ORIGINAL ARTICLE
Open Access

Evaluation of Community Perceptions and Prevention Practices Related to Ebola Virus as Part of Outbreak Preparedness in Uganda, 2020

Joseph Musaazi, Apophia Namageyo-Funa, Victoria M. Carter, Rosalind J. Carter, Mohammed Lamorde, Rose Apondi, Tabley Bakyaita, Amy L. Boore, Vance R. Brown, Jaco Homsy, Joanita Kigozi, Aybüke Koyuncu, Maria Sarah Nabaggala, Vivian Nakate, Emmanuel Nkurunziza, Daniel F. Stowell, Richard Walwema, Apollo Olowo and Mohamed F. Jalloh
Global Health: Science and Practice June 2022, 10(3):e2100661; https://doi.org/10.9745/GHSP-D-21-00661

Key Findings

  • Individuals in Ebola virus disease (EVD) high-risk districts reported greater EVD message exposure and comprehensive knowledge compared to individuals in EVD low-risk districts.

  • Individuals with comprehensive knowledge of EVD were more likely to self-report frequent handwashing with soap, avoiding physical contact with suspected Ebola patients, and avoiding funerals/burials that involve physical contact with a corpse.

  • Rejection of EVD misconceptions was not significantly associated with self-reported EVD-related prevention practices.

Key Implications

  • Efforts to promote EVD comprehensive knowledge may facilitate the uptake of EVD prevention practices.

  • Risk communication and community engagement efforts continue to be critical for EVD preparedness and response efforts, especially in high-risk geographic areas.

ABSTRACT

Introduction:

During the 2018–2020 Ebola virus disease (EVD) outbreak in the Democratic Republic of the Congo, risk communication and community engagement (RCCE) were prioritized in geographic areas in Uganda considered at high risk of introduction of EVD. To inform EVD preparedness in Uganda, we evaluated community perceptions and prevention practices related to EVD in 6 districts in Uganda.

Methods:

In March 2020, we conducted a population-based survey in 6 purposively selected districts in Uganda using multistage cluster sampling. We examined differences between districts classified as high- versus low risk for EVD in terms of their message exposure from RCCE; risk perception; and EVD knowledge, attitudes, and prevention practices.

Results:

A total of 3,485 respondents were interviewed (91% response rate). EVD message exposure was more common in the high- versus low-risk districts. EVD risk perceptions were low overall but greater in the high- versus low-risk districts. Comprehensive knowledge was significantly greater in the high- versus low-risk districts (adjusted prevalence ratio [aPR] 1.61, 95% confidence interval [CI]=1.35, 1.93). Respondents' engagement in all 3 EVD prevention practices (frequent handwashing with soap, avoiding physical contact with suspected Ebola patients, and avoiding burials involving contact with a corpse) was very low (4%). However, respondents with comprehensive knowledge were more likely to engage in all 3 EVD prevention practices compared to respondents without comprehensive knowledge (aPR 1.87, 95% CI=1.08, 3.25).

Conclusion:

Our findings suggest that while RCCE efforts as part of EVD outbreak preparedness may have contributed to higher EVD knowledge in the targeted high-risk districts, uptake of prevention behaviors was similarly low across districts. In a non-outbreak setting, implementing targeted RCCE strategies may not be sufficient to motivate people to adopt protective behaviors in the absence of a high threshold of perceived threat such as in an active outbreak.

INTRODUCTION

Ebola virus disease (EVD) is a rare and deadly infectious hemorrhagic fever disease. Four of the 6 known species of ebolavirus can cause disease in humans.1 Since 1976, EVD outbreaks have been recorded sporadically in East, Central, and West Africa.1,2 Uganda has experienced 6 documented outbreaks of EVD since 2000 that can be attributed to 3 species of ebolaviruses (Zaire, Sudan, and Bundibugyo).3 Despite substantial advances in clinical management, the case fatality ratio (CFR) for EVD remains high, reaching more than 60% for Zaire ebolavirus in 2018–2020.3 The public health response to manage EVD outbreaks requires substantial resources for coordination, surveillance, laboratory, case management, infection prevention and control, vaccination, and risk communication and community engagement (RCCE).

Human behavior can influence both the spread and containment of EVD outbreaks. Understanding how community dynamics impact behaviors is important for developing and implementing effective RCCE activities during EVD outbreaks. Assessments of the public's knowledge, attitudes, and practices (KAP) were conducted during the 2014–2016 EVD outbreak in West Africa4–7 and the 2018–2020 outbreak in the Democratic Republic of the Congo (DRC).8,9 These assessments guided the development and implementation of evidence-based RCCE strategies and other interventions to reduce transmission and end the outbreaks.10,11 For instance, KAP studies in West Africa provided information to guide culturally acceptable modifications to care-seeking behaviors and burial rituals involving contact with dead bodies, which were critical to containing the outbreak.12–16 EVD KAP studies have also helped identify misconceptions that have hindered EVD prevention and treatment efforts.4,6,17 In DRC, KAP surveys and other rapid qualitative assessments similarly informed outbreak response strategies.18 During the 2018–2020 outbreak, countries bordering DRC also conducted various community-based assessments to inform context-specific preparedness efforts to prevent and respond to possible importation events.19

Understanding how community dynamics impact behaviors is important for developing and implementing effective RCCE activities during EVD outbreaks.

