Article Text

Download PDFPDF

Representations of an Ebola ‘outbreak’ through Story Technologies
  1. Raphael Frankfurter1,
  2. Maya Malik2,
  3. Sahr David Kpakiwa3,
  4. Timothy McGinnis4,
  5. Momin M Malik5,
  6. Smit Chitre6,
  7. Mohamed Bailor Barrie7,
  8. Yusupha Dibba8,
  9. Lulwama Mulalu9,
  10. Raquel Baldwinson10,
  11. Mosoka Fallah11,
  12. Ismail Rashid12,
  13. J Daniel Kelly13,
  14. Eugene T Richardson6,14
  1. 1University of California San Francisco, San Francisco, California, USA
  2. 2School of Social Work, McGill University, Montreal, Québec, Canada
  3. 3Gardener, Koidu, Sierra Leone
  4. 4Pritzker School of Medicine, University of Chicago, Chicago, Illinois, USA
  5. 5Center for Digital Health, Mayo Clinic, Rochester, Minnesota, USA
  6. 6Global Health and Social Medicine, Harvard Medical School, Boston, Massachusetts, USA
  7. 7Partners In Health, Freetown, Sierra Leone
  8. 8Partners in Health, Freetown, Sierra Leone
  9. 9McMaster University Faculty of Health Sciences, Hamilton, Ontario, Canada
  10. 10Department of English Language and Literatures, University of British Columbia, Vancouver, British Columbia, Canada
  11. 11Africa Centres for Disease Control and Prevention, African Union, Addis Ababa, Ethiopia
  12. 12Department of History, Vassar College, Poughkeepsie, New York, USA
  13. 13Epidemiology and Biostatistics, School of Medicine, University of California, San Francisco, San Francisco, California, USA
  14. 14Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Eugene T Richardson; eugene_richardson{at}


Background Attempts to understand biosocial phenomena using scientific methods are often presented as value-neutral and objective; however, when used to reduce the complexity of open systems such as epidemics, these forms of inquiry necessarily entail normative considerations and are therefore fashioned by political worldviews (ideologies). From the standpoint of poststructural theory, the character of these representations is at most limited and partial. In addition, these modes of representation (as stories) do work (as technologies) in the service of, or in resistance to, power.

Methods We focus on a single Ebola case cluster from the 2013–2016 outbreak in West Africa and examine how different disciplinary forms of knowledge production (including outbreak forecasting, active epidemiological surveillance, post-outbreak serosurveys, political economic analyses, and ethnography) function as Story Technologies. We then explore how these technologies are used to curate ‘data,’ analysing the erasures, values, and imperatives evoked by each.

Results We call attention to the instrumental—in addition to the descriptive—role Story Technologies play in ordering contingencies and establishing relationships in the wake of health crises.

Discussion By connecting each type of knowledge production with the systems of power it reinforces or disrupts, we illustrate how Story Technologies do ideological work. These findings encourage research from pluriversal perspectives and advocacy for measures that promote more inclusive modes of knowledge production.

  • epidemiology
  • mathematical modelling

Data availability statement

Data are available upon reasonable request.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • All social phenomena are subject to interpretation; whichever interpretation prevails at a given time is a function of power, not objectivity.

  • Our attempts to understand complexity necessarily entail normative considerations.


  • This study contributes to the sociology of knowledge by examining a specific Ebola case cluster from the 2013–2016 West Africa outbreak and elucidating how various disciplinary approaches—treated as Story Technologies—parse social social phenomena.

  • It underscores the instrumental role Story Technologies play in ordering contingencies and establishing relationships in the wake of health crises.


  • The findings highlight the ideological work performed by Story Technologies and advocate for research embracing pluriversal knowledge ecologies.

The power to narrate, or to block other narratives from forming and emerging, is very important to culture and imperialism, and constitutes one of the main connections between them. —Edward Said, Culture and Imperialism (1993).


The 2013–2016 Ebola outbreak in West Africa was the largest and longest in recorded history, with more than 28 000 reported cases across Guinea, Liberia and Sierra Leone.1 Impoverished health systems, uncoordinated non-governmental organzation (NGO) and government efforts, and distrust of public health mandates all contributed to this being the first time Ebola transmission extended out from more isolated, rural areas and into densely populated urban centres, including the capitals of all three countries.2

Between December 2014 and February 2015, local health officials and US Centers for Disease Control and Prevention (CDC) partners reported evidence of 32 deaths due to Ebola virus disease (EVD) from the village of Bumpeh, Kono District, Sierra Leone. During and in the wake of the regional outbreak, we observed a number of competing narratives that described this ‘case cluster’ (or ‘hotspot’).3–5 In some ways, these narratives corroborated each other; however, they often seemed to be at odds: numbers did not line up, contexts were revealed at different scales, and ‘lessons-learnt’ focused on a variety of determinants. In response to the proliferation of such accounts, the aim of this study was to investigate the following questions: (1) How do the stories that circulate in the wake of infectious disease outbreaks inscribe various forms of order onto understandings of disease transmission and concomitant suffering? (2) What might be learnt from juxtaposing the accounts of this event emanating from various disciplines and knowledge ecologies? (3) What work do these disparate stories do in the world?

