Therapeutic citizens or therapeutic refugees? An examination of triage, refugeehood, and referral health care in Tanzania

Abstract

Referral health care refers to the movement of populations between different levels of a health care system, often to seek specialized care at higher level centers. It requires mobility of populations. Critical questions on the shortcomings for accessing referral health care, especially for refugees in sub-Saharan Africa, remain unexplored. How do global guidelines on referral health care play out in the local context of a country like Tanzania, where refugees are not legally allowed to leave the refugee camp without permission? How do health care providers navigate this complex space, and how can we leverage anthropological theories of biopower and therapeutic citizenship to better understand this context? Based on six months of ethnographic fieldwork in Tanzania between 2011 and 2021, I trace how global refugee policy manifests itself locally. I argue that referral for refugees in Tanzania is as much a political and social process as it is medical. The bureaucratic process favors certain pathologies over others and creates significant delays in referral. Triage takes on nuanced forms whereby a certain aspect of a patient’s case (i.e., prognosis, cost, age) is privileged in a context of strict refugee policy that restricts independent care-seeking. Most patients are approved for referral, but many are denied. A patient may not be sick enough to merit timely referral, or they may be deemed too sick to prevent referral altogether. Challenging other arguments of therapeutic citizenship where people living with certain health conditions acquire access to certain therapies and their right to health, I show how refugees with certain pathologies are actually excluded from care. These therapeutic exceptions produce consequences whereby bureaucracy and security have prevailed over the right to health and reinforced the power of state sovereignty in what may best be understood not as a therapeutic citizenship, but instead as a therapeutic refugeehood.

INTRODUCTION

In January 2020, a Burundian refugee whom I will refer to as Bernadette, presented to a refugee camp health center in western Tanzania. Something was wrong with her belly.

As a visiting clinician-researcher, I did not meet her until almost three weeks later. When I did, I learned that she had not been passing gas or having any bowel movements, all signs consistent with an intestinal blockage. She looked distressed, frail, and sick. Even though I did not speak her language (Kirundi), her story and body language were communicative enough: the question was not whether she needed an operation—she did. The question was one of when and where? There were several factors at play here, and the politics of referral was a major one. By politics here, I refer to the systems of power in a hierarchical health care system of referral, the unknown potential outcomes after a big operation, and my own identity as a visiting physician in this highly regulated space, the refugee camp.

First, the politics of referral also played out in the type of health center that was available to provide care for Bernadette inside the refugee camp. In the context of Tanzania’s health care landscape, this non-governmental organization-run refugee health facility is often referred to as a dispensary, but it is technically a “health center,” which is a primary health care facility where care is often limited to uncomplicated births, minor burns, wound care, and basic resuscitation. Health centers are not where complex operations are typically performed. They often have more resources than village level dispensaries, but fewer resources than district or regional level hospitals. While I was confident our team had the skillset to do the operation, I was less confident our team had all the resources for what her clinical course may entail. Operating not only requires personnel but also resources and equipment. She might need to have a breathing tube even after surgery, medications to help maintain her blood pressure, and other interventions requiring specialized equipment. The operation could be complex and if any complication occurred, it might be fatal. What is more, I was a visitor in this camp and only there for a short time. I would not necessarily be present to help manage a complication.

Second, there were questions of whether this specific humanitarian organization was going to remain the primary health servicer in the camp in the coming months. The United Nations High Commissioner for Refugees (UNHCR), in conjunction with the Tanzania Ministry of Home Affairs, negotiate which organizations work in the camp. During the time of this visit, there were discussions if a separate non-governmental organization (NGO) was going to take over as primary medical servicer in the camp. I was concerned whether a complication from surgery would muddy those waters and reflect negatively on this organization, because a negative outcome is a possibility of any big operation, which is what Bernadette needed. Despite her being a patient, Bernadette now was a risk in that a bad outcome for her could have deleterious effects on the NGO’s health mission or even my own continued volunteer participation with them. Finally, at this point in the year, the UNHCR had yet to release the referral budget, and essentially all referrals were being delayed as a result.

There is an adage in surgery that states one should never “let the sun rise and set on a bowel obstruction.” By Thursday of that week, twenty suns had risen and set since Bernadette first came to the hospital. There were layers of challenges to providing care for Bernadette including lack of resources and a complex political and social milieu. Perhaps the greatest challenge of this case was referring Bernadette to another hospital—a process I discuss in detail in the article below—because legally, as a refugee in Tanzania, Bernadette was not allowed to leave the refugee camp on her own accord to seek care.