On June 11, 2019, the Uganda Ministry of Health (MOH) declared an EVD outbreak in the Kasese district in the South Western region of Uganda.20 The index case and 2 secondary infections were all imported cases linked to the tenth documented EVD outbreak in DRC. The Uganda MOH and response partners promptly responded to and contained this EVD outbreak. As part of its preparedness efforts, Uganda prioritized targeted interventions in 30 districts considered to be at high or moderate risk of introduction of EVD cases from Eastern DRC, where the 2018–2020 DRC EVD outbreak was centered.21 High-risk districts were characterized as sharing a border with DRC and having direct access to the affected areas in DRC; moderate-risk districts shared a border with DRC but had no access to the affected areas in DRC. Low-risk districts included the rest of the country except for Kampala and Wakiso, which were designated as high-risk districts. Interventions included community-based surveillance, RCCE using mass media and community-based channels, point-of-entry screenings, infection prevention measures in health facilities, and establishment of holding centers for suspected EVD patients.21

We aimed to assess community perceptions and prevention practices related to EVD as part of EVD outbreak preparedness efforts in Uganda during an ongoing outbreak in DRC. To understand the extent of outbreak preparedness RCCE, we explored differences in message exposure and KAP outcomes between districts classified by the Government of Uganda as high versus low risk for EVD, in the context of high-risk districts more heavily targeted with RCCE than the low-risk districts.

METHODS

Study Design

In March 2020, we conducted a cross-sectional population-based KAP survey in 4 high-EVD-risk and 2 low-EVD-risk districts in Uganda. We purposively selected the 6 districts using the EVD risk classification of districts developed by the Uganda MOH.21 Selection criteria included districts that were considered high risk for EVD importation, close to the Uganda border, or a major business hub in Uganda. The selected high-risk districts included 3 districts that share a border with the DRC nearest to the epicenter of the EVD outbreak (Kasese, Kisoro, and Arua) and the capital, Kampala, which is the major destination of many travelers entering Uganda by any transportation means and the country's busiest business hub. The 2 low-risk districts selected share borders with Kenya and South Sudan (Busia and Lamwo, respectively) and are major business border crossings to and from Uganda. At the Uganda MOH's request, the sample size for this survey was powered to provide district-specific estimates for these 6 districts to inform district-level preparedness efforts. Based on sample size calculations, we estimated that 640 eligible individuals from 20 clusters were required to be approached for consent in each of the 6 districts— for a total target sample of 3,840 participants after adjusting for an expected 25% level of non-responsiveness. Details of the sample size calculations are in the study protocol (Supplement 1).

Selection of Households and Respondents

Our recruitment targets were household heads (50%), non-household head women aged 25 years and older (25%), and young persons aged 15–24 years (25%). This was to ensure diversity in the sample and include a representative number of women and adolescents/youths. To reach these targets, we randomly selected 20 clusters within districts as defined by the Uganda Bureau of Statistics and enumerated households in the selected clusters with help from local stakeholders including community leaders. Systematic random sampling was used to select up to 16 households per cluster. Households were marked as unreachable after 3 attempts to contact them on separate days in the same week. We did not replace unreachable households because this was accounted for in the sample size estimation. Two eligible individuals were approached in each selected household—first the household head followed by a second randomly selected household member who was either a woman aged 25 years or older or a young person (male or female) aged between 15 and 24 years. We alternated these second household members each time we approached a new household (if available).

Questionnaire Development and Data Collection

The survey questionnaire was adapted from EVD KAP assessments in Sierra Leone and Guinea (Supplement 2).4,6 The principal investigator led the adaptation of the survey questionnaire and modified some questions to reflect the Ugandan context. The survey questionnaire included questions on respondents' demographic characteristics, EVD awareness and knowledge, risk perception, and self-reported EVD prevention practices. Trained research assistants with previous survey experience in Uganda conducted face-to-face interviews with respondents and collected data electronically using computer tablets programmed with Open Data Kit.22 Interviews were conducted in local languages and lasted approximately 1 hour on average. We translated the informed consent and assent forms into local languages. By signing the consent forms, participants were aware that the information collected would be kept confidential and would be used to inform decision making on EVD preparedness.

Data Analysis

We analyzed the data in Stata version 16.1 Special Edition (StataCorp). We created analytical variables to: (1) enable comparison of data collected from high- versus low-risk districts, (2) account for pre-survey self-reported exposure to EVD prevention messages, and (3) measure respondents' knowledge of all 3 key measures of Ebola prevention and treatment as well as ability to reject 3 key misconceptions about EVD or lack thereof. The following analytical variables were created: (1) a new binary variable to indicate district risk category (i.e., high-districts (coded 1) versus low-risk districts (coded 0)); (2) binary variable to indicate any EVD prevention message exposure (i.e., coded 1 if received any EVD prevention message before the survey and coded 0 otherwise); (3) a composite binary variable to indicate knowledge of all 3 key measures of Ebola prevention and treatment (i.e., coded 1 for knowing that EVD is preventable by avoiding physical contact with infected corpses, early medical care reduces household transmission, and early medical care increases chance of survival; and coded 0 for not knowing all 3 knowledge measures); (4) a binary composite variable to indicate a respondent's ability to reject 3 key misconceptions about EVD (i.e., transmissible by air, spiritual healers can successfully treat EVD, and traditional healers can successfully treat EVD); (5) a binary composite comprehensive knowledge variable was then created to indicate knowledge of all 3 measures of prevention and treatment while also rejecting all 3 EVD misconceptions (coded 1 for having comprehensive knowledge and 0 for not); and (6) a binary composite variable was calculated for self-reporting of all 3 key Ebola prevention practices (i.e., coded 1 for citing more frequent handwashing with soap, avoiding physical contact with suspected Ebola patients, and avoiding funerals/burials that involve physical contact with a corpse; and coded 0 for not citing all these measures). The selected key items were based on the published literature from West Africa showing these 3 items as having performed well in serialized KAP surveys in Sierra Leone in addition to Guinea.4,6

