Rabies by Dog Bites in Benin: Determinants of Good Knowledge of the Rabies Risk among Subjects Exposed in the Plateau Department in 2023 ()
1. Introduction
Rabies, a viral zoonosis, is a neglected tropical, neuro-invasive disease caused by Rhabdovirus, commonly transmitted by the saliva or bites of certain infected warm-blooded animals such as domestic dogs and cats, which mainly affects poor and vulnerable populations [1] [2] [3] . It is present on all continents, except Antarctica, and represents a public health problem in developing countries [4] [5] [6] . The burden of the disease is borne mainly by Africa and Asia where dogs are the main vector of transmission to humans; it is responsible for up to 99% of all transmissions of rabies to humans. On these two continents, more than three (03) billion people are potentially exposed to rabies. Rabies is one of the deadliest infectious diseases, with a case fatality rate approaching 100%, but deaths due to rabies are rarely reported especially in Africa and particularly in the Republic of Benin [7] . Worldwide, nearly 60,000 people still die each year from this disease [4] . Africa is one of the continents hardest hit by rabies, accounting for more than a third of the world’s estimated human deaths each year [8] .
In Benin, rabies is endemic. Although listed among the notifiable diseases (decree n˚. 2005 - 637 of October 13, 2005), it nevertheless remains a neglected disease. There is no national rabies data management system to bring together human and animal health services. Effective pre- and post-exposure vaccines have existed for about a hundred years, but disproportionate distribution is observed in many countries in Africa. They are not readily available or accessible to those who need them [5] . However, a minority of the population is aware of the risks associated with rabies and the measures to take to prevent human infection, such as seeking post-exposure prophylaxis when a bite occurs, for example. Thus, based on the principle that improving knowledge will make it possible to minimize the burden of morbidity linked to rabies [9] , this study aimed to identify the factors associated with knowledge of the rabies risk in subjects exposed to bites of dogs in order to contribute to developing community awareness and education messages focused on these factors.
2. Materials and Methods
2.1. Community, Type and Study Objective
The study took place in five communes and thirty clusters in the Plateau department, located in the south of Benin. It was a cross-sectional description with the analytical aim and the data was collected from June 1, 2023, to August 31, 2023.
2.2. Sampling, Eligibility Criteria, Recruitment and Data Collection
The sampling method was probabilistic using a three-stage cluster random sampling technique. The cluster unit is represented by a district/village of the Plateau department and the study included 30 clusters.
At the first stage, a random choice of the 30 clusters as well as the four reserve clusters was made in proportion to the size of the villages/neighborhoods in the department. The four reserve clusters served as a replacement cluster in case for one reason or another, one of the 30 clusters was inaccessible.
In the second stage, a systematic random choice of 30 households was made in each of the 30 clusters. In fact, a sampling step was calculated by dividing the number of households in each village by 30. The first household corresponded to a number x drawn at random between 1 and k. The other households were selected by adding, each time, k to the last number.
In the third stage, a respondent aged at least 18 years old was chosen by simple randomness from each chosen household.
Moreover,
· apolygamous family was considered as a single household.
· when, for one reason or another, the cluster was inaccessible, it was replaced by one of the reserve clusters.
· when the interviewers miss a household twice (morning and evening), or when a household does not consent to participate in the survey, or when a household the members of at least 18 years of age is absent on both visits, they were considered non-respondent;
· abandoned concessions, administrative, commercial, educational and religious structures (etc.) were not considered in the sampling.
The data was collected using a questionnaire and it was an individual interviewer-respondent interview. This questionnaire included related information:
· socio-demographic (age, sex, ethnic group, marital status, type of household, profession, level of education subject, father, mother) and economic (socio-economic level) data;
· geographic data: household distance to health center, type of nearby health center, typology of nearby health facility;
· dog bites and cohabitation with dogs: Date of bite, date of consultation, first, second and third-line treatment, site of bite, nature of the animal, symptom, etc.
· monitoring and preventing human rabies: Number of dogs in the household, vaccination status of the dog, organization of vaccination campaign in the locality, stray dog or not, compliance with the dog’s vaccination appointment, etc.
The sample size n was determined by the SHWARTZ formula to which we applied a cluster effect k.
p: Estimated prevalence of knowledge of rabies risk = 1 − q = 0.5
q: 1 − p = 0.5
α = the accepted risk of error = 0.05
Zα2: gap reduced to agreed risk = 1.96
i: desired precision for the results = 0.05
k: cluster effect = 2
n: minimum number of people to survey = 768.3
Considering and adding to this size a non-response rate of 10%, the minimum number of people to survey was 845. Given that only one person per household will be surveyed, the number of households to be surveyed survey was also 845. Considering that the study will include 30 clusters, the number of households included per cluster was therefore 29. We therefore included, for practical reasons and to facilitate calculations in the field, 30 households per cluster. Furthermore, in addition to the 30 clusters to be investigated, four reserve clusters were selected to replace, if necessary, clusters that will not be accessible during collection.
