Knowledge, perceptions and effects of Ebola virus disease outbreak on the pig value chain in the agro-pastoralist district of Luwero, Central Uganda

Background Ebola Virus Disease (EVD) outbreaks have a significant impact on the health and wellbeing, and livelihoods of communities. EVD response interventions particularly affect the food value chain, and income security of pig farmers in agro-pastoral communities. Despite the enormous effort of EVD response interventions, there is paucity of information towards EVD among those involved in the pig value chain, as well as the effect of EVD outbreaks on the pig value chain. This study therefore, assessed the knowledge, perceptions on the occurrence of Ebola and its effects on the pig value chain in the agro-pastoral district of Luweero, Central Uganda. Methods A cross sectional study was conducted in two parishes of Ssambwe and Ngalonkulu, Luwero district. A total of 229 respondents were included in the study. Structured questionnaires, key informant interviews and focus group discussions were conducted to collect data. Quantitative data was analysed using SPSS version 22 while qualitative data was analysed using thematic content analysis. Results Of the 229 respondents, 95.6% could recall the occurrence of the last EVD outbreak in their locality. About 24.5% associated EVD with touching pigs or eating pork. Regarding knowledge, 194 (84.7%) correctly associated EVD with handling Ebola infected persons, 191 (83.4%) with migration of people from endemic areas, 148 (64.9%) eating monkey meat, 127 (55.5%) with eating bats, and 198 (64.9%) with conducting public meetings where there is an Ebola infected person. Out of 142 farmers, 55 (38.7%) believed that Ebola outbreaks affected demand and sale of pigs. The EVD outbreak significantly led to a reduction in the average number of pigs sold (P = 0.001), the average number of pigs bought by traders (P = 0.04), and the number of pigs sold/ slaughtered by butcher men at pork eating places (P = 0.03). Conclusion This study showed that EVD outbreak negatively affected the pig value chain i.e., the demand and supply of pigs and pork. Therefore, there is need to sensitize the stakeholders in the pig value chain on EVD in order to minimize the negative economic impacts associated with EVD outbreaks. Supplementary Information The online version contains supplementary material available at 10.1186/s12879-021-06337-8.

. The rst and biggest Ebola outbreak registered 425 cases with a case fatality rate (CFR) of 53%, while the central district of Luwero experienced two outbreaks in 2011 and 2012, respectively [9]. Considerably, Uganda remains vulnerable to Ebola outbreaks particularly in the wake of frequent outbreaks and possible spill over from her neighbours [6].
Besides human health, Ebola outbreaks affect agricultural production in the country. Ebola epidemics are associated with movement restrictions of people and their animals in and out of the affected communities, and disruption of the food value chain from the farm to consumers [10][11][12]. Considerably, families face belowaverage production levels, coupled with market disruptions lead to low purchasing power and reduced incomes [13]. In consequence, food insecurity, fear, social, political and economic turmoil escalate in the affected and atrisk communities.
Agriculture and specially, the pig industry in Uganda, is crucial role in supporting livelihoods by providing nutritional and income security for 1.1 million smallholder farmers [14]. The country's rapid growth of the pig population, at present standing at 1.1 million, is attributed to the high nancial returns, low cost of production and increased demand for pork [15]. On the whole, urban and peri-urban consumers account for 70% of consumed pork, accessed through the informal local butcheries and pork eateries referred to as 'pork joints' [16].
Evidently, the zoonotic epidemics such as Ebola cause wide economic impacts beyond the health sector. At present, information on the effect of Ebola outbreaks on the pig value chain in Uganda in general and Luwero district in particular is scanty. Likewise, little is known about the knowledge and beliefs of the rural communities along the pig value chain in Luwero with regard to Ebola outbreaks. Thus, this study assessed the knowledge, perceptions and effects of Ebola outbreaks along the pig value chain in Luwero District. In essence, this study results will guide future plans to mitigate food insecurity and other consequences associated with Ebola outbreaks in communities that rely on pig production.

Study design
This cross-sectional study was conducted in Luwero district, located at 0.8271° N, 32.6277° E in the Central Region of Uganda ( Figure.1) from May to July, 2014. Largely, the district's economic activity is based on rain-fed agriculture and agro-pastoral livestock husbandry practices for 85% of the population. Pig husbandry practiced by small holder farmers is mainly the traditional free range combined with tethering of the animals [17]. The district consists of thirteen sub-counties, of which Ebola occurred in Kakute village, Ssambwe parish, Nyimbwa subcounty in May, 2011 and Nakisamata village in Ngalonkulu parish, Zirobwe sub-county in December, 2012, [18].
One case was reported in Ngalonkalu parish involving a 12-year-old girl who later died of the disease [19]. In Ssambwe parish, the index case was a 32-year-old motorbike taxi rider Islamic man [20]. He died in hospital and was buried without following safe burial practice because the cause of death has not been established. In total six Ebola cases were con rmed, including four deaths [21].