Analyses were performed as frequencies, proportions, and 95% confidence intervals (CI) of proportions comparing high- versus low-risk districts. Results were disaggregated to show district-specific estimates (Supplement 3). We accounted for the survey design at analysis by using inverse probability survey weights. To determine associations, we fitted 3 modified Poisson regression models using generalized estimating equations with exchangeable correlation structure to account for correlation of outcomes due to clustering of respondents.23 In the first model, we examined sociodemographic correlates (district risk category, sex, age, education, and religion) of self-reported exposure to any EVD messages in the 6 months before the interview. In the second model, we examined the association between EVD message exposure and comprehensive knowledge, adjusting for demographic covariates previously listed. In the third model, we examined the association between EVD comprehensive knowledge and all 3 prevention practices, adjusting for demographic covariates previously listed. In sensitivity analysis, we modified the third model so that the 2 components of comprehensive knowledge (correct knowledge and rejection of misconceptions) were entered as 2 separate predictor variables. Statistical significance testing was based on a 2-tailed Wald-type test.

We chose a modified Poisson regression approach because the outcomes were common (i.e., >10% prevalence). When the prevalence of the outcome is greater than 10% (as for message exposure), the odds ratio exaggerates the association and is not approximately equal to the prevalence ratio or relative risk; hence we used Poisson regression with robust standard errors to get the adjusted prevalence ratio (aPR) with their 95%CI.23 We used variance inflation factors to evaluate multicollinearity in fitted models, wherein variance inflation factors >10 were indicative of severe multicollinearity. We examined the effect of missing data on regression model estimates by refitting models after completing missing data using the multiple imputation chained equation approach.24

Ethical Approval

Ethical approval was obtained from the Uganda Virus Research Institute Research Ethics Committee (UVRI REC GC/127/19/11/756), and the survey was registered with the Uganda National Council for Science and Technology (HS531ES). The Human Subjects Office of the Center for Global Health at the U.S. Centers for Disease Control and Prevention approved the survey as a public health evaluation. All respondents provided individual written or thumb-printed consent (if illiterate) before the interview. For respondents aged 15–17 years, a parent, guardian, or household head provided consent, and the respondent assented. At the end of the interviews in each household, we gave a bar of soap and an EVD education flyer.

RESULTS

Respondents' Sociodemographic Characteristics

A total of 3,485 (91%) of the 3,840 individuals approached in the 6 districts consented to participate in the survey: 634 from Busia, 593 from Lamwo, 506 from Arua, 511 from Kampala, 612 from Kisoro, and 629 from Kasese. Across all districts, 51% of the respondents were household heads, 29% were non-household head women aged 25 years and older, and 20% were young people aged 15–24 years. Sixty percent of respondents were female. Sex, educational level, and occupation were statistically significantly different between low-risk and high-risk districts (P<.001, =.022, and <.001, respectively) (Table 1).

View this table:
TABLE 1.

Characteristics of Knowledge, Attitudes, and Practices Survey Respondentsa in 6 Districts, Uganda, 2020

EVD Awareness and Risk Perceptions

Overall, 47% of all respondents were aware of the DRC outbreak at the time of the interview, more frequently so in the high- versus low-risk districts (49% versus 37%; P=.001, Table 2). Approximately half of all respondents (52%) perceived themselves to be at some risk of getting EVD; 22% (95% CI=18%, 26%) in the high-risk districts perceived themselves to be at high risk compared to 7% (95% CI=5%, 11%) in the low-risk districts (P<.001) (Table 2).

View this table:
TABLE 2.

Exposure to EVD Messages and Perceptions of Information Gap in 6 Districts in Uganda, 2020

EVD Message Exposure, Sources of Information, and Information Gaps

Half of all respondents were never exposed to EVD messages in the 6 months before the interview, which was more common in the low-risk districts (72%) compared to the high-risk districts (46% P<.001, Table 2). EVD prevention messages were more common in the high- versus low-risk districts for the following: handwashing (38% versus 15%, P<.001); avoiding physical contact with people (35% versus 15%, P<.001); avoiding unsafe burials (20% versus 8%, P<.001). Radio was the most common source of EVD information in both low- and high-risk districts (65%) followed by television (35%) and churches or mosques (27%) and by community meetings (26%). Respondents from low-risk compared to districts were more likely to want more information about EVD prevention (72% versus 63%, P=.02), transmission (61% versus 43%, P<.001), and signs and symptoms (60% versus 37%, P<.001) (Table 2).

EVD Knowledge and Prevention Practices

Overall, 5% of all respondents were able to mention a total of 4 key signs/symptoms of EVD (fever, severe headache, diarrhea, and vomiting) without any prompting by the interviewer (Table 3). Fever was more frequently mentioned as an EVD symptom in high- versus low-risk districts (46% versus 31%; P=.001). Overall, 63% of respondents had knowledge of the 3 key measures of EVD prevention and treatment, and such knowledge was greater in the high- versus low-risk districts (65% versus 55%, P=.002). The ability to reject the 3 measures of EVD misconceptions was more common in the high- versus low-risk districts (73% versus 57%, P<.001). Taken together, comprehensive knowledge was greater in the high- versus low-risk districts (48% versus 32%, P<.001) (Table 3).

