According to Bengtsson (1999) multiple case studies can be selected for literal or
theoretical replication. Literal replication cases are selected because the cases and results are predicted to be similar. Theoretical replication refers to cases that are selected because the cases and predicted results contradict each other. The cases in this study were selected under literal replication as both cases involve the experience of an EVD outbreak on the community within a public health system in western Uganda. This is referred to as a general explanation model where certain characteristics are shared by both cases, even though the cases also vary in detail (Bengtsson, 1999).
The first case study selected was in Bundibugyo district in western Uganda where an EVD outbreak caused by Ebola Bundibugyo occurred in 2007. The second case study chosen was located in mid-western Uganda in Kibaale district, where an EVD outbreak caused by
Ebola Sudan was managed in the town of Kagadi in 2012. The characteristics shared between the two case studies include contexts that demonstrate an array of similar social determinants including extreme poverty, a fragile health system, neighbouring conflict, displaced populations, and a burden of endemic disease. Shared characteristics also included the infrastructural capacity and resource scarcities within both hospitals at the centre of the outbreaks. Both cases included a district general hospital were the EVD outbreaks were managed. Both hospitals were built during the same era, were similar in design and construction, and shared similar resource challenges. Apart from ongoing historical land issues outlined in section 2.3.4.2, Kibaale district is relatively stable but shares similar levels of poverty, rural isolation, endemic burden of disease and population burden with Bundibugyo. Kibaale also hosts a refugee population from eastern DRC and has the highest fertility rate in Uganda.
Details that varied between the two case studies include timing in relation to the
geopolitics of the outbreaks. The International Health Regulations revised in 2005 were ratified in 2007, the year coinciding with the Bundibugyo outbreak. In theory, this meant that all WHO member countries including Uganda were mandated to possess suitable surveillance and response systems, despite no funding being made available. 2007 also marked the year when Uganda hosted the Commonwealth Heads of State General Meeting (CHOGM) between the 23rd
and 25th
November in 2007. This allowed consideration for how this politically important event may have influenced delays in announcement and response to the outbreak.
The EVD outbreak in Kibaale occurred five years following ratification of the 2005 Revised International Regulations in 2007. By this time, Ebola had become more recognised on the global health agenda. By now the narrative of ‘emerging infectious diseases’ was well established and available funding for responding to such outbreaks was more explicit. As a result, the 2012 Ebola outbreak was rapidly diagnosed and responded to, bringing it under control within three months of its emergence in July 2012. Focusing on the interim, intervention and aftermath periods of both outbreaks allowed for
exploration of the broader context of EVD outbreaks not normally captured during outbreak reports.
As explained in the previous section a number of constrains can limit the number of case studies chosen in a multiple case study including the availability of cases, funds and time available including ethical considerations. These practical and logistical factors were also employed when choosing the two case studies used in this study. The researcher was based at a university campus in northeastern Rwanda where the research sites could be reached by road either directly or via Uganda’s capital, Kampala. Bundibugyo is located 363km west of Kampala via the Fort Portal highway (A109) and construction of a new road between Fort Portal and Bundibugyo was completed in 2014. Bundibugyo can be reached from Kampala in less than seven hours. The town of Kagadi in Kibaale district can be reached via a non-surfaced road within one hour from the town of Kyenjojo along the Fort Portal highway. The researcher also had a co-supervisor based at Makerere University School of Public Health. Logistically the researcher used a loop from Nyagatare in
Rwanda, to Kampala, Uganda and from Kampala to Bundibugyo diverting to Kigadi town in Kibaale district on the return trip to Kampala.
3.2.3.1 Bundibugyo
Bundibugyo is located in the extreme west of Uganda bordered by Lake Albert and Hoima district to the northeast and the DRC along its western border. Bundibugyo became a district in 1974 having been traditionally part of the Kingdoms of Toro and Kabarole. It is the only Ugandan district isolated west of the Rwenzori mountains and is both
geographically and culturally part of central Africa, despite being politically within East Africa. Construction of a new road between Fort Portal to Bundibugyo and Lamia was completed in 2013 making Bundibugyo town more accessible.
