, 2010; Van Maanen, 1991). As well, some researchers find access relatively easy to attain at the macro level only to encounter problems at lower levels of gatekeeping (Van Maanen, 1991; Wanat, 2008). However, this can also work the opposite way: In some research settings, it may not be sufficient to develop excellent rapport with mid-level gatekeepers as Chambliss (1996) did in his ethnographic research on nurses, or to secure participation agreements at the most micro level, because it “may not be appropriate to accept offers of access [from meso or micro level] without macrolevel permissions” (Grant, 2017, p. 7). Indeed, ignoring top-level access can lead to a partly completed project having to be called off (Daniel-Echols, 2003). In formally structured organizations, a lesson from numerous scholars’ experiences is that all levels of access need negotiation work on the part of the researcher (Grant, 2017; Turner & Almack, 2017). It is not a unidirectional process either from top to bottom or vice-versa; attaining access at one level cannot be assumed to prompt gatekeepers at a higher or lower level to follow suit.

While there are no blueprints for negotiating access in organizational research settings (Carmel, 2011), literature on researchers’ experiences of successes and failures nevertheless yields useful generalizations on factors that shape gatekeepers’ responses. A number of scholars (Cipollone & Stitch, 2012; Scourfield, 2012) observe that for too long, discussion of the access process overemphasized the individual traits and tactics of the researchers that go into creating (or impeding) good relationships with gatekeepers. Overlooked in such discussions are the system factors—the internal politics of the organizations and how these are affected by broader social forces, such as policy shifts. For example, a context in which significant policy change is being proposed or implemented may make access negotiation more challenging as participants or gatekeepers develop a “siege mentality” toward outside observers (Troman, 1996). It is also conceivable that certain kinds of policy shifts could affect access positively, leading to greater transparency and a wish to showcase changes and achievements. At the same time, these systemic factors interact with the politics of the research topic in a way that can affect the alignment between the organizations’ goals and those of the researcher, either positively or negatively. It goes without saying that gatekeepers may have personal, idiosyncratic stances toward the outsiders’ gaze that are impossible to foresee. But many will have values and agendas that reflect the goals of their organization or group within it (e.g., along occupational lines), whereby they may see the proposed research as an opportunity for those to be validated (Clark, 2010).

In terms of access negotiation strategy, allotting an adequately long time to the research is at least a necessary (though not sufficient) condition in successful efforts (John, 1990; Smith, 2001; Troman, 1996). Time is needed for the development of relationships that can positively influence access; this is important given that “informal gatekeepers [who are] not necessarily in structural positions. . .[can] influence[] others through the strength of their personality and character” (Reeves, 2010, p. 322). As well, with more time dedicated to work on the same topic, researchers can develop a track record and reputation among gatekeepers based on prior output (Wanat, 2008). The challenge for researchers is that the academic calendar can shape their schedules to be out of sync with the time priorities of gatekeepers and research participants (Smith, 2001).

Access and IRBs

An increasingly pervasive feature of organizations that provide state-funded services to vulnerable populations, such as hospitals, schools, prisons, and so on, are IRBs. Within organizations such as hospitals, IRBs may be considered part of the macro level of access—below the top-most authorities of the hospital, but with more authorization capacity than the medical or nursing staff in charge of particular areas of care. IRBs cannot be presumed to be in sync with, or under the influence of, directors and other senior officials. The pervasiveness of IRBs as a dimension of negotiating access is a recent and still internationally variable phenomenon. In regard to health-care, social researchers have observed that hospitals in the Global South are less protective of patients’ safety entitlements vis-à-vis researchers than in the North (Van der Geest & Finkler, 2004; Zaman, 2008). “Developing countries in general . . . do not have a strong culture of research and no institutionalized body [in the form of IRBs] to regulate the issues of ethics, rights, or privacy issues” (Zaman, 2008, p. 138). There are, however, vast differences across the developing world in this regard. For example, Latin America and the Caribbean have been ahead of other regions in creating IRBs (Bartlett, 2008), with some of the smallest and poorest countries in this region, Honduras and El Salvador, establishing them in major hospitals in the early 2000s (Camp, 2008; Espinoza Pérez, Rodríguez Yunta, and Sánchez, 2011). Furthermore, across the developing world as a whole, IRBs have gradually been increasing in prevalence (AHC Media, 2007). It is a situation in transition. There is a substantial literature on how hospital IRBs function in less affluent countries for clinical or drug trial research. However, we also need to understand their gatekeeping function for social research, especially if, as was the case in my research, final sign-off from home-base university IRBs for studying health-care organizations is conditioned on obtaining written ethics approval from parallel bodies in the research sites. ¹