View this table:
TABLE 3.

Ebola-Related Knowledge, Attitudes, Intentions, and Self-Reported Behaviors in 6 Districts in Uganda, 2020

Attitudes and Behavioral Intentions Toward Care and Burial

If a family member was suspected to have EVD, 52% of all respondents intended to take the person to a health facility (50% in high-risk districts and 65% in low-risk districts, P=.003; Table 3) while 43% of all respondents intended to report the suspected case to the health authorities. Overall, 3% of respondents mentioned an intention to provide care at home, and <0.1% mentioned an intention to hide the EVD suspected family member. In a hypothetical scenario, most respondents (62%) said they would welcome an EVD survivor back into the community.

Intention to accept safe alternatives to traditional burials if a family member died of EVD was high overall (81%) (Table 3). Less than half of all respondents (39%) reported having attended a funeral/burial within the past month of the interview (for a death of any cause, not specific to EVD) (Table 3). Among those who attended a recent funeral/burial, 11% reported they directly had some physical contact with the corpse and 51% said that they observed attendees touching each other (e.g., hugging and shaking hands). Respondents in the low-risk districts were more likely to report that attendees touched each other (61%) compared to those in the high-risk districts (49%, P=.036) (Table 3).

Self-Reported Prevention Practices

Self-reported uptake of any EVD prevention practice was 67% overall, without a significant difference between high- versus low-risk districts (66% versus 72%, P=.180, Table 3). Among those who reported any uptake, 57% mentioned handwashing frequently with soap and water, 23% mentioned avoiding physical contact with suspected EVD patients, and 13% mentioned avoiding burials involving contact with a corpse. Uptake of all 3 of these prevention practices was very low in both high- and low-risk districts (3% versus 4%, respectively) (Table 3).

Multivariable Models for EVD Message Exposure, Comprehensive Knowledge, and Uptake of Prevention Practices

Association Between Sociodemographic Variables and EVD Message Exposure

EVD message exposure was significantly greater in the high-risk districts (aPR: 2.46; 95% CI=2.09, 2.91) and among male respondents (aPR: 1.09; 95% CI=1.03, 1.14) but less prevalent among respondents with no formal education (aPR: 0.86; 95% CI=0.78, 0.94), adjusting for district category, sex, age, educational level and religion (Table 4).

View this table:
TABLE 4.

Sociodemographic Correlates of EVD Message Exposure in 6 Districts in Uganda, 2020

Association Between EVD Message Exposure and Comprehensive Knowledge

There was no sufficient evidence of association between EVD message exposure and comprehensive knowledge at complete case analyses (aPR: 1.11; 95% CI=0.99, 1.24, Table 5), but the association was significantly positive after completing missing values using multiple imputation (aPR: 1.12, 95% CI=1.01, 1.24). Comprehensive knowledge was significantly greater in the high-risk districts (aPR: 1.61; 95% CI=1.35, 1.93), and among male respondents (aPR: 1.18; 95% CI=1.08, 1.28), but less prevalent among respondents with no formal education (aPR: 0.87; 95% CI=0.76, 0.98), adjusting for EVD message exposure, district category, sex, age, educational level, and religion (Table 5).

View this table:
TABLE 5.

Association Between EVD Message Exposure and Comprehensive Knowledge, Adjusted for Sociodemographic Characteristics - Uganda, 2020

Association Between EVD Comprehensive Knowledge and Uptake of Prevention Practices

Comprehensive knowledge was positively significantly associated with self-reporting of the 3 EVD prevention practices (aPR: 1.87; 95% CI=1.08, 3.25, Table 6), adjusting for district category, sex, age, educational level, and religion. None of the demographic variables were significantly associated with uptake of the 3 prevention practices. In sensitivity analyses when we examined the 2 domains of comprehensive knowledge as separate covariates in the model, knowing the 3 measures of EVD prevention and treatment (domain 1) was significantly associated with uptake of the 3 prevention practices (aPR: 2.16; 95% CI=1.08, 4.32, Table 6); however, the ability to reject all 3 EVD misconceptions (domain 2) was not significantly associated with the self-reported prevention practices. Multicollinearity was not observed in any of the multivariable models fitted.

View this table:
TABLE 6.

Association Between EVD Comprehensive Knowledge and Uptake of Prevention Practices, Adjusted for Sociodemographic Characteristics in 6 Districts in Uganda, 2020

DISCUSSION

Our population-based evaluation revealed our empirical understanding of the differences between the high- versus low-risk districts in Uganda in terms of EVD perceptions and prevention practices. Overall, although EVD exposure and EVD comprehensive knowledge were greater in the high- versus low-risk districts, uptake of prevention practices was low across all districts. Adjusted analysis suggested that comprehensive knowledge was associated with EVD message exposure and EVD prevention practices. Deeper analysis of comprehensive knowledge revealed that while knowledge of EVD prevention and treatment was associated with EVD prevention practices, this was not the case with the respondents' rejection of EVD misconceptions. Our findings may reflect the targeted approach implemented by the Government of Uganda to prioritize preparedness efforts in the high-risk districts.21

Deeper analysis of comprehensive knowledge revealed that while knowledge of EVD prevention and treatment was associated with EVD prevention practices, this was not the case with respondent rejection of EVD misconceptions.