Our overall approach posits that different disciplinary approaches to understanding (Ebola) outbreaks are best appreciated as partial curations of data rather than objective accounts of biosocial phenomena. Each curation is necessarily fashioned by, and does ‘work’ in the service of, particular political ideologies.6 In the analysis below, we consider (1) outbreak forecasts derived from mathematical models; (2) epidemiological data gleaned from active surveillance; (3) reconstructions of viral transmission based on post-outbreak antibody surveys; (4) political economic depictions of the outbreak as the embodied manifestation of historical systems of slavery, colonialism, and racial capitalism7; and (5) an ethnographic account of memories of the outbreak that have lingered in Bumpeh as affected people there process what unfolded and rebuild lives in the aftermath of the epidemic.

In relating these accounts in parallel, we argue that each is a type of what Richardson has called a Story Technology.8 Our use of the term Story emphasises the notion that these five approaches to understanding Ebola outbreaks are all essentially forms of narrative—none necessarily more value-free than the other.9 The term Technology is used in the poststructural sense: these stories are more than modes of representation—they do work in the service of, or in resistance to, power.10 Analysing outbreak narratives as Story Technologies, then, enables us to (1) trace what is centred and what is omitted by each approach and (2) appraise the politics encoded, and hierarchies normalised, by these various modes of representing the outbreak.


We investigate different approaches to representing the Bumpeh outbreak by critically examining the methodologies used and conclusions advanced by outbreak forecasting, active epidemiological surveillance, post-outbreak serosurveys, political economic analyses, and ethnographic accounts. To do this, we review the findings of key related publications and also present primary ethnographic data.

The Story Technologies employed by the CDC (sections 1 and 2) rely on the Epi Info Viral Hemorrhagic Fever (VHF) database, which was the primary data management system used for case and contact tracing during the West Africa outbreak.11 The reconstruction of transmission chains (section 3) is derived from a post-outbreak serosurvey conducted by Kelly and colleagues.4 Our employment of political economy as a Story Technology (section 4) relies on a review of related historical and theoretical scholarship. Finally, our primary ethnographic research (section 5) offers a chance to co-construct representations of the outbreak with those who suffered most from it.12 This is the longest of the five sections since it presents original research.

Our overall argument emerges through shared dialogue, exchanges, provocations, and encounters across a large network of authors and interlocutors, constituted by plural identities, positionalities, perspectives, and political positions and through a commitment to elevating forms of knowledge production and representation that are attentive to the experiences of those affected ‘first-and-worst’ by global health crises.

Patient and public involvement

Residents of Kono District were involved in the conduct and dissemination plans of our research. Groups were consulted on their experience with various research teams. Results were discussed with individuals with high network centrality.

Results: Story Technologies

Outbreak forecasting: mathematized ideological systems

During the 2013–2016 Ebola virus (EBOV) outbreak in West Africa, computational modellers proffered an array of predictions,13 ranging from the WHO’s conjecture early on that the outbreak would be contained at a few hundred cases to the CDC’s estimate of up to 1.4 million cases by January 2015.14 The premise of epidemic modelling is that simplified networks of human interactions, simulated through computer calculations, can be used to predict the scale and distribution of viral transmission. Modellers themselves have sought to codify their methodologies as a discipline and set of concepts in order to advance an applied ‘outbreak science’.15 Responders might then be able to ‘get ahead’ of epidemics by implementing control measures (at a macrolevel, in order to prevent cross-border spread, but also at a microlevel, by categorising infected individuals as ‘superspreaders,’ for example, and targeting them for early containment16). Models can be used to forecast the number of human deaths that might ensue as policy-makers weigh multiple options.