Description of the Research

The ethnography-photography endeavor at the center of my research was the final phase of a larger project on nurses’ labor conditions in El Salvador and Nicaragua. The overall study aimed at understanding how nurses’ conditions, salary, employment status, and the stresses and rewards of the work were affected by recent regime change and health-care policy shifts in the two countries that would affect the public health-care systems. Between 2010—when the project began—and 2013, I had conducted dozens of individual interviews and five focus groups with over 30 participants in total in El Salvador, and almost as many in Nicaragua. The addition of observation and photography to the larger project was aimed at showcasing what nurses’ routine responsibilities consist of, what difference their work makes to the health of the population, and how they cope with limitations in both physical and human resources. Regarding the photography specifically, my idea was that a visual presentation of nurses at work, alongside narrative description of what they do, presented in a freely accessible electronic book format, would be a way to metaphorically and literally address the invisibilization of nurses’ contribution to health care. The goal was partly one of social justice and advocacy in favor of nurses. I also conceived of it as a way to give something back to the nurses collectively for their facilitation of my research up to that point. I aimed to, and ultimately did, produce a narrative and visual electronic document, with photography done by a collaborator and friend, Jim Gronau.

As clarified in the project description I provided to my university’s Research Ethics Board (REB) and to potential participants, all patients (and their parents in the case of children) would be asked their consent to be included in any photograph of nurse–patient interaction. None of the patients’ faces would be photographed, and none would be named in the write-up of the results. Because for the nurses we would not be concealing their identities as per normal ethics protocol, additional sensibilities were required around their participation. “[T]he photographic researcher captures a person’s image and with it intimate elements of their identity that other data collection methods cannot access” (Langman & Pick, 2014, p. 709). This carries a particular onus for upholding dignity in the nature of the images that are made public (Langman & Pick, 2014). It also meant giving those who would be photographed an opportunity to approve the images we would were selecting, as well as the accompanying narrative descriptions of them.

The choice of El Salvador over Nicaragua for this final phase of the research was based on several considerations. I had developed a larger number of close contacts with nurses there compared with Nicaragua. I also knew my way around the capital city, and the country, much better through having spent several years on previous research since the mid-1990s. By 2013, El Salvador was 3 years into an expansive restructuring of its health-care system by the left-of-center government of the Farabundo Martí National Liberation Front (FMLN), elected in 2009. I hoped to do some of the observation in the primary care sector, which was the major target of the reform. Also, I sensed that authorization would be more feasible and take less time in El Salvador than in Nicaragua, based on the FMLN Health Ministry’s transparency with information regarding the health-care system overhaul, and the accessibility of its annual reporting on health indicators, personnel numbers, and so on. In Nicaragua, based on interviews with nurses and many other social actors since 2010, I came to understand the government of the Sandinista National Liberation Front (FSLN), elected in 2007, as more hermetic and defensive about anything pertaining to state resource allocation. I had been inside both private and public sector hospitals to do interviews over the previous 3 years, but the health-care reform made it logical to focus this phase of the research solely on the public system.