To our knowledge, this is the first population-based evaluation in Uganda to examine differences between high- versus low-risk districts after a year of intensified EVD preparedness efforts including RCCE interventions. Given the ongoing threat of EVD outbreaks in Uganda, our findings could serve as a baseline for future evaluations of the population-level reach and impact of RCCE interventions for EVD prevention in Uganda. The findings also shed light on areas for additional RCCE activities for consideration. In low-risk districts, for instance, we found that respondents wanted information about EVD prevention, transmission, signs/symptoms, and how to care for an EVD patient, possibly reflecting the RCCE focus placed on high-risk districts. Respondents from high-risk areas were more likely to mention fever as an EVD symptom, which may have been reinforced by body temperature screenings in these areas.21 Overall, less than 1 in every 10 respondents knew 4 of the signs and symptoms of EVD (fever, severe headache, diarrhea, and vomiting). Greater investments in the high-risk compared to the low-risk districts explain the greater EVD exposure and EVD comprehensive knowledge among respondents in the high- versus low-risk districts. Having a greater sense of personal risk and being in closer proximity to the outbreak offers another explanation. The reported low EVD message exposure in Kampala, a high-risk district, was noteworthy given it was heavily targeted with intensified RCCE efforts.

Mistrust in authorities and the health system has been shown to influence care-seeking behaviors and uptake of promoted practices.25–27 For instance, in past EVD outbreaks in West Africa and DRC, there were instances when family members hid their loved ones at home instead of reporting them to health authorities and secretly buried them instead of requesting a safe and dignified burial by specially trained teams. In our evaluation, respondents expressed willingness to accept safe and dignified burial measures if a family member died at home and intended to report an EVD suspected family member to the health authorities or take them to a health facility. However, it should be noted that in Sierra Leone, for example, although a similar assessment showed low intention to hide EVD suspected family members5 in practice, family members continued to hide sick family members and many times cared for them at home instead of reporting to health authorities. Caring for a sick family member at home is not necessarily synonymous with hiding them from authorities. There are situations where families may report suspected cases to the authorities but are unable to get a timely ambulance service.28 While waiting for medical help to arrive, families may feel compelled to provide care for their loved ones,29 which shows a need for RCCE efforts to educate on how to safely provide care as they wait for help to come. During EVD outbreaks, it is not only important to strengthen public confidence in the outbreak response services but also equally important to ensure that the promoted services can meet the public's demand and expectations for quality—including showing respect to communities, families, and their loved ones.

Although respondents were exposed to EVD messages and had EVD knowledge, many did not engage in EVD prevention practices. During all EVD outbreaks in sub-Saharan Africa, traditional burial practices involving contact with corpses have been a major source of virus transmission.13,16 In the 2018–2020 DRC outbreak, people living in unaffected areas in DRC struggled to make the connection between the spread of EVD and traditional funeral practices.30 Our findings showed that although Uganda was on high alert for EVD, traditional practices for funerals and burials continued both in the high- and low-risk districts. Most respondents intended to accept safe alternatives to traditional funerals and burials if a family member died of EVD. We found that nearly 4 in every 10 respondents had participated in a traditional funeral or burial in the past month. One in every 10 respondents reported their own direct contact with the corpse, which was a more common practice in the low-risk districts. This finding suggests that although respondents were knowledgeable, they may have felt confident that the deaths were not EVD. A possible reason for this confidence is the low number of EVD-related deaths in Uganda at the time.3 As was demonstrated during the 2014–2016 EVD outbreak in Sierra Leone, RCCE can be helpful in socializing and explaining the need for safe burials when guidance changes.13 Another possibility is that transferring knowledge alone is not sufficient if not supported by actual engagement and behavioral drill practices by the target community, as has been well documented.31

Although respondents were exposed to EVD messages and had EVD knowledge, many did not engage in prevention practices.

Misconceptions about EVD in communities can play a role in individuals' understanding of the transmission, prevention, and treatment of EVD. Studies on misconceptions about EVD have found beliefs such as spiritual prayer protects one from EVD and traditional healers treat EVD better than medical doctors.17 National EVD KAP surveys in Sierra Leone (2014) and Guinea (2015) found that participants held misconceptions about EVD transmission such as EVD can be transmitted by mosquitoes.4,6 We found that the ability to reject the 3 measures of EVD misconceptions (i.e., transmissible by air, curable by spiritual healers, and curable by traditional healers) was more common in the high- versus low-risk districts. We also found that the rejection of EVD misconceptions was not significantly associated with the uptake of EVD prevention practices. Together with our findings on burial practices, this points to the fact that science-based knowledge could coexist with religious and traditional beliefs32 and that misconceptions do not necessarily translate into lower uptake of protective behaviors. Our findings suggest more emphasis should be placed on integrating actionable knowledge and addressing misconceptions through community engagement efforts that can translate into safe and effective prevention and care practices in both high- and low-risk contexts.33

The Uganda MOH's RCCE efforts used television, radio, newspapers, posters, and door-to-door health education as part of EVD preparedness.21 While mass media may help to quickly raise awareness and create knowledge about EVD before or during the early part of an outbreak,4 improvements in attitudinal and behavioral outcomes take much longer and require more targeted comprehensive community engagement.11

Limitations

Our evaluation has several limitations. Due to the cross-sectional design, we cannot establish any causal relationships in the associations that we documented in our results. The binary variables used in the analysis may not capture the latent processes driving the perceptions and practices of the respondents. In addition, respondents may have provided socially desirable responses, especially due to their possible awareness of public health messages. Social desirability may have been more pronounced in the high-risk districts where EVD messages were widely disseminated which could have resulted in an overestimation of the association between district risk level and self-reported prevention practices. Some practical aspects of the survey's administration posed limitations. For instance, because the survey was conducted during working hours, some sampled households in Kampala did not have any eligible household members available and had to be skipped. Furthermore, female respondents were over-represented in our sample compared to the Uganda population possibly because more females were available at the time of our visit or more males declined to be interviewed, which we are unable to discern. Lastly, our results may not be generalizable to all districts categorized as high- versus low-risk for EVD in Uganda because we only evaluated 6 districts. However, a strength of our evaluation is that we have generalizable results for each of the 6 districts to inform future targeted EVD preparedness at the district level.