But the construction of these models, like the construction of other kinds of stories, requires normative assumptions in order to reduce complexity;17 18 these assumptions often become obscured as models are put out into the world with the trappings of objective science.19 For instance, there is no history, no consideration of what predisposed the region to Ebola beyond a spontaneously arising and untethered epidemic curve—a curve described as providing ‘evidentiary charisma to…recruit moral legitimacy and short-circuit scientific contestation’.20 But there are also more subtle and insidious assumptions built into the actual quantifications entailed in modelling: in the CDC model that forecasted 1.4 million Ebola cases by January 2015, the modellers note that ‘ensuring that approximately 70% of the [Ebola] patients are in ETUs [Ebola Treatment Units] is necessary, or, when ETUs are at capacity, that they are at home or in a community setting such that there is a reduced risk for disease transmission (including safe burial when needed).’ This is not a value-neutral statement: it normalises a containment-over-care paradigm that has existed since colonial times.21 Secondarily, this model lends permission for a startling death-toll. To illustrate: if we accept the projection that there would be 1.4 million Ebola cases by January 2015, the model estimates that 980 000 beds (to cover 70% of patients) would be necessary to begin to contain the epidemic in West Africa. Applying the ~30% mortality some ETUs were able to achieve in Sierra Leone,22 this would result in 294 000 deaths. The 30% of patients (420,000) unable to access ETU beds would be relegated to the 75% mortality rate of untreated EVD (315 000 deaths), a death toll that would be reduced to 126 000 deaths if they actually had access to ETU beds. In short, the model explicitly allows for 189 000 extra deaths West Africa, normalising an avertable massacre of almost 1% of the entire population of Sierra Leone, Liberia and Guinea as part of CDC containment recommendations. In contrast, the same CDC modellers certify that ‘the capacity of Ebola treatment centers in the United States (49 hospitals with 71 total beds) was sufficient to care for our highest estimated number of Ebola patients’.23

The village name ‘Bumpeh’ is not mentioned in the CDC publication; however, we might think of their model as anticipating the outbreak there in 2014, and, in the process, shaping the response to it. Accordingly, the dissonances between the logics imparted through this model and what we witnessed as health workers on the ground were manifest: many of the decisions made in the day-to-day approach to Ebola containment were decentralised—a product of conversations, improvisations, and innovations among local health workers and the few expatriates stationed there at the time, and shaped by long-term political alliances and systems, prior paradigms of public health implementation, local knowledge of the socioecological systems at play, and dynamic appraisals of the response resources available. In short, the work being done in ‘centers of calculation’ like Atlanta seemed in the initial analysis to have little utility;24 however, if we consider the necropolitical normalisation of nearly 200 000 allowable deaths, then the ways that some patients, health workers, and NGOs troubled the containment-over-care paradigm the CDC proffered come into view: Patients at times avoided going to facilities that they were well aware were not set up to provide lifesaving care (that is, were for isolation only); health workers at such isolation centres clamoured for, and often secretly provided, resuscitative care despite institutional rules against it; and some NGOs chose to focus on improvements in care and health systems strengthening rather than containment alone.25–27

In summary, while the predictions of what would unfold in Bumpeh offered by the CDC’s hyperbolic model may have been effective in conjuring fear about further spread of the epidemic and, as a result, garnered significant financial and other support from major global actors, it also delimited a certain paradigm of response and foreclosed the possibilities of others.28 The wave of intervention that did ensue—a conglomeration of transient NGO isolation centres and ETUs, and a foreign military incursion divvied along colonial-era lines—continued to advance a strategy of containment-over-care while doing little to attend to the sources of the region’s vulnerabilities to epidemic diseases in the first place.

Epidemiology: active surveillance, case investigations, and transmission trees

Two years after the Bumpeh outbreak, authors from the CDC and other international health institutions published an account (including a transmission tree) of it in the journal BMC Infectious Diseases. They used data from the VHF, case investigations, and treatment facility/laboratory records for their analysis, representing the outbreak graphically in figure 1.3 The article notes that ‘Two days prior [to the outbreak] Patient 1 had reported to the district hospital with unknown symptoms. He was clinically diagnosed with a stroke, his second, and sent home. Within 48 hours of discharge, he developed diarrhoea and vomiting and died at home.’ After his death, the district burial team was summoned and swabbed and buried the body; the study authors further note that ‘the body had been washed and prepared for burial according to traditional funeral practices that were banned after the Ebola epidemic began…’

Figure 1

Transmission tree devised by the U.S. Centers for Disease Control and Prevention based on data from the Viral Hemorrhagic Fever, case investigations, and treatment facility/laboratory records.3 Each box represents a confirmed Ebola case with *Patient ID: Gender (M=male, F=female) Age*. The length of the box represents the time the patient was in community, ranging from date of symptom onset to date of admission into healthcare facility or death. †Shaded boxes indicate the patient died, where unshaded boxes reflect the patient survived. A red outline signifies the case died in the community or was admitted to a treatment facility on the date of death. Arrows between cases represents known contact. The blue bar marks the day of the public health interventions (active surveillance and health education).