There were several steps in the planning of the observation and photography component. In May 2013, I workshopped the idea of the e-book with a group of nurses who were former interviewees, focus group participants, and key informants. I also sought their advice about how to obtain authorization. In that same time frame, I met with hospital Directors and heads of nursing at two specialist (also called third level) hospitals to verbally propose the project in general terms. If their initial verbal response was positive, my plan was to follow up with a detailed written proposal. The two specialist hospitals chosen were Hospital Nacional Benjamin Bloom, which was the national Children’s Hospital, and Hospital Rosales, the country’s main general hospital. Both Hospital Rosales and Hospital Benjamin Bloom were chosen for the sake of continuity in the larger project: I had already interviewed a number of nurses in both hospitals, and I had been inside Rosales in about a decade earlier to interview other kinds of hospital staff for a different project. In describing the project’s goal to hospital Directors (who were doctors, as is the case in all the public hospitals), I was forthcoming about my perspective that nurses are insufficiently recognized. Directors and the heads of nursing gave verbal approval and advised me to seek official authorization from pertinent committees in their institutions. For a third hospital I had hoped to include, my hand-delivered, hard-copy requests for a meeting with the Director received no response.

The hospitals can be considered the meso level of access within the public health-care system, though within each one there is also a hierarchy: at the macro level, the Directors and heads of nursing; at the meso level, the charge nurses in the wards or units; and at the micro level, the nurses who volunteered to be shadowed and photographed, and the patients whose consent I also needed to photograph nurses interacting with patients. It did not occur to me at the time to seek access to the hospitals through the more macro level Health Ministry authority structures, because I believed that it would not be difficult to attain access at the hospitals themselves, and I saw this as better for collaborative relations.

During the month of June 2013, I sent a relatively detailed description of the project to those whom I had already met, as well as heads of committees to whom I had been referred. In addition to hospitals, I wanted to include primary care settings. Based on the advice of several nurses, I decided to pursue access to that level through the overarching Health Ministry authorities. The Ministry, as the macro level within the system, is itself hierarchically subdivided, with the Minister and two Vice-Ministers as the top-most authorities, and a number of additional divisions. Among these are the Nurses’ Unit, whose Director, Guadalupe Ibarra, ² I was advised to approach for support and direction for the project. Ibarra, who represented the highest level of nurse management in the public health-care system, responded very positively to my emailed description. She advised me to send a brief letter to the Minister and Vice-Minister of Health, and subsequently edited my draft versions of each. She also committed to advocating for my project with those officials to obtain authorization. Upon my arrival in July, Ibarra would connect me with an adjunct entity called the Nurses’ Research Nucleus, with the rationale that working through the Nucleus would not only be helpful to me but would also “raise the profile” of the Nucleus. This was a contribution I was very pleased to make as part of a research bargain. She also indicated that she and the Nucleus leader would advocate for the Ministry-level authorization for the project. In essence, within the macro level of the health-care system, it appeared that one set of gatekeepers was offering to support me in approaching the highest level gatekeeper, working through the formal command chain.

Three Access Negotiations

Upon returning to El Salvador on July 1, I began meeting with members of gatekeeper groups in the two hospitals (with Jim Gronau in attendance at most) and with Ibarra’s colleagues in the Health Ministry Nurses’ Unit (which I attended alone). My hope was that we would have the last 2 weeks of July to shadow nurses in three or four clinical environments during four or five shifts in each setting. In what follows I describe the positive signals of access I encountered in each of the three groups of gatekeepers with whom I attempted to negotiate, as well as the hurdles in each case. I will also suggest factors that made the difference in being able to conduct the research.

Hospital Rosales

As mentioned, at the macro level of this organization, the Hospital Director verbally indicated approval of the basic idea of the project. During the first week of July, I had three meetings with hospital officials—two doctors and one nurse who made up the institution’s Professional Development Unit—for advice about the research ethics process in this hospital. They instructed me on the logistics of making a proposal to the Committee for Ethics in Clinical Research, and pledged to facilitate an expedited hearing. It seems that these individuals did ensure a faster process than might have otherwise been the case, and that they did so to accommodate my time constraints. They also implied that they would speak in favor of the project. One of these officials was herself also a member of the Committee.