In this article, I provide an ethnographic account of the referral process in a Tanzanian refugee camp. This is in line with a growing movement of critically analyzing processes, power, inclusion and exclusion in health programs and health care access (Boeger, 2020; Nguyen, 2013; Petryna, 2013; Strong, 2020). To build on the body of literature at the nexus of biopower, therapeutic citizenship, and refugee studies, I draw on ethnographic fieldwork conducted in a refugee setting in Kigoma, Tanzania, using participant observation, interviews, and clinical record review to better understand the referral health care system. By approaching referral health care through anthropological theories of biopower and therapeutic citizenship, unique questions arise at the intersection of referral health care and refugeehood—that is, the state of being a refugee. How do global guidelines on referral health care play out in the local context of a country like Tanzania? How do health care providers navigate this complex space and decide who gets referred and who does not? Selecting patients for referral is as much, if not more, political than it is medical. This triaging takes on nuanced forms whereby a certain aspect of a patient’s case (age, prognosis, where they are along a spectrum of treatment) is privileged in a context of refugee policy that restricts freedom of movement for refugees and prevents independent care-seeking. This prioritization—or triaging—of some diseases or patients in is what I refer to as therapeutic privilege. This therapeutic privilege exists along a spectrum of diseases, diagnoses, treatments, and outcomes. In the end, most patients are accepted for referral, but many are denied. By prioritizing certain diseases or treatment modalities, some patients are inherently included while others are excluded because they are too sick or too far along in the natural history of their disease. In other classical arguments of therapeutic citizenship, patients receive health care based on having a particular disease (e.g. radiation sickness, HIV/AIDS) and in doing so can realize their right to health care in a unique way. On the contrary, in this context refugees with certain pathologies are actually excluded from care. These patients are excluded from care in a logic that on a macropolitical level favors state sovereignty over the right to health care and on a micropolitical level does not allow refugees to independently seek care. Referral health care for refugees in Tanzania becomes a window into the larger politics of refugee hosting, and what emerges is a phenomenon whereby certain pathologies of illness create not therapeutic citizens, but instead therapeutic refugees—some of whom are left waiting and some of whom die.

BACKGROUND, SETTING AND HISTORICAL CONTEXT

Ecology of Referral

National healthcare schemes utilize systems of referral to manage large burdens of disease. A large body of literature explores the general concept of referral (Font et al., 2002; Mwabu, 1989). Importantly, the World Health Organization defines referral as “a process in which a health worker at…one level of the health system, having insufficient resources (drugs, equipment, skills) to manage a clinical condition, seeks the assistance of a better or differently resourced facility at the same or higher level to assist in, or take over the management of, the client’s case” (WHO, 2010, p. 2). An example of the referral process is when health providers at a health center refer cancer patients to higher level care facilities that are equipped with resources such as chemotherapy. Many of these specialized facilities in Tanzania are located in major cities, such as Dar es Salaam. Referral health care is an important aspect of a national health system because a well-designed referral healthcare system will allow for optimization of patient care at each individual level (e.g. primary, secondary) while simultaneously referring patients to higher levels of care when needed without overwhelming such higher levels (Simba et al., 2008).

The ecology of healthcare in Tanzania can best be understood geographically. In Tanzania, the health care system is divided into multiple levels, such as health dispensaries, health centers, district hospitals, regional hospitals, and zonal or national hospitals, with varying services available at each level of the health care system (Jumbam et al., 2020). If a patient requires treatment for a condition that a health center or dispensary cannot treat, he or she may be referred to a district level hospital or higher level of care. All higher-level health facilities (district, regional, zonal, national hospitals) theoretically offer all services provided at their respective lower-level centers. In addition to being referred to a higher level of care, some patients may self-refer and present directly to a district, regional or national hospital (Simba et al., 2008).

The Refugee Camp

Uniquely, the main health facility in this refugee camp is classified formally as a health center, despite its staff and services mirroring that of at least a district level hospital (e.g., laboratory services, hernia repair, laparotomy, tubal ligation). In this specific camp, the main facility has over 100 beds and is supported by other health posts in the camp at large, which together have over 150 inpatient beds. Outpatient consultations number over 25,000 monthly (TRCS, 2020). Between 2000 and 2020, there were over 10,000 operations performed in this refugee camp in Tanzania.

The population is approximately 130,000 refugees from the Democratic Republic of Congo and Burundi, and it is the largest and oldest refugee camp in Tanzania (UNHCR, 2020). Established in 1996, it has been in continuous existence and is classified as a protracted refugee situation (Milner & Loescher, 2011). The UNHCR in conjunction with its operational and implementation partners provides general services for the camp population, including health, water and sanitation services, counseling, and food distribution. The government of Tanzania is responsible for overseeing administrative activities and determines both national refugee policy and provides permits for those who wish to leave the camp for both medical and non-medical reasons. The health care services are provided free of charge to both refugees and Tanzanians primarily by a humanitarian organization that has been working in the refugee camp for over 15 years and works in partnership with the UNHCR and other entities. To deliver patient care, the organization primarily employs Tanzania staff, including doctors and nurses, many of whom are young or live far from their families for only a modest salary. For patients that cannot be provided services at the camp level, they are referred elsewhere per a set of international and local guidelines, and such referral health care is paid for by the UNHCR and partner humanitarian organizations and remains free for the refugee patient. These guidelines, though, are subject to local policies on refugee movement. Tanzania’s history of hosting refugees dates back decades, and its policies have evolved over time as it transitioned from a socialist government at the time of independence to a neoliberal one in the 1980s and the present day.