The article makes little emphasis of what transpired at the hospital—a hospital which, we know from working in Kono at the time, was largely shut down to patients as an international NGO sanitised and reconfigured it for Ebola screening. In the coming days, more people from the village of Bumpeh would fall ill; initial case investigations did not clearly identify (or, perhaps, did not log properly in the VHF database) epidemiological links. This was interpreted by the study authors as evidence that ‘delayed reporting of additional symptomatic individuals and contacts of suspected Ebola cases hindered public health officials’ ability to respond swiftly.’ As case-counts increased, the District Ebola Response Center (DERC) organised a 1-day educational campaign, during which community health workers (from an NGO with which some of the authors of this paper were affiliated) and health officers went in teams from house-to-house to observe, interview, and educate community members; several patients with Ebola were identified and transported via ambulance to an ETU. This was a harrowing journey at the time: the authors note that at least one subsequent Ebola case from the village is suspected to have contracted the disease from the ambulance ride. The majority of people taken away from their families on that day never returned alive.

In order to assess the effects of these educational and behavioural interventions on the outbreak in Bumpeh, the CDC authors identify two outcome variables: ‘(1) the amount of time after symptom onset that a suspected case-patient remained in the community before isolation or death and (2) whether the case-patient had a known epidemiologic link.’ These variables might be thought of as proxies for the ‘functioning’ of an Ebola response system—how quickly were patients removed from households, and how complete was the data collected about them. But we might also consider how the construction of these empirical questions already point to a paradigm of sense-making and storytelling: the functioning of the Ebola response system is collapsed onto the behaviours of and the utility of surveilling those it seeks to serve. Little is made of what occurred at the hospital when the man presented with a stroke, other than a note that a number of patients with Ebola were not recorded as meeting case criteria on initial presentation: both the requisite resources (or lack thereof) in order to bring patients promptly to care and the deeper affective and historical determinants of a patients’ eagerness to call for removal, isolation, and likely solitary death are unexamined.25 29

Ultimately, the purpose of such an epidemiological investigation is to render a synchronic report on salient features of the outbreak: reproductive numbers, transmission chains, epidemiological linkages and proximal behavioural aetiologies. In keeping with this disciplinary approach, the authors reported 50 cases of EVD, 32 deaths, and a reproduction number (R) of 0.93 (95 % CI 0.15 to 2.3) concluding, ‘Initial case investigation and contact tracing were hindered by delayed reporting and under-reporting of symptomatic individuals from the community. Active surveillance and health education contributed to quicker identification of suspected cases, interrupting further transmission’.3 This mode of representation buttresses a neoliberal, developmentalist paradigm of low-cost, rapidly implementable and time-bound responses that, it is claimed, can function effectively even amidst impoverishment and structural barriers to care. In other words, health outreach and technical advising can ‘contain’ outbreaks whose determinants stretch back in time and across continents.30

Seroepidemiology: dynamic networks and spectra of disease

Six years after the Bumpeh epidemic, Kelly et al. published a distinct epidemiological analysis of it (including a rather different transmission tree, figure 2) in the journal Open Forum Infectious Diseases.4 This study team used EBOV antibody surveys to supplement interviews of households and showed there were many missed cases and epidemiological links in the previous CDC assessment. In essence, Kelly and colleagues redescribed the same ‘outbreak’ as consisting of 106 cases and an Rt that varied from 0.42 to 12.3 depending on the generation. Their research found 40 people who were close contacts of officially reported Ebola cases and who had serological evidence of infection—suggesting that official case counts (including those reported in the prior CDC paper) were over 50% too low. When interviewed, these individuals, who had not been classified as Ebola cases in prior reports, recalled experiencing a range of symptoms at the time (from none at all, to arthralgias and headache, to debilitating fever and fatigue).

Figure 2

Transmission tree devised by Kelly et al. using EBOV antibody surveys to compliment household interviews.4

The authors’ behavioural data suggested that Ebola survivors and their close contacts had a range of exposure levels, even while cohabitating and offering improvised domestic care for each other, suggesting that people were attuned to risks and modes of transmission and may have adapted their caregiving practices at the time (that is, before the outreach interventions described in the prior CDC paper.) Integrating these data with serological evidence, the team concluded that there may be ‘a dose-dependent relationship between exposure risk and severity of [Ebola] illness.’