Also at the macro level of the administrative structure, I met with Hospital Rosales’s Head (Jefa) of nursing, whom I had interviewed at length in 2011. She appeared highly receptive, taking time for several long conversations in which she described worsening conditions for nurses in the country, and explicitly reinforcing the importance of my project. She also made suggestions as to which units of the hospital to focus on, and connected me with nurses in charge of those units. It made sense to me at the time to pursue these leads based on the positive verbal signals I was getting from the people who advised me on the ethics process. At the micro level, the charge nurses of two units also responded positively to the project. I accepted an invitation by one of them to explain the project to a group of nurses in her unit—the Emergency Department—as a preliminary step to inviting volunteers.

In the first of my two meetings with the Ethics Committee members in this hospital, I faced a set of predictably adversarial questions. However, I was taken aback by the accusatory assertion by one committee member that my photographer had been seen taking his camera out of his backpack and taking photos inside the hospital. I responded that this could not be correct, because Jim and I understood very well that such behavior would be unethical. In fact, whenever Jim accompanied me to meetings at this hospital, he never brandished his camera inside the hospital or on its grounds.

At the subsequent meeting a week later, the committee provided a written verdict that technically granted authorization but with several very restrictive conditions. Three of these, in particular, made for a research bargain that I could not accept: a payment to the hospital of $500, the requirement that anything I would publish from the research be approved by the Committee, and the need to have at least one witness for every patient’s consent. From the committee’s perspective, the decision was undoubtedly based on a sense of duty to protect their patients, and it is entirely possible that previous negative experiences either with researchers or photojournalists had left them appropriately wary. But for me, the conditions they imposed would “interfere too much with the principles of the proposal” (Horwood & Moon, 2003, p. 107) and would have been challenging to deal with in the time that remained. Only later I discovered that if I had sought authorization from the top-most authority in the Health Ministry instead of requesting access at the hospital, my access to Rosales would have been covered by an overarching permission. But once the ethics review started in Rosales, Ministry officials were unwilling to override it. Ultimately, I decided to exclude this hospital from the research.

Hospital Benjamin Bloom

At the macro level of administration of this hospital, the verbal response to the idea of the project when I pitched it in May was not only positive but also very enthusiastic. The Director and two additional doctors on the administrative team of the hospital gave me a full hour of their time. They committed to supporting the project, and instructed me to direct the full written proposal to the Public Relations (PR) Committee, which is the same body that handles local journalists’ requests. In the verbal research bargain that we reached at that meeting with the Director and his colleagues, the only thing asked of me was the provision of a copy of the final e-book for the hospital’s use.

When I was back in El Salvador in July, it became clear that the written proposal at Bloom would not be subject to a detailed or adversarial ethics review process. Jim and I met only with the head of the PR Committee, whose principal condition for our access was something to which we were already bound by my own university IRB: not identifying patients in the photography or narrative. He additionally said that we would not be authorized to do the work during night shifts, and that he personally would have to accompany us to each care area at the start of each shift we would be observing. For reasons he did not make clear, he did not end up doing that, and I was able to observe during a night shift in one of the units.

Below these macro-level structures of gatekeeping at this hospital, I had a good relationship with a nurse who was not part of the administrative structure, but who had worked in a supervisory capacity there for over 20 years. This nurse, Magalí Quiñónez, reinforced my access effort at Bloom in direct conversations with the Director. Three factors here seem important: the nature of my relationship with Quiñónez, the respect she enjoyed in and beyond her profession, and the nature of inter-occupational hierarchy in this hospital. To backtrack to the initiation of the larger project, my first nurse interview in El Salvador, in May 2010, was with Quiñónez, a highly activist nurse who wore many hats: In addition to her full-time position at Bloom, she also held a teaching position in the national university’s Nursing program and an elected position on the Executive of the National Association of Nurses of El Salvador (ANES). Quiñónez, who became a key informant and also a friend, represented a critical, change-seeking perspective on the nursing profession. In this view, the urgent need for improved conditions for nurses and greater valuation of their expertise were seen as entwined with the quality of nursing care. There is little doubt that Quiñónez’s advocacy and activist orientation facilitated an alignment with the goals of my research, and that this in turn positively disposed her toward the project.