The population of Bundibugyo according to the 2014 national census was estimated at 224,387 and 79% of the population are under 30 years (UBOS, 2017). Eighty eight percent of the population of Bundibugyo is engaged in crop growing. The main crops grown are coffee, cocoa, beans, rice, cassava, matoki, and vanilla. According to Uganda Bureau of Statistics (2009), Bundibugyo has 34 health units serving a population of over 200,000 (Figure 3.2). The doctor to population ratio is approximately 1:70,500. The most commonly reported causes of ill health and mortality are malaria at 68.4%, respiratory diseases at 12.5% and diarrhoea at 6.9%. HIV is reported at 0.1% but is likely to be under reported. Infant mortality is 102/1000 births and maternal mortality is 505/100,000 births.
Figure 3.2: Health Facilities in Bundibugyo District, Western Uganda (Source: Okware, 2015)
Bundibugyo district has experienced a long history of conflict and is particularly affected by events in neighbouring DRC. For example on the 5th
of July 2014, a week prior to data collection, a rebel group stormed the town of Bundibugyo directing attacks on the police force and military barracks and a total of 85 people were killed (BBC, 2014). Previous
attacks in the area occurred between the Allied Democratic Forces (ADF), an anti - Ugandan government rebel group supported by the government of Sudan who fought the government of Uganda during the Sudan 2nd
civil war (1983-2005). In the 1990’s, tens of thousands of local civilians were displaced by ADF insurgents and in 2007, intense battles between the ADF and the Ugandan military occurred inside the Semuliki National Park close to Bundibugyo town. In 2013, resurgence of the ADF in eastern DRC displaced over 60,000 refugees from DRC into Bundibugyo, putting excessive pressure on an already weak health system.
An EVD outbreak was officially announced on 29th
November 2007 following a suspicion of an unknown disease that was first reported in June 2007. The Bundibugyo case study involved three study sites: Kikyo Health Centre IV, Bundibugyo General Hospital and the District Health office, which were the main sites from where the outbreak was managed. In total 132 cases were reported and 42 people died during this outbreak. The last case was reported on the 3rd
January 2008 and the outbreak was declared officially over on the 20th
February 2008.
3.2.3.2 Kibaale
The second case study is based in Kibaale District in Midwestern Uganda bordered by Lake Albert to the west, Hoima district to the north, Kiboga district to the east, and Mubende District to the south. Kyenjojo, Kabarole, and Bundibugyo districts border the southwest of Kibaale district (Figure 3.3).
Fig. 3.3: Kibaale districts south east of Lake Albert
Kibaale district was created from Hoima district in 1991 and consists of 3 counties, 18 sub- counties, and 2 town councils. Uganda Bureau of Statistics (2009) conveys that the mainly rural population of 582,000 is mostly engaged with subsistence crop production (UBOS, 2009). The population growth rate is 5% with 51% being female. The population density increased from 20 persons per square kilometer in 1969 to 98 in 2002. In the 2002 consensus 88.2% of the population, live in temporary dwelling units (mud and wattle houses with thatch or iron sheet roofs). Only 0.4 % of households have electrical light, 64% latrine coverage and only 5% have access to hand washing facilities. The district lacks surfaced roads.
There are 34 public health unit facilities in Kibaale district. The doctor to population ratio is 1: 42,000. Life expectancy is lower than the national average at 49.1 years. Malaria accounts for over 40% of reported morbidity, acute respiratory disease for over 20%, anaemia 8%, intestinal worms 5.5% and a HIV prevalence of 6.4% is reported (UBOS, 2009). Infant mortality rates are 75/1000 births and maternal mortality rates are
435/100,000. Contentious ethno-political instability has frequently surfaced in Kibaale District, referred to in chapter two (Section 2.3.4.2) as part of the “Lost Counties” issue. An EVD outbreak was declared on 28th
July 2012 resulting in 17 fatalities out of 24 cases, 12 of these deaths occurred within one family. The outbreak was officially confirmed over
on the 24th
August 2012. A total of 13 probable, 11 confirmed cases and 17 deaths were reported during this outbreak (WHO, 2012b).