Evolution of Refugee Policy in Tanzania and the Humanitarian Response

Tanzania has long been viewed by the international community as a haven for refugees (Morel, 2009). During its peri-independence era, Tanzania’s socialist practices guided its refugee hosting practices (Eriksson et al., 1981). President Julius Nyerere was focused on a pan-Africanist vision of unity and developing Tanzania as an agricultural powerhouse. Thus, he incorporated refugees into his larger nation-state building plan and provided them with ample land and opportunity (Malkki, 1995; Tague, 2019). Scholars refer to Tanzania’s Ujamaa (socialist) era practices as ‘Open Door,’ which allowed for more freedoms for refugees, including freedom of movement (Chaulia, 2003; Tague, 2019). Ujamaa practices set the stage for a humanitarian response to decades worth of conflict in the region.

Yet, as Tanzania’s economy was failing in the 1980s and the imposition of the World Bank and International Monetary Fund programs were being forced on reluctant African governments, so, too, did Tanzania transition away from socialism and into post-socialist and neoliberal policies (Kweka, 2007). The increased focus on neoliberalism shifted resources away from state-sponsored government and social support programming. Health care metrics deteriorated, and new refugee policies and practices had an increasing focus on security (Ahluwalia & Zegeye, 2002; Chachage, 1993; Turshen, 1999). In Tanzania’s current practices, refugees are not allowed to leave the camp without formal approval or else be faced with fines or imprisonment. Unlike Lebanon where Syrian refugees may live, work, and more independently seek health care, in Tanzania local integration is not an option in the 2003 National Refugee Policy (Doocy et al., 2016). Recent practices by Tanzania suggest an increased focus on security. Examples include the forced repatriation of Rwandan refugees in the 1990s, a withdrawal of its participation as a pilot site in the Comprehensive Refugee Response Framework, and reports of forcibly returning Mozambican refugees (Fellesson, 2019; Schlein, 2021; Whitaker, 2002).

A larger examination of humanitarianism in east Africa points to NGOs in conjunction with other transnational institutions (e.g. United Nations) and state governments as key actors in the delivery of social services including health care. Examples include longstanding refugee camps and humanitarian missions in Tanzania, Kenya, Malawi, Uganda and several other countries. While this transnationalism humanitarianism—broadly defined as “an ethos, a cluster of sentiments, a set of laws, a moral imperative to intervene, and a form of government” (Ticktin, 2014, p. 274)—has existed for many years, the current geopolitical milieu in Tanzania places a larger focus on security and control. Here, anthropological theories of biopower and therapeutic citizenship offer a useful backdrop to understand how such a focus on security may be relevant to health.

Biopower and Therapeutic Citizenship

Classically, biopower is a theory coined by Michael Foucault where he sought “to explain how biological and medical data are used by institutions of the modern world to define, count, divide, and—in a word—discipline populations” (Farmer et al., 2013, p. 26). In doing so, Foucault explored how “institutions constricted norms about what is considered…healthy and sick” and “the right of death and power over life” (Foucault, 1978, p. 133). Rabinow and Rose have argued that contemporary biopower “entails a relation between ‘letting die’ (laissez mourir) and making live (faire vivre)—that is to say strategies for the governing of life” (Rabinow & Rose, 2006, p. 195). Therapeutic citizenship has emerged as a form of biopolitics where institutions have a “power over life” and can decide who lives and who dies. Generally, patients make claims to an institution (e.g. government, NGOs) on their right to health not because of their status as a citizen of the state, but rather because of their diagnosis of a disease (e.g., HIV/AIDS, radiation sickness). In Petryna’s Life Exposed, individuals suffering from radiation sickness made claims on the state for health benefits (Petryna, 2013). In Nguyen’s The Republic of Therapy, therapeutic citizenship emerged in the absence of the “state” when individuals partook in clinical trials and provided testimonies for NGOs in order to gain access to therapy (Nguyen, 2013, p. 109). In other words, biological or therapeutic citizenship may be a sense of entitlement on the part of a patient to access certain health care and therapies. But, therapeutic citizenship may also be understood as the phenomenon that some patients are being granted access to therapies by a state or NGO—or in the refugee context, a transnational institution—based on the presence of a particular pathology. The presence of a particular health condition privileges, in some cases, this access to therapies, and in some ways serves as a form of triage.

Triage and Therapeutic Privilege

Triage is a dynamic concept. It usually refers to the “sorting” of individuals or as Nguyen (2010) has described, “an operation that differentiates people into groups based on specific criteria, such as those who require immediate medical attention from those who do not;” it selects “those who would receive the treatments and those who would not” (Nguyen, 2013, p. 109; Robertson-Steel, 2006, p. 154). In the process, value is placed on particular disease processes and thus on life itself, and in this process certain therapies become privileged.

Historically, medical ethicists have used the term therapeutic privilege in a clinical situation in which a doctor may choose to withhold information about a patient’s diagnosis for fear that such information if disclosed to the patient would be detrimental to their health. In this privileging, there is a balance to be had between the ethical principles of autonomy, beneficence, and justice. A parallel to this balance may be a useful way with which to approach referral health care for refugees in Tanzania, as often, in contrast to individual approach to the patient observed in the therapeutic privileging in the clinical setting, in humanitarian settings, population level decisions are often the focus.