Eligibility criteria
The study population involved actors along the pig value chain who ful lled the eligibility criteria. These were small holder farmers with at least 2 pigs. Pig traders involved in sale of pigs, owners of pork-eateries and consumers of pork. Also, households that had registered Ebola related illness or death were included in the study.
While small holder farmers with 2 pigs or less, had other livestock other than pigs, and respondents that did not consent or did not report Ebola related illness/death were not considered for the study.

Sampling strategy of study respondents
This study used both qualitative and quantitative methods. A structured questionnaire was used to collect quantitative data from individual respondents along the pig value chain. From each parish of Ssambwe and Ngalonkalu, an index villages where Ebola outbreak occurred, were purposefully selected. The criteria for selecting the remaining three villages in each parish was basing on a shared common border(s) with the index villages for Ebola and having a major road connecting the index to other villages. The adult population (> 18 years) in the two parishes was estimated at 6000 persons with 60% engaged in any form of livestock farming, and 40% (1440 persons) in the pig value chain [15]. A total of 227 respondents was targeted representing about 16% of population actively engaged in the pig value chain. The key aspects considered included respondents' sociodemographics, level of knowledge on Ebola, perception, and effects on pig production, trade and consumption of pork.
The qualitative study utilized focus group discussions (FGDs) and key informant interviews (KIIs) to get an indepth understanding of the provided data. A total of eight key informants that were conversant with the chronology of Ebola outbreaks in Luwero district were selected for in-depth interviews. These included the community development o cer, animal husbandry o cers, local council leaders, the district health o cer, and village health team members.
A total of four FGDs, two in each parish, composed of 8-10 respondents were conducted for the pig farmers, pig traders (butchers and pork eateries) and pork consumers. An open-ended interview guide was used by a moderator in the local language. FGDs were used to explore the knowledge, possible risk factors and effects of Ebola virus disease along the pig value chain. Deliberately, respondents from the two index villages, and women were among the participants. A digital recorder was used to capture the discussions which were later transcribed into English. An informed verbal consent was obtained prior to this. Each session lasted ninety minutes and ended when subsequent discussions produced no new information [22]. Each FGD was comprised of different categories of respondents which was anticipated to obtain contrasting perspectives that could enrich the discussion. Deliberate efforts were made to include females in the FGDs in the male-dominated community. Mixed gender groups tend to improve the quality of discussions and its outcomes [23].

Data management and statistical analysis
Qualitative data from recorded audios was transcribed verbatim and cross-checked to ensure completeness of data by experienced research associates. A thematic content analysis approach was used for analysis of key emergent themes [24].
Quantitative data was entered in Microsoft Excel and exported to SPSS version 22 for statistical analysis.
Descriptive analysis was used to summarize demographic characteristics of respondents.
Respondents' knowledge about Ebola Virus Disease was categorised as agree, disagree or I don't know.
Univariate analysis of categorical data was used in identi cation of potential risk factors to Ebola. The Chi-square test and Fischer's exact tests were used for cross tabulations. Co-e cient regression was used to investigate associations between the variables and the outcomes. Variables were considered signi cant at p-value < 0.05. Graph Pad Prism version 6 software was used to carry out repeated measures of analysis of variance (ANOVA) to test the effect on the pig sales at various levels in the pig value chain. Variables were considered signi cant at pvalue < 0.05.