Second, Quiñónez enjoyed widespread respect among her nursing peers as well as in the medical profession, not least because of the competence she had acquired in every highly complex area of care in this hospital. Third, the doctors’ structural relationship with nurses as an occupational group in Hospital Bloom seemed to be unusually horizontal. The Director himself observed in our meeting that nurses in this hospital have more autonomy and respect than at any other public hospital. This point that was echoed by the Director and Vice-Director of nursing at this hospital. Bolstering this claim was evidence of a more oppressive and conflictive situation for nurses at Rosales; in interviews and focus groups, several nurses told of the administration attempting to add beds to units, and to open new units, without expanding the nursing workforce in the hospital. Quiñónez herself described a jocular, peer-like interactions with the hospital Director, as for example, when she chided him not to forget his appointment to meet me (which he did on two occasions). This combination of formal and informal relations may have allowed my research proposal to bypass the impersonal committee process, which may or may not have resulted in access. Although there was never any written statement of permission from this hospital, the verbal permissions from PR, combined with a signed statement from the Ministry, were sufficient to enable entry.

An additional distinguishing feature of Hospital Bloom that may have led to the expedited, PR Committee route for my research, instead of the IRB, was that the hospital was more open to the presence of foreigners, in turn because of positive experiences in recent years. Through established partnerships with U.S. hospitals and medical charity organizations, Hospital Bloom has benefited from the presence of U.S. specialists—doctors and nurses—who have volunteered their time to perform complex procedures on children in this hospital, while simultaneously training their Salvadorean counterparts. While we were in the ICU (intensive care unit), a group of about 30 U.S. citizens linked to a charity that had funded a special heart surgery program were permitted to observe and even take photos in the unit, supervised by the head of the PR Committee. To my knowledge, Hospital Rosales had not enjoyed these kinds of international partnerships.

Attaining access in Bloom, however, was not a completely smooth or linear process from macro to micro levels. For one, there were time delays in arranging meetings with key people, including hospital Director, to pitch the general idea of the project. On two occasions he did not arrive for appointments I had been granted; it took Quiñónez’s reminding and nudging to make the meeting finally happen. Also at the macro level, the Sub-Director of nursing was not at all receptive to the project, and seemed intent on discouraging me. For example, even after I had been given verbal authorization by the PR person and from the Health Ministry (to be described below), she told me that I could not yet start the research.

Attaining access in the form of cooperation at the micro level at the two areas of care I chose in this hospital—the ICU and the Oncology department—proved to be relatively unproblematic. The charge nurses were receptive and helpful in allowing me to make my pitch to the nurses on duty. In each of my requests for volunteers for shadowing, two or three nurses stepped forward. Throughout the shifts we observed, I had to ask permission of a number of patients and their parents to be able to photograph nurses providing them hands-on care. In the case of children, if they were above a certain age, a nurse participant pointed out that parental consent was less important than that of the children themselves. This was an important lesson on how academic IRBs’ concerns do not always mesh with ethical realities on the ground.

Health Ministry Nurses’ Unit

The Nurses’ Unit in the Health Ministry oversees the conduct of the nursing profession across all the public sector hospitals and primary health-care settings. At this macro level of access, which was essential to observing nurses in primary care settings, what had earlier looked like a very optimistic scenario of access negotiation soon turned into a puzzling withholding of support. At a brief initial meeting with Ibarra at the Nurses’ Unit, she presented me immediately to the Research Nucleus whose regular meeting was about to begin. The seven members in attendance and the Chair, Graciela Rodríguez, expressed a positive view of the project and a willingness to provide logistical support. The group decided that for the hospitals, they would await the results of the processes I had launched there. For the primary care sector, they would help me select sites at their next weekly meeting. However, there were already clues at this first meeting regarding a noncollaborative, disapproving stance on the part of Rodríguez, who asked me how it had been possible for me to do the earlier interviews and focus groups in 2011 without any official permission. About a week later, it became apparent that the Nucleus had not moved toward seeking the Ministerial authorization that Ibarra had assured. Rodríguez informed me by telephone that authorization would likely take weeks to obtain, much longer than the month I had programmed to be there.