CONCLUSION

EVD comprehensive knowledge and message exposure were more common in high- versus low-risk districts following RCCE efforts as part of EVD outbreak preparedness in Uganda. The RCCE efforts in Uganda included extensive use of the media and door-to-door education to share information on EVD signs and symptoms, seeking care, and safe burials. Our findings suggest that while RCCE efforts may have contributed to higher knowledge in the targeted high-risk districts, uptake of prevention behaviors was similarly low across districts. In a non-outbreak setting, implementing targeted RCCE strategies may not be sufficient to motivate people to adopt protective behaviors in the absence of a perceived threat such as in an active outbreak. Additional strategies may be required for comprehensive community engagement in prevention practices.

Acknowledgments

The authors acknowledge the contributions of the following staff from the Infectious Diseases Institute: Gloria Akol, Lydia Akori, Angella Alapo, Jeff Amanya, Andrew Amara, Ogwal Ambrose, Peter Anabo, Robert Angurini, Heddy Anguyo, Fred Anguzu, Angella Apecho, Winfred Oprah Arach, Patricia Asiimwe, Sharon Atim, Martin Atugonza, Jackline Atukunda, Saviour Ayikoyo, Josephine Bagonza, Noah Bagonza, Moses Baluku, Richard Maliyamungu Bayo, Valentine Birungi, Doreen Bitwayi, Loice Boonabaana, Tom Duhimbaze, Patrick Eradu, Constance Iradukunda, Harriet Itakariot, Catherine Kyarisima, Moses Mabonga, Uzziah Muthahinga Masereka, Dezanje Mbambu, Joshua Muhindo, Vincent Muhindo, Allan Mutabazi, Nuhuh Mutebi, Stella Nabatyanga, Flavia Nabuuma, Lydia Nagaddya, Moureen Juma Namada, Florence Alanyo Nighty, Kenneth Nkumire, Simeo Ochieng, Opiti Abraham Okello, Jasper Billy Okwir, Moses Osinya, Lydia Nadera Otimong, Deutschmann Paadeh, Paula Tabu, Hilder Tindimwebwa, Brenda Wagaba, and Collin Wanyama Wandera. We thank the following staff at the district health offices in Uganda: Dr. Yusufu Baseke, Dr. Drileba Bishop, Dr. Stephen Nsabiyunva, Dr. Daniel Okello, Dr. Charles Akiya Oyoo, and Dr. Dick Wamayena. Finally, the authors thank the over 3,000 respondents who agreed to participate in the survey.

Funding

The U.S. Centers for Disease Control and Prevention funded the Infectious Disease Institute to implement the survey (Cooperative Agreement #NU2GGH001744).

Disclaimer

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.

Author contributions

MFJ, VC, VRB, ML, JH, RC, EN, JM, RA, JK, RW, MSN, and TB designed the assessment. AO, ML, ANF, RC, RA, JM, EN, DFS, VN, MSN, RW, and JK led the training of the data collection teams with remote support from MFJ and AK. AO, ML, MSN, EN, JM, JK, RA, ALB, RW, and DFS supervised and oversaw the data collection. JM analyzed the data with support from ANF and MFJ. All coauthors contributed to the critical interpretation of the results. JM, ANF, and MFJ led the writing of the manuscript with substantial contributions from all co-authors. All coauthors reviewed and approved the final version of the article.

Competing interests

None declared.

Footnotes

  • ↵* Co-first authors.

Notes

Peer Reviewed

First published online: June 6, 2022.

Cite this article as: Musaazi J, Namageyo-Funa A, Carter VM, et al. Evaluation of community perceptions and prevention practices related to Ebola virus as part of outbreak preparedness in Uganda, 2020. Glob Health Sci Pract. 2022;10(3): e2100661. https://doi.org/10.9745/GHSP-D-21-00661

  • Received: September 25, 2021.
  • Accepted: March 29, 2022.
  • Published: June 29, 2022.

This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly cited. To view a copy of the license, visit https://creativecommons.org/licenses/by/4.0/. When linking to this article, please use the following permanent link: https://doi.org/10.9745/GHSP-D-21-00661