Results
Respondents' socio demographic characteristics Table 1 summarizes the socio-demographic characteristics of 229 respondents that were involved in the study.
Slightly more respondents 121 (52.8%) were from Ssambwe parish, almost half 115 (50.2%) were males, most 141(61.6%) attained at least primary level education and majority 184 (80.4%) were farmers. Regarding involvement in pig value chain, most 142 (62%) of the respondents were pig farmers, and few 20 (8.7%) were pig traders.  Respondents' knowledge of clinical signs associated with Ebola Regarding clinical manifestation, bleeding from body openings, fever, vomiting blood and diarrhoea were the signs that most respondents highly associated with Ebola. However, other signs such as muscle pain, headache, rash and sore throat were not associated with Ebola (   With reference to the perceived risk factors and practices of acquiring Ebola virus, there was no correlation to all factors (Table 4). This was also in concurrence with one of the respondents quoted as in the FGD, "....…………we don't believe that one can contract Ebola from eating monkey meat, we have eaten bush meat for years and none of us has contracted Ebola." Effect of Ebola outbreak on the demand and sales of pigs and consumption of pork Out of 142 farmers, 55 (38.7%) believed that Ebola outbreaks affected demand of pigs (Table 5). Similarly, from a Focus Group Discussion, where a study participant was quoted, "…….. Ebola outbreaks greatly affected demand for pigs and other farm products. Traders from Kampala did not come to buy pigs during that Period, they feared to acquire Ebola disease." Out of the 20 traders, 17 (85%) were in agreement that Ebola reduced pig sales (Table 5). There was a signi cant reduction (P<0.001) in pork consumption during Ebola outbreaks. Out of the 67 consumers in this study, 15 (22.4%) did not consume pork during the Ebola outbreak (Table 5).
In concurrence with FGD, one respondent was quoted as, "……I feared to consume pork because I thought I would contract Ebola disease; we were told to avoid consuming any meat during that period." As in Table 6, there was a signi cant reduction (P = 0.001) in the average number of pigs sold during Ebola outbreak period. There was also a signi cant reduction (P = 0.04) in the average number of pigs bought by traders during the Ebola outbreak. Similarly, it is clearly seen that the number of pigs sold/ slaughtered by butcher men at pork eating places reduced signi cantly (P = 0.03). The unit cost of a kilogram of pork reduced during the outbreak. The income generated from pork sales reduced signi cantly during the outbreak when compared to the period before and after the outbreak (Table 6) Discussion Highly infectious emerging zoonotic diseases such as Ebola are a major cause of morbidity and mortality in Africa [25] as shown by the recent outbreaks in West and Central Africa [7,26]. The past six Ebola outbreaks in Uganda accentuate the country's vulnerability to further outbreaks, especially due to spill-over from the Democratic Republic of Congo, a neighbouring country prone to frequent Ebola outbreaks [27,28]. Also, Uganda's location in an area with great biodiversity, together with a fast-rising population density that comingle closely with animals are high risk factors for Ebola and other emerging zoonotic diseases [27,29].
Our focus on the Ebola outbreak with respect to the pig value chain, was prompted by Uganda's fast growing pig sector, that continues to register increasing pork consumption rates (average of 3.4 kg/person per year), the highest in East Africa [30]. Primarily, the national Ebola control programs are designed to protect the human health and remotely consider enterprises that support farmers' livelihoods. However, the neglect of the community's survival and sustenance options, especially the pig value chain could lead to adversative effects on livelihoods. With the recent zoonotic disease prioritization [31] and having the multi-sectoral-multi-agency One health strategic plan in place [32], Uganda has the opportunity to embrace a One Health approach while handling Ebola and other priority zoonotic diseases for the bene t of the community. Effective post-Ebola intervention plans in Uganda and elsewhere, must encompass the primary sources of livelihoods for such agro-pastoral communities in order to mitigate the negative impacts of the disease. For instance, farmers could be provided with breeding pigs, access to credit, and training for the small holder pig enterprises. This is in concurrence with strategies put in place in West Africa in lieu of efforts to stimulate social and economic recovery of vulnerable communities affected by Ebola [33]. Ebola control programs cognisant of the inter-sectoral collaboration that would assure sustainable pig production, supply and marketing for the small holder pig farmers in Uganda.
With regards to knowledge, a majority of the respondents were able to recall previous Ebola outbreaks in Luwero that occurred more than two years ago. This was contrary to common health care events, often not recalled accurately within a few months of occurrence [34]. Most likely, the impact of Ebola outbreaks associated with economic turbulence and social stigma in communities appear to last longer. Indeed, people tend to remember disastrous events associated with excessive changes in routine life, even after several years of the incident [35]. One such event was the introduction of 'safe burial' practices for Ebola victims [36]. Whereas this was needed to prevent disease spread, the communities misconstrued the practice as being disgraceful and demeaning to the dead. This is because the bereaved family members were denied a chance to pay their last respect to their dead, (pray, wash, dress, perfume the corpses) and mark the graves of the deceased [36]. In effect, this caused psychosocial trauma among relatives and community members that might have led to long-term memories which lasted for several years, even after the Ebola outbreak.
The fact that respondents were able to identify the major signs and symptoms of Ebola such as fever, vomiting, bleeding and diarrhoea was not unexpected. These were frequently shared on the information education and communication materials that were available in the community health centres and public spaces in Luwero [37].
Communities that are able to identify the main signs of Ebola can ably avoid the infections when such cases occur. However, other important signs which are frequently associated with common ailments such as malaria could not be identi ed, suggesting that the community could mistakenly handle Ebola patients with minimal or no precautionary measures. This is particularly possible if Ebola presents with nonspeci c symptoms that can easily be mistaken for other endemic diseases like malaria, typhoid fever, yellow fever or, even measles or meningococcemia in children [38]. Efforts should be put in place to educate the public to consider patients with such signs as suspected cases, especially when there is an Ebola outbreak in the community.
Besides, the health centres, a signi cant proportion of patients suspected to have Ebola could be treated by traditional healers or witchdoctors. A number of communities in Uganda and elsewhere in Africa still associate Ebola outbreaks with witchcraft [20]. Traditional healers have evolved with communities in Africa, and continue to offer rst line of health care services in Uganda for unknown ailments believed to be associated with witchcraft to appease the spirits [20,39]. Nonetheless, escalation of Ebola infections in West Africa and DRC was linked to traditional healers due to inability to adhere to preventative measures [40]. Training traditional healers on the basic preventive measure for Ebola and closely monitoring their operations would minimize spread of the disease in the community.
This study observed mixed perceptions on the transmission of Ebola with a majority linking transmission with eating of bats, monkey and bush meat. This was in conformity with the study conducted by [5,41,42], who noted that consumption of bush meat could contribute to infection. Conversely, the respondents did not consider eating of game meat as a high risk factor, probably because in earlier outbreaks that occurred in Kibaale district [43] and also in Luwero district [20] the victims could not be linked to eating of wild animals such as monkeys before falling ill. Meanwhile, as eating of bats or other wild animals could not be regarded as a high risk for acquiring Ebola, hunting and eating of game meat continues unabated, exposing the vulnerable communities to future Ebola infections. Besides hunting game meat for authentic values, most communities depend on wild game as source food and income security. Efforts to provide alternative livelihoods and food security could protect vulnerable communities from Ebola.
Our ndings showed that the unit cost of pork was signi cantly reduced during Ebola outbreak. Also, the average number of kilograms of pork sold as well as the nal income signi cantly reduced during the outbreak. This was in conformity with the assertions that beyond the public health impacts, emerging global infectious cause wider socioeconomic consequences often disregarded in impact assessments [44]. This could be attributed to the human health sector-based Ebola preparedness and prevention practices and the public behaviour of stigmatizing the pigs as the source of Ebola that led to avoidance of eating pork. Evidently, zoonotic infectious disease outbreaks have been associated with substantial economic impacts due to disruption of the livestock markets (44). In consequence, the reduced sales and incomes negatively impacted the household incomes. In accord with prior reports from West Africa, household and community incomes dropped signi cantly during the Ebola outbreak [45] leading to increased food insecurity and household vulnerability.
A plausible explanation for the reduced sales of pigs/pork was due to the misconception that pigs were the source of Ebola, creating fear of contracting the disease and avoiding eating of pork by the consumers. It was not clear why the community associated pigs with the Ebola. However, the two Ebola outbreaks in Luwero were apparently preceded by African swine fever (ASF) epidemics with wide spread mortalities of pigs in the community. Besides, pigs with ASF show signs of haemorrhages on the skin and in internal organs, labored breathing with frothy nasal discharges mixed with blood, and death of most affected animals within 2-5days [46]. Farmers have frequently observed reddening of the ears, clinical signs consistent with of ASF [47,48]; but more less the same sign (bleeding) is observed in human cases with Ebola. As a result, the ASF in pigs and Ebola in humans are erroneously believed to be caused by the same agent that originates from pigs. To clear this misconception, a deliberate social-anthropological understanding of the agro-pastoral communities with regard to such diseases should be given priority attention to minimize losses to the pig industry in future.