The addition of serosurveys and in-depth interviews to the epidemiological approaches discussed in section 2 can also be critiqued as highly reductive, effacing our understandings of the sociohistorical forces that determine epidemics by fetishising individual behaviours, exposures, and immune responses;31 32 however, Kelly and colleagues’ redescription of the Bumpeh outbreak does extend our understanding of Ebola dynamics beyond a model of atomised individuals engaging in billiard-like interactions.33 The story they tell suggests that Ebola manifests as a spectrum of disease across more widely dispersed social networks—intersubjective webs—made up of dynamic and adaptive individuals integrating information gleaned from public health messages with personal experiences of the healthcare system, cultural practices, and epistemologies of healing.25 29 34–36 The authors bring us closer to a description of Ebola transmission that centres the complexities of unfolding social relations (describing Ebola as a ‘caregiver’s disease’, for example). But their rendering of Ebola is still apolitical and largely ahistorical. They do not offer any structural-level suggestions for future outbreak response, and instead land at a rather modest call for wider distribution of personal protective equipment (PPE) during outbreaks, as ‘reducing exposure risk among household members unable to quarantine in separate locations or forced into caregiving roles while awaiting ambulances and safe transport to Ebola treatment centers has the potential to prevent severe and deadly EVD’.4

Political economy: active underdevelopment and illicit financial flows

A number of scholars have explored the proximal determinants of the 2013–2016 Ebola epidemic, centering, for instance, the roles of dysfunctional healthcare delivery systems, a highly mobile and interconnected population travelling via porous borders, debilitated government institutions, burial practices that involved contact with contagious corpses, and unsafe and poor-quality provision of care to infected individuals.37–42 More distal determinants, including legacies of enslavement, colonial and neocolonial resource extraction, structural adjustment, and illicit financial flows have received less attention.43–52

In our own research in Bumpeh in the aftermath of the epidemic, the ongoing presence and remnants of these sociohistorical forces were visible. Bumpeh is on the outskirts of Koidu, a heavily mined region of eastern Sierra Leone. The towering mounds surrounding the Koidu Holdings diamond mine (figure 3), a foreign-owned venture that has rarely paid income tax on its extraction, cast a long shadow on the roads into the village.53 This mine is storied to have originated as a government payment for a mercenary force brought in to fight rebel forces during the country’s civil war (1991–2002) and continues to use colonial-era strategies for land reclamation in order to expand access to diamond-rich territory.51 The grounds around Bumpeh are pockmarked with ‘artisanal’ mines, where young and old alike toil day in and day out with very little to show for it. In the far corner of the village, people continue to live in concrete row houses (in various states of decay) built by the National Diamond Mining Company (NDMC) to house workers before the country’s civil war. Nearby, almost entirely overtaken by forest are the decades-old ruins of an idyllic settlement for European and elite NDMC staff, complete with movie theatres, dance and mess halls, European-style schools, gardens, tennis courts and swimming pools, and a golf course.54

Figure 3

Diamond mine in Koidu as viewed from a UN helicopter. Photo credit: ET Richardson.

It is evident to social scientists, geologists, and the local population alike that there is abundant mineral wealth in Bumpeh.51 55 But throughout the past century, this wealth has been systematically expropriated from the region and then out of the country without the taxation necessary to develop and sustain an effective local health system.51 52 56 When Ebola emerged in the region and the Koidu Government Hospital—operating largely without power or PPE, and with only two regularly working doctors for approximately 500 000 people—became overwhelmed, it is no wonder that the virus spread rapidly through homes and families.42 (At the same time, Ebola did not spread widely in any Global North country and was contained relatively quickly in Nigeria and Senegal, evidence that Ebola transmission can be rapidly halted with effective, resourced healthcare systems.) This dismal state of healthcare within Bumpeh is also a legacy of structural adjustment,57 verticalised health programmes, and the gutting of infrastructure during the civil war. But the effects of resource expropriation cannot be overstated: in the years leading up to the Ebola epidemic, the net amount of wealth leaving Sierra Leone via illicit financial flows—as on the rest of the continent (figure 4)—is estimated to have exceeded all development aid afforded the country, suggesting that Sierra Leone continues to develop the global North rather than the other way around.51 In summary, a political economic analysis demonstrates that the ‘underdevelopment’ of Kono’s health systems—and the Bumpeh outbreak that eventuated—arose and is inextricable from these histories of predatory accumulation, which operate according to the broader logics of colonialism and racial capitalism.50 58–60

Figure 4

Intercontinental pillage: ‘African countries received $161.6 billion in 2015—mainly in loans, personal remittances and aid in the form of grants. Yet $203 billion was taken from Africa, either directly—mainly through corporations repatriating profits and by illegally moving money out of the continent—or by costs imposed by the rest of the world through climate change’.52

But these overarching descriptions of how capitalism generates outbreaks and excess mortality in places like Bumpeh still do not capture the ‘whole’ picture and are necessarily meta-narrativist. Purely political economic accounts risk effacing granular narratives of individual experience, while also sweeping away the agency and perspectives of the people of Bumpeh. Like epidemiological stories, political economic portrayals can leave depictions emanating from the community of Bumpeh itself underexamined.