REFERENCES

  1. 1.↵
    What is Ebola virus disease? Centers for Disease Control and Prevention. Accessed May 4, 2022. https://www.cdc.gov/vhf/ebola/about.html
  2. 2.↵
    1. Murray MJ
    . Ebola virus disease: a review of its past and present. Anesth Analg. 2015;121(3):798–809. doi:10.1213/ANE.0000000000000866. pmid:26287303
    OpenUrlCrossRefPubMed
  3. 3.↵
    History of Ebola outbreaks. Centers for Disease Control and Prevention. Accessed May 4, 2022. https://www.cdc.gov/vhf/ebola/history/chronology.html#anchor_1526565058132
  4. 4.↵
    1. Jalloh MF,
    2. Sengeh P,
    3. Monasch R,
    4. et al
    . National survey of Ebola-related knowledge, attitudes and practices before the outbreak peak in Sierra Leone: August 2014. BMJ Glob Health. 2017;2(4):e000285. doi:10.1136/bmjgh-2017-000285. pmid:29259820
    OpenUrlAbstract/FREE Full Text
  5. 5.↵
    1. Jalloh MF,
    2. Bunnell R,
    3. Robinson S,
    4. et al
    . Assessments of Ebola knowledge, attitudes and practices in Forecariah, Guinea and Kambia, Sierra Leone, July-August 2015. Philos Trans R Soc Lond B Biol Sci. 2017;372(1721):20160304. doi:10.1098/rstb.2016.0304. pmid:28396475
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. Jalloh MF,
    2. Robinson SJ,
    3. Corker J,
    4. et al
    . Knowledge, attitudes, and practices related to Ebola virus disease at the end of a national epidemic–Guinea, August 2015. MMWR Morb Mortal Wkly Rep. 2017;66(41):1109–1115. doi:10.15585/mmwr.mm6641a4. pmid:29049279
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Li W,
    2. Jalloh MF,
    3. Bunnell R,
    4. et al
    . Public confidence in the health care system 1 year after the start of the Ebola virus disease outbreak–Sierra Leone, July 2015. MMWR Morb Mortal Wkly Rep. 2016;65(21):538–542. doi:10.15585/mmwr.mm6521a3. pmid:27254016
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Vinck P,
    2. Pham PN,
    3. Bindu KK,
    4. Bedford J,
    5. Nilles EJ
    . Institutional trust and misinformation in the response to the 2018–19 Ebola outbreak in North Kivu, DR Congo: a population-based survey. Lancet Infect Dis. 2019;19(5):529–536. doi:10.1016/S1473-3099(19)30063-5. pmid:30928435
    OpenUrlCrossRefPubMed
  9. 9.↵
    Ebola outbreak in the Democratic Republic of the Congo declared a public health emergency of international concern. World Health Organization. Accessed May 4, 2022. https://www.who.int/news/item/17-07-2019-ebola-outbreak-in-the-democratic-republic-of-the-congo-declared-a-public-health-emergency-of-international-concern
  10. 10.↵
    1. Bedrosian SR,
    2. Young CE,
    3. Smith LA,
    4. et al
    . Lessons of risk communication and health promotion–West Africa and United States. MMWR Suppl. 2016;65(3):68–74. doi:10.15585/mmwr.su6503a10. pmid:27386834
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Bedson J,
    2. Jalloh MF,
    3. Pedi D,
    4. et al
    . Community engagement in outbreak response: lessons from the 2014–2016 Ebola outbreak in Sierra Leone. BMJ Glob Health. 2020;5(8):e002145. doi:10.1136/bmjgh-2019-002145. pmid:32830128
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Gillespie AM,
    2. Obregon R,
    3. El Asawi R,
    4. et al
    . Social mobilization and community engagement central to the Ebola response in West Africa: lessons for future public health emergencies. Glob Health Sci Pract. 2016;4(4):626–646. doi:10.9745/GHSP-D-16-00226. pmid:28031301
    OpenUrlAbstract/FREE Full Text
  13. 13.↵
    1. Lee-Kwan SH,
    2. DeLuca N,
    3. Bunnell R,
    4. Clayton HB,
    5. Turay AS,
    6. Mansaray Y
    . Facilitators and barriers to community acceptance of safe, dignified medical burials in the context of an Ebola epidemic, Sierra Leone, 2014. J Health Commun. 2017;22(sup1):24–30. doi:10.1080/10810730.2016.1209601. pmid:28854130
    OpenUrlCrossRefPubMed
  14. 14.
    1. Fallah M,
    2. Dahn B,
    3. Nyenswah TG,
    4. et al
    . Interrupting Ebola transmission in Liberia through community-based initiatives. Ann Intern Med. 2016;164(5):367–369. doi:10.7326/M15-1464. pmid:26746879
    OpenUrlCrossRefPubMed
  15. 15.
    1. Richards P,
    2. Amara J,
    3. Ferme MC,
    4. et al
    . Social pathways for Ebola virus disease in rural Sierra Leone, and some implications for containment. PLoS Negl Trop Dis. 2015;9(4):e0003567. doi:10.1371/journal.pntd.0003567. pmid:25886400
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Tiffany A,
    2. Dalziel BD,
    3. Kagume Njenge H,
    4. et al
    . Estimating the number of secondary Ebola cases resulting from an unsafe burial and risk factors for transmission during the West Africa Ebola epidemic. PLoS Negl Trop Dis. 2017;11(6):e0005491. doi:10.1371/journal.pntd.0005491. pmid:28640823
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Kpanake L,
    2. Gossou K,
    3. Sorum PC,
    4. Mullet E
    . Misconceptions about Ebola virus disease among lay people in Guinea: lessons for community education. J Public Health Policy. 2016;37(2):160–172. doi:10.1057/jphp.2016.1. pmid:26865320
    OpenUrlCrossRefPubMed
  18. 18.↵