Conclusion
There was a signi cant reduction in pig sales, price per unit cost of a kilogram of pork and pork consumption. On health seeking behavior, traditional healers were preferred as the rst line of health care providers to Ebola patients. These may in the long run escalate the outbreak of Ebola since they do not adhere to preventative measures during their ritual practices. There was no correlation of the perceived risk factors to Ebola outbreak and therefore this indicated that probably another unknown factor may have led to Ebola outbreaks in Luwero district. However, the association of Ebola Virus Disease with pigs is detrimental and could affect the entire pig value chain in terms of pig production, pig sales and pork consumption.

Recommendations
There is need for a One health inter-sectoral collaboration of agricultural and health sectors to while designing preparedness and prevention programs for Ebola and other emerging zoonotic diseases in Uganda. Farmers should be sensitized that pigs are not a source of Ebola. Traditional healers should be trained and their role monitored during future Ebola virus disease outbreaks. Socio-anthropological studies need to be conducted in order to understand the way communities perceive Ebola Virus disease.
collection and writing the manuscript. GT and CDK participated in the conceptual design and analysis of data. MO and SM supervised the process of data collection, write up and reviewed the paper for substantial intellectual content. All the authors read and approved the nal version of the manuscript.
Map of Uganda showing Zirobwe and Nyimbwa sub-counties in Luwero District. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.