Ethnography: memories, critiques, and Ebola’s lingering presences in Bumpeh

“Pa Bangura was someone you could call on at any time to drive you [anywhere]. You’d call and he’d say: ‘Ok, just a minute, let me get my things,’” Alusine, a friend of the taxi driver Pa Bangura, told our research team in Bumpeh some 4 years after the outbreak. One afternoon, Bangura picked up a sick woman outside of town and drove her to the hospital for care. Alusine did not know who this woman was—had Bangura met her on the road? Was she fleeing another village known to have patients with Ebola? Whatever her story, she got into Pa Bangura’s car, and as they made their way on the rutted road to Koidu Government Hospital, she vomited all over the passenger seat.

After dropping her at the hospital, Pa Bangura returned immediately to Bumpeh. He called a few of the men who stayed near his compound to help wash out his car. We spoke to several friends of those men: they assured us they had known about and feared the illness that was sweeping through Sierra Leone at that time and were aware that it spread by exposure to infected body fluids. But Pa Bangura was known as a generous taxi-driver; he frequently brought sick patients and their families to the hospital, and of course there were a lot of sick people in Kono District at that time who did not have Ebola. ‘What to do?’ Pa Bangura was torn. But he needed to continue his driving business, the village needed his taxi services, and his car needed to be cleaned.

A week or so after that, Pa Bangura began to feel ill. He first reported a fever and body aches. Days passed with no improvement in symptoms. Bangura’s friends insisted that he go to the doctor, so the man took a motorbike taxi to the same hospital he had driven to a week earlier. There, he learnt that an international NGO had recently moved into the hospital and shut it down to sanitise and ‘reset’ its flow in accordance with infection control protocols. By the time Bangura got there, all patients were turned away at the gates. If triage staff suspected that they had Ebola though (ie, they fit WHO criteria as ‘suspect cases’), they were sent to another district (Kenema) for isolation, testing, and treatment.

Bangura, however, had long had a relationship with one of the doctors at the hospital, Dr. Mahmood. Dr. Mahmood had treated Bangura before for hypertension, and when he examined Pa Bangura at the hospital, he noted that Bangura seemed weaker on one side of his body and thought he may have had a stroke. Because stroke was not part of the WHO case definition for EVD at the time, and since Bangura could not be admitted to the hospital for further workup, the doctor contacted a nurse from a clinic near Bumpeh to provide Bangura with supportive treatment at home.

Pa Bangura’s condition worsened, and later that week, he died at home. A friend or family member called the national Ebola hotline (figure 5) and reported that the man had died, requesting a team to remove and bury Pa Bangura’s remains safely. But since the family felt assured that Pa Bangura had been diagnosed with a stroke and not EVD, a small group of Bangura’s closest peers also held a traditional funeral as a last respect to their friend, during which the corpse was washed. Then, the family carefully laid the corpse outside again, where it was swabbed and buried by the Ebola burial team the following day.

Figure 5

There are many ways to depict Ebola dynamics.6

A few days later, the family learnt that the swab had tested positive for Ebola. The nurse who had taken care of Bangura was the first to hear, and she called a friend: “Did you hear the results, did you hear the results of the test?” Alusine, who had attended the burial and had helped care for Pa Bangura when he was sick said: “When we heard the news, we thought: ‘We are all finished!’”

The Ebola survivors and those who had lost family members described two ensuing weeks of senseless terror and death. Alusine fell ill, as did his wife and infant daughter. Pa Bangura’s nurse contracted the disease and died as well. That week, we were told, many of those suspected of being ill were transported in sweltering ambulances packed with seven or more people. Several of those who tested negative at the destination ETU but later became ill were thought to have contracted the virus from the journey in the packed ambulance. Alusine’s wife and daughter both succumbed to the disease far away from home and from him.

‘The doctors—they’re the ones that killed them both,’ we were told by several friends and family of Pa Bangura and his nurse. The Sierra Leonean doctor who had cleared Bangura dared not set foot in that town again. But survivors also focused on the fact that the international NGO had commandeered the public hospital and shut down all clinical services. So much ‘Ebola money’ had flowed to these NGOs during the epidemic, we were told, evidenced by the waves of Europeans and Americans that appeared in Kono, the Land Cruisers and the hotels that politicians had managed to repurpose in record speed to be rented out to the NGO workers. As coauthor SDK explained, reflecting on our conversations with the survivors: “The attitude of some of the expatriates that came!…Just like our big shots here, we know, they were renting out their vehicles [and making so much money]. I think this was not lost on the common people—they have eyes to see! You really start…thinking whether this [situation] was man-made, whether it was intentional.” Indeed, SDK’s interpretation is not a conspiracy theory but rather dovetails neatly with that of the political economic Story Technology: his comments bring into sharp relief the mutual interdependence of Bumpeh’s impoverishment and vulnerability to public health crises, the economic histories that have produced that vulnerability, and the (well-funded) waves of humanitarian workers and institutions that profit off of these acute-on-chronic catastrophes.30 61 62