With a focus on population level health, the referral process in Tanzania is complicated with multiple actors. While literature has explored Tanzania’s refugee policies evolution in the post-socialist era, key gaps remain on how said policies might affect health and the therapeutic landscape for decision-making regarding which refugees are referred for care and which ones are denied referrals. In this article, I draw on anthropological theories and ethnographic methods to better understand the referral process for refugees in Tanzania and the ways in which the health conditions and decisions made about which patients get referred become a window into the larger politics of hosting refugees in Tanzania. I focus on non-emergency referral cases, although many of the referrals needed may still be time sensitive, like that of Bernadette. In this process, the social and political aspects of referral are privileged as much, if not more, than the medical aspects, and the decisions made may exclude patients with certain illnesses from accessing care in what might best be appreciated not as a therapeutic citizenship, but instead as a therapeutic refugeehood, especially for those refugee patients who are either too sick or not sick enough to be referred.

METHODS

Data Collection

Methods included interviews, participant observation, and clinical record review. The work is based on approximately six months of fieldwork in Tanzania that I have conducted at various times over the last decade (2011–2021). Much of this fieldwork involved participant observation and interviews—core aspects of ethnography—and detailed record review. Interviews and/or meetings were held with health care staff and others (clinical officers, nurses, medical doctors, and academicians, n=11) identified because of their observed role as health care team members or experts in referral or refugee studies. Most interviews were audio recorded, but non-recorded interviews/meetings took place. Interviews were conducted in Kiswahili or English. Participant observation occurred primarily in the operating theatre/suite, outpatient clinic, morning report, attendance at a medical referral committee meeting, and even personally escorting one patient on a referral to a district hospital. These settings provided rich context on understanding how referral and decisions about referral function. This work also involved review of over 200 patient clinical referral records. Two documents were considered more formally for analysis: UNHCR’s Principles and Guidance for Referral Health Care for Refugees and Other Persons of Concern and the “Standard Operation Procedures for Medical Referral System: Refugee Operating for Northwestern Tanzania and urban Refugees, 2021–2022” (SOP, 2021; UNHCR, 2009)

Data Analysis

Analysis primarily took the form of memoing and coding with the support of Nvivo for data management. Analysis was inductive and deductive based on the aforementioned theoretical frameworks (Charmaz, 2006) in that I approached the themes with a focus on the anthropological theories presented here while also being open to other themes that arose during the research process about how referral works. Some, not all, clinical records were coded in Nvivo based on symptoms and diagnoses. Data analysis was performed solely by the author, but translations of interviews were performed by the author and a research assistant. Names used in this article are pseudonyms and any identifying information has been altered for anonymity purposes. All words used in quotation are either the exact words by a participant or my best attempt at recording them. Field notes were typically taken in short-hand and expanded upon at the conclusion of the day’s events. Memos were written during fieldwork and outside of fieldwork as well during the analysis phase.

Positionality and Ethical Clearance

I am an American-born dual citizen of the United States and Nigeria. I am a licensed physician in the United States and Tanzania. Under the umbrella of the NGO running the health mission, I participated as a volunteer clinician and surgeon in caring for patients during my fieldwork, including seeing patients in clinic, operating, and recommending patients for referral. I have volunteered with this organization several times over the last five years. By my nature of being present in the main hospital or operating room and being fluent in Kiswahili, I was privy to conversations about surgical care, treatment plans, and referral. As a visiting doctor, I referred multiple patients like Bernadette to higher levels of care. While I did feel an ethical obligation to treat the patient in front of me, including referring patients, I recognize such an obligation cannot be fully divorced from my interests in studying the social science of referral health care, specifically how my positionality influenced what clinical conditions I thought merit referral and the type of information and spaces I did and did not have access to, as well as my interpretation of what I observed. My understanding of which cases might be appropriate for management in the refugee camp or at various levels of the health care system must be understood with an appreciation for my status as a visitor with limited knowledge compared to my local collaborators. In navigating these two identities (surgeon and researcher), I relied heavily on local expertise and guidance from collaborators, as well as self-reflection.

Ethical approval was obtained from the Johns Hopkins Medicine Institutional Review Board (IRB00258009) and the Tanzanian Commission on Science and Technology (2020-391-NA-2011-143). A permit to enter the refugee camp was obtained from the Ministry of Home Affairs of Tanzania.

RESULTS

Several key findings emerged including: 1) medical and surgical referral was a complex process; 2) several factors contributed to the decision of whether a patient should be referred, including age, prognosis, previous admissions, and accompanying paperwork and 3) triage occurred at several points along a spectrum of diagnoses, treatments, prognoses and potential outcomes.