Worried and unsure of what to do, I called Quiñónez to ask her advice. She in turn contacted a close friend who worked in the Health Ministry department above the Nurses’ Unit. I had also met, interviewed, and socialized with this person, Delmy Hernández, on several occasions.

At a social dinner that Quiñónez arranged the following day with Hernández and several others, she pledged to unblock the process. A day later, she forwarded me an email exchange in which she advised Ibarra that the Ministry had authorized my project including the use of photography, and instructed her to begin helping me immediately with the logistics. Ibarra’s reply to Hernández reassured her that the Vice-Minister herself, Dr. Menjívar, had indeed authorized the research, and that the Nucleus was already “working with” me and would start programming visits the very next day. However, at that July 12 meeting with the Research Nucleus, Rodríguez began by dictating a new plan for my research, while scolding me for not seeking permission for the earlier phases of the project. Her new plan omitted the two hospitals where I was immersed in procuring access. I would not be able to do the research at Benjamin Bloom because “the specialized hospitals are difficult and different.” When I asked her if she was not aware of the Ministerial authorization for my project, she stated as follows:

No, that won’t help. People in high positions don’t always know. We’re just respecting the process. . . You started the process [the larger research project] without authorization, so you won’t even be able to publish results here. . . You should start at the primary level of care instead of the hospitals. . . You should have gotten permission in 2010 [to do the interviews and focus groups] from the National Ethics Committee.

With this reproach, Rodríguez seemed to be revealing the core of her resistance to the project. At about that moment, Hernández happened to walk by the meeting room. Detecting that it was not going well for me, she joined the meeting and in a brief, somewhat heated exchange with Rodríguez, politely but firmly overrode her. She stayed in the meeting to participate in the discussion about field site selection—which hospitals, and which primary facilities, would be the most appropriate sites to focus on. Before the meeting ended, Hernández directed Rodríguez to help me coordinate the specific locations, including introducing me to relevant nurses within the system. The following day, Hernández personally delivered the Ministry’s letter to my home, signed by its Chief of Communications. Although this was a blanket authorization to conduct the research in its facilities, it did not override the process that was still underway at Hospital Rosales. Through the Ministerial authorization and the Nurses’ Unit support I was able to conduct the research in three additional clinical settings, two in the primary sector and one in a second tier hospital.

Conclusion

In looking at the nature of the challenges I encountered in the access efforts for this project, some of the factors that shaped the responses pertain to decisions made in the conduction of the research. Others have to do with organizational and system-level characteristics, and the political and policy context of the time period. Among the factors that worked in favor of access attainment is the fact that the larger project had been underway for almost 3 years. During this time, I produced and socialized results in several formats, including a lengthy written summary that I circulated among key informants, heads of the National Association of Nurses of El Salvador, and others, and several public presentations to audiences of nurses I had interviewed. This earlier groundwork, in turn, was facilitated by the fact that access is generally easier for interviews and focus groups than for observational methods (Grant, 2017), though this is not always the case (John, 1990; Wanat, 2008). My research summaries may have demonstrated to key individuals that my project aimed at supporting the critical, change-oriented perspective that these particular nurses adhered to. As well, the prior components of the project enabled me to develop relationships with those supportive individuals. The interventions by Quiñónez and Hernández, in particular, also underscore the importance that informal relations between researcher and gatekeepers, and among gatekeepers at different levels, can have in the access-seeking process (Chambliss, 1996; Reeves, 2010; Zaman, 2008). These can operate outside of the institutional command chain. In addition, it is possible that my status as an academic researcher from outside the nursing profession and the country, who was offering to shine some light on nurses’ conditions and contribution to health beyond their own social realms, positively affected receptivity on the part of those nurses who advocated for and cooperated with the project. In this respect, it may have mattered that no one had conducted systemic research on nurses’ labor conditions in El Salvador (or Nicaragua) before; gatekeepers in contexts and countries that have previously been under-researched may regard new research projects with more openness, ³ perhaps because the outcomes are an unknown quantity.