Inclusion criteria
People aged at least 18 years old and crossing the following criteria were included in this study:
· residents in the Plateau department for at least six months;
· present in their dealership on the day the collection teams visit;
· having given their free and informed consent to the study;
· chosen from one of the households drawn according to the sampling plan.
Non-inclusion criteria
The respondents with below criteria were not included in the study:
· having visitor status in the sampled households;
· suffering from an apparent cognitive disorder that does not allow them to understand and answer the questionnaire;
· not belonging to the households sampled according to the sampling plan;
· belonging to households absent from their compound when collection agents visited.
2.3. Data Processing and Statistics Analysis
The questionnaire used as part of this study (available in the appendix) was designed by the author, pre-tested, validated and incorporated into the KoboCollecttoolbox server which served as collection support. The collected data was extracted from the kobotoolbox server. They were analyzed with R software version 4.3.1. Descriptive statistics were used and results were presented as mean ± standard deviation for quantitative variables with a normal distribution and as median for those with an asymmetric distribution. The frequencies of qualitative variables were expressed as percentages.
The dependent variable of the study was “good knowledge of rabies.” It is a binary qualitative variable with two modalities: “good” and “bad”.
A subject was said to have good knowledge of the rabies risk when (A) he correctly answered 50% of the knowledge items provided, (B) he recognized at least 3 signs of human rabies, (C) he knew that human rabies is a fatal disease and that he should get vaccinated after exposure to a dog bite.
The explanatory variables were composed of socio-demographic and economic factors, the practice of dog bite prevention, the characteristics of the animal service, individual factors, and as well as factors linked to the rabies surveillance system in the department.
For the analytical aspect, the Pearson chi-square test was used to test the associations in univariate analysis between the dependent variable and each independent variable. At this level, all variables with a p < 0.2 are taken to construct the logistic regression model. For all analyses, the significance threshold was set at 0.05%.
2.4. Ethical Considerations
The study received a favorable opinion from the local ethics committee for biomedical research at the University of Parakou, which judged it to be compliant and respect fundamental ethical principles. It also received authorization from the Ministry of Health of Benin. In addition, the free and informed consent of the respondents was obtained before each interview.
3. Results
3.1. Description of the Study Population
This study included 900 people residing in the five (05) communes of the Plateau department.
3.1.1. Sociodemographic Characteristics
The average age of the subjects was 40.24 ± 14.52 years with extremes ranging from 18 years to 86 years. The most represented age group was: [30; 50 [years. Male subjects were the most represented with a proportion of 52.56%, i.e. a sex ratio (Male/Female) of 1.11. Of the surveys, 58.78% were Christians. In the study population, 80.89% of those surveyed lived in rural areas. Among the respondents, 46.22% were heads of household, the wives of household were surveyed in 34.78% of cases and the child of household in 18.78% of cases. Overall, 77.0% of habitats were collective. The distribution of subjects surveyed according to socio-demographic characteristics is summarized in Table 1.
The average household size was 5.97 ± 2.57 people with the extremes of 1 to 20 people as presented in Table 2.
![]()
Table 1. Distribution of subjects surveyed on rabies in the Plateau department according to socio-demographic characteristics in 2023 (n = 900).
![]()
Table 2. Distribution of subjects surveyed on rabies in the Plateau department according to household characteristics in 2023 (n = 900).
3.1.2. Socioeconomic Characteristics of Households
More than half of the subjects surveyed in the 900 households are not educated (55.33%) while 28.45% would have stopped their studies in primary school, 15.00% in secondary school and 1.22% have reached a level of university. The median monthly household income of the 120 subjects was 50,000 CFA francs with an IIQ: [30,000; 80,000] F CFA as indicated in Table 3.
3.2. Prevalence of Dog Bites and Knowledge of Rabies Risk Among Exposed Subjects
A total of 37 cases of dog bites were recorded in the study population. The prevalence of dog bites was 4.11% (95% CI = [3.00; 5.63]) (Figure 1). The prevalence of good knowledge of the rabies risk after dog bite in the study population was 66.11% with 95% CI = [62.96 - 69.13].
Table 3 presents the prevalence of dog bites in each commune of the Plateau department in 2023. Among subjects who were bitten by a dog, the male/female ratio was 1.46. The mean age in this population was 43.08 ± 13.10 years.