The Complexity of Referrals and Case Examples

The referral process in Tanzanian refugee camps is defined in the Standard Operating Procedures (SOP) as “a clinical decision that shall be taken by a medical expert” and “based on the prognosis, funding availability and the availability of services” (SOP, 2021, p. 1). In the refugee camp, the process for referral involves a treating physician examining a patient. If deemed appropriate in the judgment of the physician, the physician fills out a formal referral form (see Figure-1 and Box-1 as examples), and then a case is brought before a Medical Referral Committee (MRC) for discussion. The decision is made on both objective criteria and subjective criteria as I outline in this section. Per the local Tanzanian SOP, the MRC is comprised of multiple committee members representing different government and NGO agencies, health partners, and the UNHCR. The referral committee meetings occur usually monthly, last several hours, and can involve discussing easily over 150 patients. Some patient discussions are less than a minute. Others may last several minutes. Most patients are approved, but it was not uncommon for a patient to be deferred or denied referral (see Case 2 below). In a single month, over 150 referrals may be presented or approved, and approximately 9 out of every 10 patients gets approved. Even after approval, refugees may experience delays. Cases from all three camps in Tanzania are presented at MRC meetings. If the decision is made to refer a patient, a date and transportation are organized to send the patient. Referrals may be made directly to a district hospital, regional hospital, zonal hospital, or national hospital. Intercamp referrals are also a possibility. Any referral involving a patient already outside of the camp (e.g., already referred and admitted at another hospital) requires approval for an additional referral.

Figure 1-.

Medical Referral Form

BOX-1: Example of Approved Referral Case.

Abbreviations: HHD = Hypertensive heart disease; BMC = Bugando Medical Centre

As an example, the following are two detailed, but typical examples of cases presented—one approved and one denied (see also Box-1 and Box-2, respectively).

BOX-2: Example of Denied Referral Case.

Abbreviations: BMC = Bugando Medical Centre

Case 1

Bahati was a 46-year-old male with known hypertension that subsequently led to heart disease. He was previously sent to a zonal hospital called Bugando Medical Centre (BMC)—the premiere health institution in western Tanzania—in the previous year where he was admitted for 26 days with a diagnosis of stroke. He received numerous blood tests and imaging of the brain at the time. Now, he was being again presented before the referral committee because he was experiencing symptoms consistent with heart disease and a stroke. During the referral committee meeting, there was a long discussion about what the previous recommendations were on the discharge summary and questions about what occurred during his previous admission. At one point, Dr. Joyce, the presenting doctor, recommended he be sent to a different hospital apart from BMC, but the final decision was to send him back to the hospital that he was attended to previously (BMC). In fact, his case could be classified as a heart disease case, and it was the fourth heart disease case in a row during that meeting to be presented and approved to go to BMC—perhaps contributing to the positive referral decision. Additionally, there is no equipment or technology available to confirm worsening heart disease or a stroke at the camp level health center.

Case 2

Kiza, an older 74-year-old man with back pain and a bulging disc was presented next (Box-2). Kiza was also previously seen at BMC in just the previous month for the same back problems and given a follow up appointment for the following month. A senior referral committee member commented that for follow-up cases, “there is not much discussion. We just send.” This led to one of the longer discussions of the meeting about re-examining the need to send the follow-up cases because as a different committee member noted, “there are currently many, and we (the committee members) have to consider the budget. If it’s follow-up for procedure [fine], but if it’s not [a] procedure, we really have to consider them.” The remainder of the discussion for his case questioned what would be expected for a repeat visit to BMC because it wasn’t the “first” or “second” time he was being sent to BMC. Was he simply being sent to be “given medication and be sent back?” The conversation shifted, and the senior referral committee member asked “What assessment was done by the camp doctors? Do we really need to send the patient back or can we manage?” A third member added that there was a discharge summary provided that stated there were no further treatment recommendations. Finally, the discussion concluded inquiring as to whether the medications prescribed for the patient were even available in the camp and whether the prescribed medication combination was adequate. The committee decided Kiza’s problem could be managed at the camp level, and Kiza was not referred.

Referral is a bureaucratic process. For patients like Bahati, the Ministry of Home Affairs of Tanzania must approve any referral for a refugee to leave the camp. Movement in and out of the camp is highly regulated, and even in an informal conversation with one NGO driver, he stated “everything that goes into the camp…[even] stationary” has to be regulated. The space of the refugee in Tanzania’s larger immigration policies is restrictive in that refugees must live in the designated areas by the government (Kamanga, 2005). On paper and even in practice sometimes, these processes can be highly regulated and bring organization and support to a specific population. While these processes can be regulated at some level, it often led to significant delays in care up to several months for patients, and sometimes this was even to get a diagnosis. The bureaucratic process of referral in some ways functioned as a rate limiting step in advancing a patient’s care to the more appropriate level. Since patients are not allowed to legally leave the camp on their own accord, this also created significant delays in accessing health care. In other words, refugee patients lacked the legal rights of citizens to freely seek care in health centers outside of the refugee camp without governmental approval, and the referral process may deem their health condition not important enough for prompt medical attention.

Factors Affecting Referral Decisions

Several factors affected referral decisions including prognosis, limited resources, age of the patient, previous admissions, and accompanying paperwork.

Prognosis and Natural History of Disease

Prognosis played an important role in determining if a patient would get referred. This took three primary forms: 1) the patient’s current health status including presence of other medical conditions; 2) likelihood of improvement; and 3) patient’s current treatment plan and where they are within said treatment plan. Both the UNHCR guidelines and local SOPs also written by the UNHCR state that “cases should not be referred for health care if their health condition has reached such an advanced stage of deterioration that survival or recovery is unlikely (e.g., cases of advanced renal failure, far advanced malignant cancers).” In other words, someone may be too sick by the MRC committee to be deemed appropriate for referral.