    2018- Congo. Ebola virus disease–Democratic Republic of the Congo. World Health Organization. August 17, 2018. Accessed May 4, 2022. https://www.who.int/emergencies/disease-outbreak-news/item/17-august-2018-ebola-drc-en?msclkid=ff5c13e8cd6211eca04d5da15b00ae96
  19. 19.↵
    UNICEF Eastern and Southern Regional Office. Risk Communication and Community Engagement for Ebola Virus Disease Preparedness and Response: Lessons Learnt and Recommendations from Burundi, Rwanda, South Sudan, Tanzania and Uganda. UNICEF; 2020. Accessed May 4, 2022. https://www.unicef.org/esa/media/6391/file/UNICEF%20ESA-Risk%20Communication%20and%20Community%20Engagement%20for%20Ebola%20Virus%20Disease%20Preparedness%20and%20Response.pdf
  20. 20.↵
    Ebola virus disease-Uganda. World Health Organization. June 13, 2019. Accessed May 4, 2022. https://www.who.int/emergencies/disease-outbreak-news/item/13-june-2019-ebola-uganda-en?msclkid=a996312bcd6311eca3ed46bedc9e869c
  21. 21.↵
    1. Aceng JR,
    2. Ario AR,
    3. Muruta AN,
    4. et al
    . Uganda's experience in Ebola virus disease outbreak preparedness, 2018–2019. Global Health. 2020;16(1):24. doi:10.1186/s12992-020-00548-5. pmid:32192540
    OpenUrlCrossRefPubMed
  22. 22.↵
    Open Data Kit. Accessed May 4, 2022. https://opendatakit.org/
  23. 23.↵
    1. Yelland LN,
    2. Salter AB,
    3. Ryan P
    . Performance of the modified Poisson regression approach for estimating relative risks from clustered prospective data. Am J Epidemiol. 2011;174(8):984–992. doi:10.1093/aje/kwr183. pmid:21841157
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. White IR,
    2. Royston P,
    3. Wood AM
    . Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377–399. doi:10.1002/sim.4067. pmid:21225900
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Thiam S,
    2. Delamou A,
    3. Camara S,
    4. et al
    . Challenges in controlling the Ebola outbreak in two prefectures in Guinea: why did communities continue to resist? Pan Afr Med J. 2015;22(Suppl 1):22. pmid:4695515
    OpenUrlCrossRefPubMed
  26. 26.
    1. Masumbuko Claude K,
    2. Underschultz J,
    3. Hawkes MT
    . Social resistance drives persistent transmission of Ebola virus disease in Eastern Democratic Republic of Congo: a mixed-methods study. PLoS One. 2019;14(9):e0223104. doi:10.1371/journal.pone.0223104. pmid:31557243
    OpenUrlCrossRefPubMed
  27. 27.↵
    1. Richardson ET,
    2. McGinnis T,
    3. Frankfurter R
    . Ebola and the narrative of mistrust. BMJ Glob Health. 2019;4(6):e001932. doi:10.1136/bmjgh-2019-001932. pmid:31908869
    OpenUrlFREE Full Text
  28. 28.↵
    1. Nuriddin A,
    2. Jalloh MF,
    3. Meyer E,
    4. et al
    . Trust, fear, stigma and disruptions: community perceptions and experiences during periods of low but ongoing transmission of Ebola virus disease in Sierra Leone, 2015. BMJ Glob Health. 2018;3(2):e000410. doi:10.1136/bmjgh-2017-000410. pmid:29629189
    OpenUrlAbstract/FREE Full Text
  29. 29.↵
    1. Schmidt-Hellerau K,
    2. Winters M,
    3. Lyons P,
    4. et al
    . Homecare for sick family members while waiting for medical help during the 2014–2015 Ebola outbreak in Sierra Leone: a mixed methods study. BMJ Glob Health. 2020;5(7):e002732. doi:10.1136/bmjgh-2020-002732. pmid:32694222
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Muzembo BA,
    2. Ntontolo NP,
    3. Ngatu NR,
    4. et al
    . Local perspectives on Ebola during its tenth outbreak in DR Congo: a nationwide qualitative study. PLoS One. 2020;15(10):e0241120. doi:10.1371/journal.pone.0241120. pmid:33091054
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Carey RN,
    2. Connell LE,
    3. Johnston M,
    4. et al
    . Behavior change techniques and their mechanisms of action: a synthesis of links described in published intervention literature. Ann Behav Med. 2019;53(8):693–707. doi:10.1093/abm/kay078. pmid:30304386
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Falade B
    . Religious and traditional belief systems coexist and compete with science for cultural authority in West Africa. Cult Sc. 2019;2(1):9–22. doi:10.1177/209660831900200102
    OpenUrlCrossRef
  33. 33.↵
    1. Osborne J,
    2. Paget J,
    3. Giles-Vernick T,
    4. et al
    . Community engagement and vulnerability in infectious diseases: a systematic review and qualitative analysis of the literature. Soc Sci Med. 2021;284:114246. doi:10.1016/j.socscimed.2021.114246. pmid:34311391
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Evaluation of Community Perceptions and Prevention Practices Related to Ebola Virus as Part of Outbreak Preparedness in Uganda, 2020
Joseph Musaazi, Apophia Namageyo-Funa, Victoria M. Carter, Rosalind J. Carter, Mohammed Lamorde, Rose Apondi, Tabley Bakyaita, Amy L. Boore, Vance R. Brown, Jaco Homsy, Joanita Kigozi, Aybüke Koyuncu, Maria Sarah Nabaggala, Vivian Nakate, Emmanuel Nkurunziza, Daniel F. Stowell, Richard Walwema, Apollo Olowo, Mohamed F. Jalloh
Global Health: Science and Practice Jun 2022, 10 (3) e2100661; DOI: 10.9745/GHSP-D-21-00661
del.icio.us logo Twitter logo Facebook logo Mendeley logo

Statistics from Altmetric.com

Jump to section

Subjects