Knowledge, attitudes and practices (KAPs) and risk factors of brucellosis at the human-animal interface in the Nile Delta, Egypt

Background Brucellosis is a highly contagious zoonosis affecting human and almost all domestic species. It is a multi-burdens disease leading to severe economic losses due to disability in humans in addition to abortion, infertility and reduced milk production in animals. An Important element for effective prevention and control of brucellosis is to improve knowledge, attitude and practices of the community. Objective(s) This study aimed to assess the knowledge, attitudes and practices (KAP) related to brucellosis at human-animal interface and to determine the risk factors for human infection in the Nile Delta, Egypt. Methods A matched case-control study was conducted at the main fever hospitals located in 6 governorates in the Nile Delta, Egypt between June 2014 and June 2016. Face-to-face interviews with cases and controls was done using a structured questionnaire. Differences in proportions of KAP variables among the cases and controls were evaluated by Pearson’s Chi square test and a p value <0.05 was set as a level of significance. A multivariate conditional logistic regression analysis was built to determine the risk factors for Brucella spp. infection among study participants. Results A total of 217 cases and 434 controls matched for age, gender and sociodemographic characteristics were enrolled and interviewed. In total, 40.7% of the participants owned animals in their households and lived in shared accommodation with animals [48.8% of cases vs 36.9% of controls; (p= 0.003)]. The majority (78.1%) used to accommodate cows and buffaloes with sheep and goats. Human brucellosis cases experienced more animal abortions comparing to the controls [(23.5% vs 9.7%, respectively), (p= 0.0003)]. The majority of the participants (82.4%) did not notify authorities in case that abortion occurs in their owned animals. Apparently, 67.4% of the participants [(70.0% of the cases vs 66.1% of the controls) (p = 0.315)] had not ever heard about brucellosis. The overall mean practice score regarding animal husbandry, processing and consumption of milk and dairy products was significantly lower among cases comparing to controls (−12.7±18.1 vs 0.68±14.2 respectively; p<0.0001). Perceived barrier for notification of animal infection and/or abortion was significantly higher among cases (p= 0.034) and positively correlated with participants’ education. Results of univariate analysis showed that participants who have animals’ especially small ruminates were at a higher risk of getting Brucella spp. infection than others. In the proposed multivariate conditional logistic regression model, the predictors of having brucellosis infection were consumption of unpasteurized milk, having consumed dairy products in the last 3 months before the study, consumption of yoghurt or home-made cheeses and involvement in contact with animals [OR (95% CI) = 4.12 (1.62 - 10.75); 2.71 (1.06 – 6.93); 2.51 (1.21 – 5.24); 1.96 (1.17- 3.30), p<0.05; and 4.97 (2.84 - 8.72)], respectively. Participants who take more protective measures against infection were at a significant lower risk of being diseased with brucellosis; [OR (95% CI) = 0.23 (0.10 - 0.58); p<0.001], respectively. A model predicting risk factors for brucellosis among those who own animal showed that frequent abortions per animal increased the chance for brucellosis infection among human cases by 49.33 fold [(95% CI)= (8.79 – 276.91); p= 0.001] whereas the practice protective measures with animals was protective for humans as well [OR (95% CI)= 0.11 (0.03-0.45); p= 0.002]. Conclusion Consumption of dairy products stands side by side with the contact with infected animals particularly aborted ones as the major risk factors for Brucella spp. infection among humans in Egypt. On the other hand, there is a poor knowledge, negative attitudes and risky behaviors among villagers which increase the magnitude of the risk of brucellosis transmission at the human-animal interface. This supports the need for integrating health education in the national brucellosis control programs in Egypt with a special emphasis on hygienic animal husbandry, disease notification and benefits of animal vaccination. Author summary Zoonotic brucellosis has a vast global burden and remains neglected in many areas of the world despite notable advances in disease containment strategies. Despite the implementation of a national brucellosis control program in Egypt, the challenges for the disease eradication are intractable and multifaceted. We modeled in the present study the multivariate factors for brucellosis persistence in Egypt which apparently pointed to lack of basic understanding of the nature of brucellosis, traditional practices, beliefs and risky behaviors being undertaken on farms and at households across a wide region of the country. Predominantly, consumption of dairy products from unregulated sources; underreporting animal infection and abortion; underutilization of animal vaccination service; unsanitary disposal of abortus; use of milk of infected/aborted ruminants and lack of protective measure when practicing animal husbandry. Together, these conflict with disease intervention strategies and contribute to disease spread and re-emergence. The proposed model can provide a framework for future containment strategies that should be adopted to support and enhance the adherence to the current national brucellosis control program.

322 (p=0.034) and positively correlated with participants' education (Table 5). Knowledge was strongly 323 associated with participants' practice, perceived barriers, as well as perceived susceptibility and risk 324 ( Figure 3).

378
Results of multivariate logistic regression analysis are shown in (

415
Knowledge about the disease and preventive herd management practices have previously 416 been identified as the most important factors required for minimizing the disease risk in animals [24] 417 Apparently, brucellosis as a disease of animals is not known by our study population, since more 418 than two thirds of the participants had not heard of the disease or its possible transmission. This is 419 consistent with previous reports in Egypt [

503
Only a small proportion of the study respondents perceived that brucellosis was a serious 504 disease in both animals and humans and that animal husbandry is a risky practice. Accordingly, they 505 had unfavorable attitude towards good practices in prevention of brucellosis. Several known high-506 risk behaviors were common self-reported practices among the study participants, particularly 507 cases. They were more likely to engage in risky practices that could expose them to infection. This 508 was evident from actions most of them would take when confronted with an aborting animal in their 509 herd. The majority would not seek governmental veterinary services and thus would not notify the 510 disease. Some manage animal infection or abortion on their own and some call private veterinarians.

551
Although having infected animal in the households, and notification of authorities for of a 552 case of brucella were removed from the logistic regression model for predicting brucellosis infection, 553 we do not think that these are not considered as risk factors for brucellosis infection but the results 554 we had may be due to the fewer number of respondents to these questions.

555
The number of abortions per animal increased the chance for infection among human cases.
556 This is an important finding in the present work especially for people who insist to keep aborted 557 animals [for fattening or reproduction] and do not like to slaughter them. It is worth noting that 558 livestock owners deny having animal abortion to be able to sell them or their milk [12].

559
Animal vaccination appeared as a protective factor in our proposed logistic regression model 560 for risk factors of human infection. This finding supports the economic and health benefits of animal 561 vaccination in reducing the occurrence of human brucellosis [52]. Importantly, a significant number 562 among those actually practicing animal vaccination did not list this practice among the measures 563 that they adopt for protecting themselves, their household member or their animals against 564 brucellosis. This means that villagers can follow a preventive measure without being aware of its 565 benefits probably because this was not properly explained to them.

566
The key limitation of the present work is the self-reporting on practices by the respondents 567 that was subject to recall bias, Hawthorne effect and the face-to-face interview situation.
568 Observational checklist could have enhanced this type of bias in assessing attitudes and behaviors.

584
The prospective of this work is to collect information from different community sections to 585 assess the economic impacts of brucellosis on small livestock holders and on the national level. Data 586 of the current study will be the corner stone for the following section of the project which aims at 587 building a more realistic to apply and economically efficient socioeconomic model for brucellosis 588 control in the Nile Delta region and consequently in Egypt to substitute the current national 589 brucellosis control strategy. 599 Informed written consent was obtained from the individuals participating voluntarily. Data sheets 600 were coded to ensure anonymity and confidentiality of patient's data.

601
This article does not contain any studies with animals performed by any of the authors.