In the end, amidst all this ‘Ebola Bisnes’ (as our interlocutors and others have called it),63 the temporality of the humanitarian response—the rapid incursion into and seizure of the hospital, the shutting down of all routine services, and the nosocomial transmission during patient transport—inadvertently propagated the disease through Bumpeh. Some we spoke with in Bumpeh seemed sympathetic to the NGO that had shut down the hospital. They knew that the original facility—without proper equipment or triage procedures—did not have the capacity to keep patients and health workers safe. And yet, others we talked to emphasised that Pa Bangura had fulfilled demands from so many different quarters—to offer a lift for the sick passenger, to notify his friends and family that he was not feeling well, to go to the hospital to seek care given his knowledge of the disease. He had been ‘cleared’ of Ebola and in the process became what some have unjustly deemed a ‘superspreader’.8 16

The circulating accusations within Bumpeh that the doctors in the hospital were ultimately responsible for all the death that ensued also speak to debates about the limited scope of ‘expert’ knowledge on the virus: at the time, the WHO had excluded the now-recognised symptom of stroke from its criteria for Ebola screening. The evolving description of Ebola offered by VHF experts in Brussels and Geneva that contributed to Pa Bangura’s misdiagnosis was identified by Ebola survivors we spoke with as the accepted origin of the Bumpeh epidemic.

Today, the survivors persistently reoriented our questions to present conditions: the poverty that had ensued in the aftermath of the epidemic, and the further retreat of post-Ebola social support services. “Before [Ebola], we had more food—three meals a day,” one man who had cared for Bangura and contracted the virus told us. “But now—it is difficult. Really, really hard.” In addition to the emotional toll of the loss of loved ones and the memories tended in their aftermath, the sudden absence of multiple breadwinners thrust many families from tenuous agricultural existences to brutal peri-urban poverty. Those who merely lost family members without a single survivor did not even receive the minimal NGO and government support offered post-discharge. In Bumpeh today, the inhabitants of these homes—colloquially called ‘death houses’—remain destitute and embittered in a marginalising and anaemic post-Ebola humanitarian regime.

Five years later, Alusine told us, nearly all the aid had dried up for everyone. The stories we heard of family members lost in the subsequent years to motor vehicle accidents, tuberculosis, malnutrition, malaria, and other unnamed illnesses underscored that Ebola was exceptionally targeted for fleeting assistance (but which, as lived experience, was in some ways inseparable from the ongoing violence of day-to-day life.) Researchers, including those from our own group, came in from time to time to appraise the communities’ economic status, psychological well-being, health sequelae, and the antibodies of Ebola left in the survivors’ blood. At one poignant and painful moment, witnessing a study team go house-to-house remunerating a blood draw with a small handout of cash as part of the Kelly study cited above, a young woman remarked to her child loudly for the whole research team to hear: ‘You’re hungry? Go sell blood, buy bread.’


Instead of trying to analyze complex phenomena in terms of single or essential principles, these approaches acknowledge that it is not possible to tell a single and exclusive story about something that is really complex.—Paul Cilliers, Complexity and Postmodernism: Understanding Complex Systems (1998)

The Ebola outbreak in Bumpeh has no perfect representation; however, a set of particularities emerged during our review of various attempts to depict it: a renowned community member and a sick woman in need of a ride into town; international ‘experts’ and an inadequate case definition; a cultural practice (burial) and an underdeveloped health system; a history of human and resource extraction and a paradigm of containment-oriented global health.

Our interest in this paper has been to focus attention on how Story Technologies are used as visibility regimes to make ‘sense’ of phenomena.64 The first two technologies we examined, mathematical modelling and epidemiological outbreak investigations, present proximal, apolitical portrayals of epidemic emergence and transmission. These partial depictions are amplified through both academic and media echo-chambers, doing work in the world as Story Technologies.

With regards to such practical consequences, the catastrophising modelling beckoned in a humanitarian response paradigm, which was buttressed by epidemiological studies insisting on the need for more health education and outreach as keys to epidemic containment. As Story Technologies, both decouple analyses of power from disease dynamics, reifying particular risk factors at the expense of others.65 This serves to maintain status quo relations of transnational inequality by vitiating our understandings of more distal determinants, including legacies of enslavement, colonialism, and extractive capitalism.6 If we therefore accept that these stories are by no means objective or value neutral—exemplified by the CDC’s touting the ‘benefits’ of a model that, of all those published, was least consistent with the observed epidemic28—and instead serve as tools beckoning particular forms of engagement, might there be more space for the elevation of other forms of storytelling, such as those that emphasise the historical situatedness of Ebola and calls for more reparative regimes of global justice?

