War and disease have intertwined across history to the point that untangling the affects one has upon the other may appear incalculable. But UF Emerging Pathogens Institute researcher Burton Singer’s latest project does just this: it takes the measure of conflict and civil strife upon an ongoing Ebola epidemic.
The Democratic Republic of Congo, a central African country, has been roiled by armed conflict and civil strife for years, while also weathering multiple outbreaks of highly contagious Ebola and other infectious diseases, such as measles, cholera, pneumonic plague, sleeping sickness and river blindness.
The latest Ebola outbreak, which first bubbled up in April 2018 in North Kivu and Ituri provinces, has festered into the continent’s second worst Ebola epidemic. It is dwarfed only by the 2014 West African outbreak.
To the casual observer and reader of official reports, Singer says, it may look as if the DRC’s armed conflicts are periodically worsened by outbreaks of highly contagious Ebola. The skirmishes tend to center upon politics, ethnicity, land ownership, economics and control of extractive natural resources.
But Singer’s research, which was recently published in the Proceedings of the National Academy of Sciences, is the first to quantifiably demonstrate that the periodic repetition of disruptive, often violent, events in the DRC’s North Kivu province has promoted an Ebola epidemic to become entrenched, rather than the prolonged disease outbreak itself fueling social and political destabilization.
Singer, the study’s corresponding author, is an adjunct professor with the EPI and UF’s College of Liberal Arts and Sciences mathematics department. His collaborators include Yale University researchers Chad Wells, the lead author, Abhishek Pandey, Martial Ndeffo Imbah and Alison Galvani. The group also included medical doctors Bernard Gaüzère and Denis Malvy who worked on the ground in the DRC with Médecins Sans Frontières.
“We have all this literature that qualitatively is correct, which says where there is deep and prolonged civil unrest, real impacts on the health spheres will follow,” Singer says. “But once you have that piece, there is nothing in the journalistic literature that I’ve seen, or in the epidemiological literature, to explain the mechanism. How exactly does this work?”
Singer is a member of the National Academy of Sciences, and his resume roams from ecology and evolutionary biology to demography, medicine and social statistics. Prior to joining the EPI he held an eclectic collage of academic positions: the Charles & Marie Robertson professor of public and international affairs at Princeton University, chair of Yale University’s department of epidemiology and public health, and he served as an associate dean for Public Health in Yale’s School of Medicine. This was all preceded by 18 years on the faculty of Columbia University.
Having “retired” ten years ago, Singer has had the unbridled freedom to focus his time strictly on research where he uses mathematical tools to understand infectious disease processes, and is engaged in an in-depth reconsideration of the evidential basis of clinical medicine in collaboration with his former Yale colleague, Ralph Horwitz of Temple University.
In his latest study, Singer and his colleagues matched qualitative information documenting specific disruptive events in North Kivu province, with visitation trends to Ebola treatment centers and measurements of Ebola cases over time. He and his colleagues then led the development of a mathematical model that linked specific spikes of Ebola cases in North Kivu province to the preceding violence or social disruptions.
“When you look only at the reports about the disturbances, and the number of Ebola cases around those times, you lose all the context of how the unrest builds up to these explosive events,” Singer says. “What we did, heavily relying on ethnographic input from Bernard Gaüzère of MSF, was quantify how the bid up to the events themselves caused disease control measures to decline.”
A short history
Ebola first emerged in 1976 as a viral hemorrhagic fever disease, sickening a cotton factory worker in Nzara, South Sudan. It’s strongly suspected that the virus spreads to people from fruit bats, which were known to roost in the cotton factory’s eaves. The virus rides in bodily fluids such as blood, vomit, feces, and semen to hop from host to host. After an outbreak that died out, it next popped up in what was then northern Zaire (which is now the DRC) and then Uganda. By the mid-1990s, it was popularized globally in Richard Preston’s bestselling book, The Hot Zone, although it is endemic only to Africa.
Sometimes, but not always, victims bleed internally or externally. When it kills, death comes within 6 to 16 days of symptoms beginning. But slightly less than half of those who become infected survive, only to later battle joint and muscular pain, inflamed livers, hearing loss, problems with their vision and difficulty regaining weight. Survivors, and their family members, sometimes face discrimination and fear from others in their communities.