In terms of education, 43.65% were uneducated, 70.27% lived in collective housing and 83.02% were in rural areas.
![]()
Figure 1. Prevalence of dog bites in the Plateau department in 2023 (n = 900).
![]()
Table 3. Prevalence of dog bites by municipality in the Plateau department in 2023.
3.3. Factors Associated with Good Knowledge of the Rabies Risk among Subjects Exposed to Human Rabies
Several variables were statistically associated with good knowledge of human rabies in the Plateau department. These are the municipality of residence, age, area of residence, level of education, existence of a veterinary service, cohabitation with a dog, presence of a dog in the neighborhood and knowledge of a victim of a dog bite. Bivariate and multivariate analyses are presented in Table 4 and Table 5 respectively.
![]()
Table 4. Bivariate analysis between good knowledge of the rabies risk in the population of Plateau and the presumed associated factors in 2023.
![]()
Table 5. Multivariate model between good knowledge of rabies risk in the population of Plateau and the presumed associated factors in 2023.
*RCA: Adjusted coast ratio
Compared to the subjects surveyed in the commune of AdjaOuèrè, the subjects surveyed in the commune of Sakété are 2.15 times more likely to have good knowledge of the rabies risk and this is statistically significant (p = 0.01).
Compared to subjects under 30 years old, subjects aged 30 years and over are more likely to have good knowledge of the rabies risk and this is statistically significant (linear trend p < 0.01).
Compared to subjects from rural areas, subjects living in an urban environment are 2.04 times more likely to have good knowledge of the risk of rabies and this is statistically significant (p < 0.01).
Compared to uneducated subjects, subjects with a primary school education are 2.79 times more likely to have good knowledge of the risk of rabies and this is statistically significant (p < 0.01).
Compared to uneducated subjects, subjects with secondary education are 5.58 times more likely to have good knowledge of the risk of rabies and this is statistically significant (p < 0.01).
Compared to uneducated subjects, subjects with a higher education level are 11.46 times more likely to have good knowledge of the rabies risk and this is statistically significant (p = 0.03).
Compared to subjects living in an environment where a veterinary service is non-existent, subjects living in an environment where a veterinary center exists are 3.66 times more likely to have good knowledge of the rabies risk and this is in a statistically significant manner (p = 0.04).
Compared to subjects who did not live with a dog, subjects who lived with a dog were 2.09 times more likely to have good knowledge of the rabies risk, and this was statistically significant (p < 0.01).
Compared to subjects whose neighbors did not have a dog, subjects whose neighbors had a dog were 1.70 times more likely to have good knowledge of the risk of rabies, in a statistically significant manner (p = 0.02).
Compared to subjects who do not have knowledge of a dog bite victim, subjects who do are 4.75 times more likely to have good knowledge of the risk of rabies and this is statistically significant (p < 0 .01).
4. Discussion
The main limitation of this study is its nature. Being transversal in nature, we were only able to take a snapshot of the situation present on Plateau at the time of the study. Thus, the associations observed between good knowledge of rabies and the explanatory variables could also vary over time.
4.1. Sociodemographic Characteristics of the Subjects Surveyed
The average age of the subjects surveyed was 40.24 ± 14.52 years with extremes ranging from 18 years to 86 years. It is higher than the average age found by Carrara et al. in 2012 (37.7 years) reviewed from 1990 - 2012 [10] . The difference observed could be due to the fact that Carrara et al. in their study also included subjects aged under 18 years while the present one had exclusively included subjects aged at least 18 years. It remains lower than that found by Barbosa Costa et al. in Cameroon in 2018 (43.6 years) [11] . The average age of the subjects found by Niang et al.in 2020 in the community of Sokone in Senegal was 29.1 years [12] .
The age group most represented in the present study was: [30; 50[ years corresponding to the youth of the active population in our cities and countryside.
Male subjects were the most represented with a proportion of 52.56%, or sex ratio (Male/Female) of 1.11. In the review by Carrara et al. [10] , the sex ratio (Male/Female) was 3.5. Niang et al.in their study in 2020 reported a sex ratio (Male/Female) of 1.5 close to that found in the present study [12] .
Christianity is the dominant religion in the present study (58.78%) unlike the study reported by Niang et al.in 2020 in Senegal where the Muslim religion was predominant with a proportion of 91.10% [12] . Although secular, the populations of southern Benin are predominantly Christian in contrast.
4.2. Socioeconomic Characteristics of Survey Subjects
The median monthly household income was 50,000 CFA francs, IIQ: [30,000; 80,000] and more than half of the households had a monthly income close to the Beninese Guaranteed Interprofessional Minimum Salary.