Dr. Abdul described the effect of co-existing medical problems on a patient’s likelihood of being referred. “For example, [if] the patient has more than 4 diagnoses (diabetes, hypertension…with HIV) … they used to say they have a poor prognosis, and it has to be managed in the camp.” Another clinical officer in the camp, Dr. Robert, confirmed this: “people whose prognosis are not good’ do not get referred. Dr. Robert’s comment was commonplace in conversations around referral. Prognosis, though, seemed to be open to some level of interpretation. For example, Dr. Robert retold a story of a child he cared for with hearing loss who was initially denied a referral but after some convincing and efforts on Dr. Robert’s part, he was granted one.

“There was a child who was deaf. Later, they said that even if we take that child anywhere, he won’t have any benefit. But, later the child was given a referral. He went and got help, and now the child hears. Now, if they had that same mentality, the child would not have been helped, so I see it as any man can be helped.”

In this case, Dr. Robert thought the child should have a referral based on a similar patient he had cared for previously. In this case, there was a disagreement about the patient’s prognosis that influenced both whether the patient was given a referral initially and the timing of the referral. But sometimes, establishing a diagnosis requires equipment and technology not available in the camp—which in other circumstances would aid in determining a prognosis. According to Dr. Nia, a doctor working in the camp for a year and a half when we spoke, establishing a diagnosis may take up to “three months.” If a diagnosis is needed to make a referral more urgent or appropriate, this presented an additional problem especially if the referring doctor did not think there was anything to be done for a patient. As Dr. Robert remarked, “[i]f we see that we have the ideology that we are the last resolution but even the equipment we do not have, we will hurt the sick.” If no benefit was seen, Dr. Robert noted this was only hurting those who were already sick.

Despite the goal of objectivity, there are inherently ways in which individual subjectivity of a person in a position of power came to light in the process, such as during the MRC meeting (Willen, 2012). The committee meeting itself is unique in that it brought together an expert panel described above “to take objective technical decisions” while also promoting distance between the treating physician and referral decision (SOP, 2021, p. 1). On paper, health professionals with direct contact with patients are asked to present the patients and not participate in decision-making. In practice, the presenting physicians did more than just present a patient’s history and hospital course and remain silent, but they often participated in the ensuing discussion. The providers themselves cannot be seen as separate from even the moral obligations they feel as clinicians or individuals working in this context (Malkki, 2015). There were implicit and explicit ways in which clinicians who participated in the discussion participated in a politics of “life itself” and brought their own worldviews and moral compass into the triaging of patients (Nguyen, 2013, p. 89; Ticktin, 2011). An example of this is the way the age played a role in discussion for younger patients, such as the case of a young woman patient with an arm injury. There was a palpable desire to help a “young” patient who had documented—subjective as it may be—evidence that there was no further treatment available for her. At the same time, for the deaf child described by Dr. Robert, at first the referral was denied because the referring physician did not see a clear benefit of the child being sent to a tertiary care facility. It was only after the advice and persuasion of Dr. Robert that the child was referred.

Natural History of Disease and Treatment

In the case of Kiza, he had already been referred and was subsequently denied a referral because there was no certainty that his outcome would be different if referred again. He presented on the latter end of a spectrum of treatments already utilized, and the ability to get a second opinion through the referral process from a different hospital was almost non-existent. Part of this was related to the limited resources available in general in the Tanzanian health care system. Another part was because this patient had already been to this particular hospital. There was no space for a second opinion. On a spectrum of severity, his main complaint was pain, and this was recommended to be managed at the camp level, despite him having an appointment to return to the higher-level care center. In other words, his persistent problem was not severe enough to merit a referral.

Limited Resources

One of the most common themes that arose in conversations and observation was a lack of resources, including diagnostic testing, transportation, specialist services, and the overall budget for referrals (transportation and cost). For example, in recounting a story of a patient who was having difficulty breathing and who needed oxygen supplementation, Dr. Abdul said:

“Currently we don’t have an ambulance with well-equipped facilities, like oxygen cylinders or portable oxygen machines. For example…[a] patient was desaturating (had low oxygen levels). So, we had to wait for an ambulance from another district to come and fetch the patient. Unfortunately…when the car arrived, the patient had already died.”

Here, a patient needed a referral but there was no transportation available. Sometimes, this happened with emergency cases, such as in the case described above. Other emergencies, as Dr. Robert noted, like intestinal obstructions, which is what Bernadette had, don’t just take “one hour” or “two hours,’ but instead will “usually take six hours or one day” for someone to get referred. Patients have died while waiting to be referred.