There may also be system factors at play. If the policy context of the research settings matters (Troman, 1996), it is reasonable to speculate that attaining macro-level access to El Salvador’s health-care system was a function of the relatively open nature of the regime. Three years into its recently launched expansion of the health-care system, the FMLN government had become accustomed to sharing information about it with outsiders, including, for example, other Latin American policy-makers seeking lessons for their own countries. There may well have been a general sense among Health Ministry officials that there was nothing to conceal and much to be proud of in its public health care settings.

Regarding the difficulties I encountered, perhaps the main cautionary note concerns the amount of time that I had allotted for the ethnographic piece. Complex bureaucracies have time priorities that may not mesh with those of researchers bound to academic calendars (Delamont, 1992). Notwithstanding the weeks spent workshopping, corresponding with, and meeting with people in gatekeeper roles throughout May and June 2013, I underestimated the legwork required for authorization, especially for dealing with the formal committees. A more realistic scheduling of the work would have permitted more hours of observation in the settings I chose, or observation in additional clinical settings. In general, an oversight of qualitative researchers in regard to their time commitment to a research endeavor is that they do not always “think[] about how they sometimes fail to ‘give’ themselves access” (Riese, 2018, p. 5). As well, not fully grasping the complexity of access in the health-care system as a large bureaucratic setting led to failures of interpretation: In both the Rosales case and the Nurses’ Unit, I misread positive initial verbal signals, incorrectly assuming that there would be a basic alignment among those who were expressing verbal support and those with the ultimate decision-making authority.

Having (or not) a supportive and influential contact within an organization may go some way to explaining the contrasting outcomes in Hospital Rosales and Hospital Bloom. Although my interactions with the Head of nursing in Rosales were quite friendly, it was not as close a relationship as I had with several nurses in Hospital Bloom. There was really no one to advocate for the project in Rosales in a way that might have generated more enthusiasm for it at the macro level of the hospital, or in its ethics committee process, whereby the latter might have been more lenient in the terms it offered in the access “bargain.” The differing responses at the two hospitals also suggest that organizational cultures, including the nature of relations between occupational groups, can differ even within the same institutional system. A system factor that also played a role in the Rosales case was the unwillingness of the Ministry to override Rosales’s IRB process, suggesting that the autonomy of doctor-led organizations is respected by top-level health-care system authorities.

At the Research Nucleus in the Health Ministry Nurses’ Unit, Rodríguez enjoyed a level of autonomy and authority in her role, whereby she could have delayed the project indefinitely were it not for the crucial relationships and interventions that broke the impasse. She was not in step with her colleague Ibarra in seeing the project as in some way a benefit to the Research Nucleus, an opportunity to collaborate in the research in ways that could have gone beyond gatekeeping support and logistical assistance. Her noncooperation with me, and her push-back on higher level authorization, suggests that gatekeepers have idiosyncratic agendas, interests, and motivations whose sources we can only speculate about. It is impossible to know whether her resistance to the higher level gatekeeping was a function of inter-occupational rivalry and tension typically found in large health-care organizations (Carmel, 2011).

Concerning the relative importance of macro, meso, and micro levels of gatekeeping, in the present era of the increasing prevalence of IRBs attached to research settings, research involving ethnographic observation cannot sidestep the macro level. This is not least because obtaining IRB certification in the research sites is often built into the requirements of academic IRBs. This means macro-level authorization must be sought alongside cooperation from gatekeepers who are closer to direct participants in the study. But my experience suggests that macro level, official approval may be influenced by informal relations and organizational subcultures. The line of questioning I faced from the committee in Hospital Rosales suggests that IRBs themselves are not immune to such factors. More research is certainly needed on ethnographers’ encounters with IRBs in organizational settings. The objective is not to sidestep entities that have been established to protect vulnerable populations we may seek as research participants. Instead, understanding the complexities of this additional layer of gatekeeping can equip researchers with the ability to pivot access strategies in response to diverse scenarios we encounter in the field, and to learn more about our topic from opportunities and hurdles that are beyond our control.