In 2017, a ring vaccination trial in Guinea and Sierra Leone demonstrated nearly 100 percent efficacy against Ebola when administered 10 or more days prior to exposure. Two UF researchers affiliated with both the EPI and the Colleges of Public Health and Health Professions, and Medicine, Ira Longini and Natalie Dean were involved in the design of the vaccine trial and were coauthors on The Lancet article which reported its success.
In the decades since Ebola made its entrance, healthcare workers, epidemiologists and researchers have devised protocols to strategically contain and control outbreaks. Some of these measures include healthcare workers using protective gear, early detection and isolation of infected individuals, ring vaccination campaigns, and using contact tracing to track down people who may have been exposed to an infected person in order to determine their status.
Ebola has raged in the DRC’s North Kivu and Ituri provinces for the past 19 months. Since inception, there have been 3,180 confirmed cases and 2,197 fatalities (as of Nov. 21, 2019). At least 10 healthcare workers had been targeted and killed by the study’s conclusion.
In contrast, the country’s previous Ebola outbreak in Equateur province in 2018 lasted 2.5 months, with 39 cases and 19 dead. The main difference between Equateur province’s ability to curtail its Ebola epidemic, and North Kivu province’s inability to do so, centers on maintaining disease control measures: Equateur performed adequate contact tracing and ring vaccination campaigns comparatively unimpeded, whereas North Kivu’s efforts are periodically interrupted.
To piece together how violent conflicts and the Ebola outbreak may have affected each other, Singer’s team made a timeline of events in the DRC and then conducted an “ethnographic appraisal” of the local conditions that both preceded and followed these events. The ethnographic work included vital first-hand observations recorded by Bernard Alex Gaüzère, M.D., who was a senior Ebola doctor with Médecins Sans Frontières (also known as Doctors Without Borders) in Katwa, DRC between January and February 2019. His testimony is included in the study’s supplemental materials. Next, they modeled Ebola transmission and control efforts.
They found that following major conflicts, such as violence enacted against an Ebola treatment center, the speed of isolating new cases was then hampered, and the effectiveness of vaccine campaign efforts also declined. The epidemic’s declining trajectory was then reversed, and cases would predictably spike.
The research team showed that unrest in North Kivu could delay outbreak reporting to authorities by up to three months. In addition, Ebola treatment centers were destroyed in coordinated attacks, roads were barricaded, and healthcare workers were targeted for intimidation or violence. All of which added up to people not seeking diagnoses or care in Ebola treatment centers, and disrupted contact tracing and vaccine campaigns.
Disruptive events extended the time from a patient’s symptom onset to their isolation on average from 8.13 days to 10.04, or just shy of two days lag time on average. This extended the time they were able to transmit the disease to others.
The effectiveness of a vaccine campaign prior to disruptive events was measured at 52 percent, but this was reduced to a minimum of 4.8 percent after disruptions, and averaged out to just 29.5 percent when all disruptive events within the study time frame were considered.
Last, the team also tracked something known as the reproductive number which measures the number of new cases spawned by each new infected victim. Ideally, when disease control measures are working, the reproductive number falls below one. But when epidemics are growing, it rises above one. The team’s model shows the reproductive number repeatedly declining toward zero, only to spike above one in a series of fluctuations where the spikes are linked in time with the buildup of community anxiety and subsequent outbreak of violence.
“We see this yo-yoing back and forth in these graphs, and if it was natural processes alone at work, that would never happen,” Singer says.
What these numbers demonstrate is that disease control measures successfully resulted in a declining incidence of Ebola, only to have the civil unrest upend the trend and fuel the epidemic. The model used in this study could be adapted to explain other infectious epidemics which are enmeshed in civil strife or armed conflict, Singer says, such as cholera in Yemen.
“One of the things that really led us to go after this,” Singer says, “Is the fact that the WHO publishes detailed and timely reports on the major attacks. We then paired those with on-the-ground observations from a member of Doctors Without Borders about what was happening in the lead up to these attacks. His narrative is critical. Without it, we would not have had the detail to begin calibrating at what point community anxiety reaches a level that interferes with vaccine distribution and with people going to health centers.”
By: DeLene Beeland