Another major theme that emerged was the budget. First, this involved the overall cost of a particular referral (transportation to the higher-level hospital and any included costs of further management). Second, there was an intent to maximize the number but also consider the type of referrals (e.g., pathology). The number of cases referred was a function of the budget available. There were more cases that needed referral than what can be afforded. For example, for this refugee camp, there were approximately 130 cases to present during a single MRC meeting. Ninety of these cases were “eye” cases for which upon hearing the number, one committee member remarked, “If we send all to Kabanga [a district hospital] …we will have used up the referral budget.” The estimate to send all 90 would cost about 60 million Tanzanian shillings (equivalent to $25,800). In response, all ninety cases were deferred to be discussed later and the recommendation was that the surgical ones should be selected. None were individually decided upon at that meeting. Treating eye cases may be relatively cheap compared to treating cancer, for example, but there seemed to be a desire to diversify the cases referred.

Age

Age played a role in the decision-making in three ways. This included what the diagnosis may be, what the available treatment options are, and as a moral imperative. First, since many imaging services are not available at the camp level, many patients were referred for additional investigations to determine a diagnosis. For example, a patient named Patrick was an elderly 76-year-old man who underwent urethroplasty—a procedure to fix a defect in the channel connecting the bladder to the penis. He had a urinary catheter in place that was draining foul smelling pus, likely indicating an infection. In discussing his case, one referral committee member commented that he is “worried he might develop cancer with this age” and thus the patient should be referred. In this case, the older age of the patient came into the decision-making as part of what his diagnosis might be. It might be cancer and thus a referral was favored. Prognosis requires a diagnosis and here age influenced the prognosis and favored a referral.

Secondly, the age of a patient influenced what treatment options were available. For example, three cases presented during a particular MRC meeting were patients with cryptorchidism—an undescended testicle. In these cases, the age of the patient was important because for the two-year-old child, corrective treatment was still an option, but for two other patients who were in their 30s already, the procedure would not help with fertility but could help to prevent cancer. Here, age was important because it dictated the treatment options available to the patient and thus influenced the referral decision. In the above scenario, all three cases were approved for referral.

Lastly, age was also mentioned as a reason to support patients as a moral imperative. An example of this was the case of a young, teenage girl who was injured in 2015 in Burundi and has had a persistent arm injury with worsening deformity. The treating physician stated that she should be sent to BMC to be helped “because she is only 16.” In this case, the patient had been sent to a local hospital previously where it was advised that she live with her deformity and not undergo surgery. The physician presenting her case was arguing that her condition continued to worsen, but the previous paperwork indicated that surgery may only make things worse. In the end, the committee decided to reach out to BMC through back channels to get feedback in advance, prior to referring the patient. They noted, though, not to send the previous discharge summary that indicated the other hospital’s plan of (in)action (i.e. not undergoing surgery).

DISCUSSION

A close examination of referral from the perspective of those involved highlights that global policies are enacted in unique ways based on the local (Tanzanian) social and geopolitical context (Milner, 2014). My focus in this paper is not simply to highlight that difficult decisions and uneasy compromises abound in the context of limited resources—for this has been shown across many settings including in Tanzania (Emanuel et al., 2020; Fassin, 2012; Strong, 2020). Rather, my goal is to argue that patients with certain diagnoses did not have improved access to health, but instead were actually excluded from care in what can best be understood not as therapeutic citizenship but instead as therapeutic refugeehood. A particular therapeutic ecology—or therapeutic landscape—exists in the context of refugeehood in Tanzania (Gesler, 2018; Langwick, 2008, 2015). This therapeutic ecology is at once both a product of a restrictive geopolitical context and a perpetuator of significant limitations in the realization of one’s right to health care. It is this ecology (restrictions on movement, lack of resources, lack of finances) that privileges certain therapies over others and allows physicians to enact biopower by favoring patients along a particular point in a spectrum of diagnoses, severities of illness, treatments, and outcomes. This ecology allows for physicians or MRC members to decide that age is the critical aspect of someone’s case to encourage a referral or at the same time decide that pain is not deserving enough to necessitate one—all while patients cannot independently seek a second opinion. In this context, refugee patients in western Tanzania inhabit a unique problem space whereby one may gain access to a diagnosis—if at all—but still be subsequently excluded from actually receiving care in a timely manner, all in accordance with a physician or referral committee-deemed prognosis. Here, the frontline health care workers functioned as “street level bureaucrats,” deciding who gets referred and who does not in a system that privileges certain aspects of a patient’s case or prognosis (Lipsky, 2010).

Challenging other arguments of therapeutic citizenship where people with certain health conditions are granted access to therapy, in this refugee setting certain pathologies actually led to a therapeutic refugeehood where refugees were excluded from care based on their diagnosis or prognosis. While there is some similarity to other contexts, such as the work of Miriam Ticktin, who writes that “sick…bodies [can] cross borders while impoverished ones cannot” (Ticktin, 2011, p. 3), in this case, sometimes even the sick bodies were not able to “cross borders.” This created a unique space for refugee patients who need access to care not available at the camp level, especially in this restrictive geopolitical context. This is not to say that deciding which cases to send to higher levels of care is divorced from medical knowledge and expertise—quite the contrary, and the MRC meeting even became a site of teaching and knowledge exchange. But, in the context of restricted freedom of movement for refugees, medical referral was never simply a “clinical decision” that is “based on prognosis, funding availability and the availability of services” (SOP, 2021, p. 1). It was equally social, political, and subjective in the space of the refugee camp. Decisions made often became “arbitrary and contingent”—in other words, subjective (Fassin, 2012, p. 72). The arbitrary is seen in the sense that the same person may evaluate a patient and deem them a good candidate for referral. Someone else may not. The contingent is multifactorial: contingent on the history of present illness, contingent on resources and budgets available at that time, and contingent on whether a patient is a refugee or Tanzanian because Tanzanians are allowed to move in and out of the camp freely for health services, including independent care-seeking to other hospitals.