More than half of the subjects surveyed are not educated (55.33%), those educated had 28.44% a level of study corresponding to primary school, 15.00% to secondary school and 1.22% have reached a level of university. These observations would be due to the predominantly rural study area.
4.3. Prevalence of Dog Bites in the Community
In the Plateau department, the prevalence of dog bites was estimated at 4.11% in 2023. This proportion is similar to that found by Niang et al. in 2020 in Senegal (4.55%; 37/813) [12] and lower than those of Hergert and Nel Louis in 2013 in South Africa on the one hand and Ahmed et al.in Ethiopia in2021 on the other hand (28.2%; 94/326) [13] [14] . The low prevalence of dog bites in the Plateau department could be explained by the low reporting of bites cases due to endogenous beliefs and the lack of perception by communities of the risk of developing human rabies. Sociological gravity assimilates dog bites to spiritual and mystical facts requiring a discreet solution. However, the difference observed between the proportion of bites in the present study and that found by Ahmed et al. in Ethiopia, could be explained by the size of our different samples. The sample of this study carried out in the Plateau is approximately 2.5 times larger than the one carried out in Ethiopia.
4.4. Factors Associated with Good Knowledge of Human Rabies
Factors that predict good knowledge of human rabies in the community of the Plateau Recovered department are the municipality of residence, the area of residence, the type of habitat, the age group, the level of education, the existence of a veterinary center, cohabitation with dogs, availability of dogs among neighbors and knowledge of a dog bite victim.
The subjects’ municipality of residence influenced their knowledge of human rabies in the Plateau department. This observation is accepted since in the literature, certain diseases occur disproportionately in countries and within a country in a different way from one region to another. This observation was also made in the department of Plateau with regard to the knowledge of human rabies. The multivariate logistic regression analysis performed in the study by Rinchenet al. showed that residency was significantly associated with adequate knowledge of rabies [15] . The same observation was made by Madjadinan et al. in their study in Chad [16] . These conclusions are also supported by the results found by Awuniet al. in Ghana in 2017 who found that district of residence was statistically associated with knowledge of rabies using the Pearson chi-square test [3] .
Age was corroborated with knowledge of certain illnesses. Through the present study, it appears that younger subjects are likely to misunderstand rabies. In a study carried out in India [16] , Madjadinan et al. also came to the conclusion that the age of the subjects influences their knowledge related to rabies. This observation is consolidated with regard to the results found by Leblanc et al.with his study in Madagasca in 2018 which are unanimous. Indeed, according to their conclusions, lower knowledge scores were associated with younger respondents [17] .
The type of habitat is a determinant of health [18] and can therefore influence health knowledge, as is the case in this study.
The role of education in the knowledge of certain diseases such as human rabies is admitted. Our study confirms that uneducated subjects are the most susceptible to ignorance of human rabies. This remark is also shared by Herbert et al., Sambo et al., Bouaddi et al. who reached similar conclusionsrespectively in Tanzania (2009 and 2010), in India (2010) and in Morocco (2018) [9] [19] [20] . Furthermore, other authors who have investigated the same theme as us have not reached the same conclusion.
Cohabitation with dogs can influence knowledge of rabies, as is the case in the Plateau department. Knowledge of a person who was the victim of a dog bite that resulted in death from human rabies could also influence knowledge of human rabies. The multivariate logistic regression analysis performed in the study by Rinchenet al. in Bhutan in 2017 showed that having seen a case of human rabies was significantly associated with adequate knowledge of rabies [15] and the results from the analysis by Al-Mustapha et al. with the same regression method in Nigeria in 2020 [21] corroborates the susceptibility of dog owners to have good knowledge of canine rabies.
Other studies found that non-dog owners were 1.6 times more likely to have heard of rabies than dog owners [13] .
5. Conclusions
The results of this study prove that the population in rural areas is very exposed to human rabies due to cohabitation with dogs with unknown vaccination status, and this constitutes a real public health problem.
In view of the results obtained, despite the good overall knowledge of human rabies by this study population, it is urgent to strengthen awareness-raising actions aimed at rural and illiterate populations on the risk of developing human rabies following the bites of dogs.
The strengthening of existing surveillance strategies for the control and eradication of human rabies will involve the application of the One Health approach but it should be recognized that the eradication of the phenomenon is not limited to the intensification of awareness and monitoring efforts. A more holistic and integrated approach is needed, involving collaboration between health authorities, local communities, veterinarians as well as technical and financial partners.
Appendix
Community Questionnaire