In section 3, we juxtaposed the CDC’s scriptings of the Bumpeh outbreak with a more expansive—but by no means value-neutral—assemblage of facts about the disease cluster, gathered by academic investigators using antibody surveys. Seroepidemiological insights blur the binary categories used in acute outbreak response–case versus no case, symptom versus no symptom, exposure versus no exposure—helping to reveal the illusory nature of how viral diseases, in social contexts, leave their marks on bodies. But despite expanding on knowledge gleaned during active surveillance (section 2), these methods—from a political economic vantage (Section 4)—still filter out information that would reveal the global North’s complicity in outbreak determinants (figure 5).

Political economic approaches, when used to unmake the coloniality of knowledge production, arguably offer more just redescriptions of epidemic phenomena.66 This Story Technology can help uncover how legacies of colonialism actively shape social life and vulnerabilities to outbreaks. The primacy of these sociohistorical (rather than virologic) forces might then beckon reparative responses.67

The final Story Technology we feature, ethnography, is a mode of knowledge production that offers a chance to co-construct knowledge with interlocutors.68 Our own adoption of this technology emerges from well over a decade of dialogic encounters in Kono District across lines of difference and is attentive to personal memories and lived experiences. Corroborating chains of events, in this Story Technology, is less important than tending to the sense-making that unfolds in the aftermath of crisis.

As a technology, ethnographic storytelling has the potential for strategic resistance to status-quo relations of inequality by promoting emergent and experience-near political critiques. And while it also produces partial renderings of complexity, it has the potential to address, head-on, the challenges of equitable knowledge production through its centering of reflexivity and elevation of subaltern voices. These are the very conundrums on which ethnographic methods dwell—rather than claiming, through tinkering with methodology, to skirt them. In this way, ‘methodology’ can be thought of as a collective, situated mechanism of inquiry that incorporates multiple perspectives and recognises knowledge-producers outside of existing hierarchies, while prioritising their equitable participation.


As Edward Said has observed, the power to narrate—to tell certain types of stories and to erase others under the guise of objectivity—is a primary means through which culture legitimates imperialism.69 In this study, we have accounted for the narrative powers of various academic disciplines as Story Technologies, calling attention to the instrumental—in addition to the descriptive—role they play in ordering contingencies and intersubjectivity in the wake of health crises.64

Our overall objective was to detail the discursive struggles (ie, uncover the machinations of power within claims to objectivity70) that take place in accounting for health phenomena, with the goal of democratising knowledge production about epidemic determinants.71 72 As partisans of solidarity, our account of the value of cooperative human inquiry has only an ethical justification, not an epistemological or metaphysical one:73 truth in this case is the pluriversal, dialogic movement towards agreement on ever more just redescriptions of social phenomena.6 73 74

We end by examining the potential effectiveness of the approach used in this paper as a Story Technology, employed towards achieving global health justice. Its potential success could (1) destabilise the epistemic privilege and legitimacy afforded to the technologies with the most material and institutional backing both rhetorically and in the minds of practitioners who might be encouraged to alter their worldview and pursue alternative methodologies for epidemiologic research and response; (2) present a counternarrative that respects the dignity of marginalised knowers (an attempt at epistemic justice in the face of cultural hegemony); and (3) offer a platform for the cross-pollination of multiple forms of knowledge production that may redescribe the interrelation of the viral, the historical, the political, and the phenomenological in a more just fashion.

But there are hurdles to such success, including (1) dissemination on an inaccessible academic stage; and (2) the mismatch of epistemic currency between academic discourse and knowledge production often derided as folklore. Addressing these potential failures cannot be wholly accomplished with this technology; it can only be ameliorated when used as part of a larger arsenal for reparative justice and decoloniality.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

The parts of this study subject to ethical review were approved by the Sierra Leone Ethics and Scientific Review Committee and the Harvard University and University of California, San Francisco Institutional Review Boards. Participants gave informed consent to participate in the study before taking part.



  • JDK and ETR are joint senior authors.

  • Handling editor Seye Abimbola

  • Twitter @mfallah1969, @unsymbolize

  • Contributors ETR designed the study. RF, MM, TM, MMM, SC, DJK and ETR conducted the literature search. RF, SDK, TM, SC, MBB, YD, DJK and ETR collected data. All authors interpreted the results. All authors critically revised the article. All authors approved the final version. ETR is

    responsible for the overall content as guarantor.

  • Funding This study was supported by NIAID K08 AI139361, NIAID K23 AI146268, NIGMS R01 GM130900 and NIGMS T32GM007618.

  • Map disclaimer The depiction of boundaries on this map does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. This map is provided without any warranty of any kind, either express or implied.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.