Triage, then, is inherently linked to a patient’s circumstance and that of the referral committee, all the while influenced by the geopolitical context within which refugee patients in Tanzania exist. Applying broadly the way triage has been explored by others (Fassin, 2012; Redfield, 2013) to this refugee context highlights what I argue is a related medical concept that has important implications for triage in this context: therapeutic privilege. I offer the concept of therapeutic privilege—where one ethical principal (e.g. patient autonomy) is sacrificed for another (e.g. beneficence)—as a parallel to the humanitarian reasoning that occurs within this context. In the refugee setting, decisions about care in general are often considered on a population level rather than an individual one. Whereas many debates about therapeutic privilege concern an individual patient and their access to knowledge of a disease, in the refugee camp we see it on a different scale and with respect to a population as opposed to an individual. Therapeutic privilege takes on two distinct forms in this context: one is practical, one is ontological.

Practically, therapeutic privilege refers to the actual prioritization—or privileging—of certain therapies or prognoses over others. Referral was based on the factors such as age, prognosis, previous documentation, and other elements of bureaucracy. Because of these factors, certain therapies were privileged to promote greater access to health care based on cost, all while creating therapeutic refugees whose conditions render them not referrable. More serious, individual cases may consume a large proportion of the budget and thus a decision was deferred. Similarly, the 90 “eye cases” presented in a single meeting would consume all the budget before other patients or pathologies are presented. The premise is that one does not want to spend too much money on a single case—or even single type of case. Therapeutic privilege was not just a balancing of ethical and moral principles in choosing a specific therapy within a distributive justice space. Therapeutic privilege—or prognostic privilege—also related to the ways in which patients at a particular point along the spectrum of diagnoses, disease severities, and treatment plans were prioritized over others. This therapeutic or prognostic privileging produced therapeutic refugees—those whose diagnoses and prognoses created significant delays in their care or excluded them from care altogether.

In the context of limited resources and the refugee setting, the global and local policies suggest that the sickest patients should not be referred to health care outside of the camp (SOP, 2021; UNHCR, 2009). The premise is that these resources would be better utilized if a patient was less sick. Ethically, treating one sick patient is sacrificed in exchange for treating another (or multiple), less sick patient(s). In places such as France, Fassin and Ticktin have demonstrated how a “logic of extremity” and “unusual pathologies” permitted the granting of state-sponsored benefits or a resident and work permit to those individuals who are deemed so sick they need even more support (Fassin, 2012, p. 91; Ticktin, 2011, p. 4). Similarly, in the United States, organs donations are often allocated to the sickest patients at the top of the organ recipient list (Liu et al., 2021; Schaffer et al., 2003; Van Meter, 1999). In this refugee context, the opposite may be said of how the UNHCR guidelines played out on the local level. This challenges other narratives of therapeutic citizenship where much needed health care is provided to individuals on the basis of having a particular disease or severity of disease (Leclerc-Madlala, 2009; Mattes, 2011; Nguyen, 2013; Nguyen et al., 2007; Petryna, 2013). In this context, the sickest patients are not moved to the top of the referral list—quite the contrary, they are excluded from referrals or the right to health care in what can best be understood not as a therapeutic citizenship, but instead as a therapeutic refugeehood.

CONCLUSION

We decided against operating on Bernadette in the refugee camp for all the reasons outlined in this article (lack of resources, my positionality as a visiting surgeon under the umbrella of an NGO, her prognosis being appropriate for referral). From what I was told, after several weeks of waiting, Bernadette was indeed referred to a district level hospital. There, it was deemed they did not have the resources available to treat her. She was referred to the regional level hospital. There, it was also deemed that they did not have the resources available to treat her. She died while waiting to be referred to the zonal level hospital.

Referral for refugees in Tanzania is as much a political and social process as it is a medical or surgical one. A certain aspect of a patient’s case (prognosis, cost, age, where they are along a spectrum of treatment or the natural history of disease) allowed for most patients to receive referrals, others to be deferred, and many to be denied. A patient may not be sick enough to merit urgent, emergent, or timely transfer, or they may be deemed too sick to prevent referral altogether. Therapy was privileged at a particular point along a spectrum of diseases and prognoses. These therapeutic exceptions produced consequences whereby bureaucracy and security prevailed over the right to health care, illuminated the paradox of poor prognoses in this setting, and reinforced the power of state sovereignty seemingly over all else.

Research Highlights.

• Refugees in Tanzania are subject to a forced encampment policy

• Certain health conditions seemingly exclude refugees from accessing referrals

• A patient may not be sick enough to merit timely referral

• A patient may be deemed too sick to prevent referral altogether

• Referral for refugees is as much a political problem as it is a medical one