The Langue de Barbarie is in the Saint-Louis region on Senegal’s north coast, approximately 250 km from Dakar. The city of Saint-Louis is located in the estuary of the Senegal River. It extends over an area comprising three entities: the Langue de Barbarie, the island of Saint-Louis and the large suburb of Sor. The Langue de Barbarie is where the Senegal River meets the Atlantic Ocean. It is a very narrow strip of land bounded by the Atlantic Ocean on the west coast and by the Senegal River on the east coast. The northern limit of the Langue de Barbarie is the border with Mauritania, less than 3 km from Goxu-Mbacc, while in the south, the strip of land ends at the mouth of the Senegal River.

The population of the Langue de Barbarie is divided between the 3 main neighbourhoods: Guet-Ndar/Hydrobase, Santhiaba and Goxu-Mbacc. It is subdivided into 12 sub-neighbourhoods ( Table 1 ).

The population of the Langue de Barbarie was estimated at more than 60,549 in 2013, according to figures from the general census of population and housing, agriculture, and livestock (RGPHAE 2013). It represented almost a quarter of the population of the commune of Saint-Louis, which stood at 253,392.

The Langue de Barbarie is part of the Saint-Louis health district. In terms of health infrastructure, the Langue de Barbarie currently has:

On the other side of the river, on the island, the Guet-Ndar neighborhood faces the Saint-Louis regional hospital.

In terms of education infrastructures, according to data for 2022 from the education and training inspectorate of Saint-Louis (IEF), the Langue de Barbarie has 11 elementary schools, 2 secondary schools, 3 pre-schools and several daaras (koranic school).

Regarding socio economic aspects, the three neighborhoods of the Langue de Barbarie have the highest concentration of fishermen in Senegal. Most of their inhabitants depend on sea fishing for their livelihood [9] .

Indeed, the male population is almost exclusively made up of fishermen, with more than 86% of male heads of household engaged in fishing [10] .

Artisanal processing of fish products is mainly carried out by women and is a source of income. It is an old tradition among the women of the Langue de barbarie, particularly in Guet-Ndar, who are known for their long experience and know-how in this profession [11] .

This is a descriptive and analytical cross-sectional study in the Langue de Barbarie, Saint-Louis.

The study population consisted of Women in Reproductive Age (WRA), mothers or guardians of daughters aged between 9 and 14.

Inclusion criteria: All mothers or guardians of girls aged between 9 and 14 who have lived in the Langue de Barbarie for more than 6 months at the time of the study and who have given their consent to take part in the research.

Non-inclusion criteria: Any mother or guardian absent or unavailable at the time of the survey.

The minimum sample size was calculated using the Schwartz formula:

$n=t^2\times p\times \left(1-p\right)/m^2$

The minimum sample size calculated was 768, rounded to 800 to take into account of possible non-responses. The sample was distributed proportionally to the population size by neighborhood in the Langue de Barbarie.

A cluster survey was used with:

To choose the starting point, the sub-neighborhoods were divided into sectors with an average of thirty concessions in each sector. The first concession was chosen at random by elementary sampling in the sector, which was also identified at random by drawing lots.

At the concession level, households are numbered, and one household drawn at random. All the women that met the selection criteria at the level of the household selected at random were surveyed. A mother or guardian who met the inclusion criteria would only be surveyed once. If she had several daughters/wards in the 9 to 14 age group, a draw was made to choose the girl around whom the interview would revolve.

Data collection was carried out from 2 to 19 January 2024 based on anonymous questionnaires administered by interviewers to women in the targeted households. This questionnaire was structured in 4 parts: identification of the questionnaire, the respondent’s profile including her socio-demographic characteristics and those of the girl about whom the interview focused, knowledge of the HPV vaccine and Senegal’s vaccination strategy, the respondent’s relationship to cervical cancer and HPV vaccination including the girl’s vaccination status reported and the number of doses received.

The questionnaire was configured on tablets using Survey 123 software to allow direct entry of information into the database.

The team of interviewers was made up of 3 men and 7 women who were trained in the use of the tablets for data collection and in the questionnaire so that the questions could be better formulated in the local language. The questionnaire was pre-tested after the training in the Khor area, another locality in Saint-Louis, with around twenty women, which enabled adjustments to be made to the wording of some of the questions and the answers given by the women during the pre-test. The interviewers were distributed throughout the different zones and local resource persons were identified to facilitate their introduction to the local population. The data collected was protected and stored on the password-protected platform, with the possibility of monitoring the progress of the survey in real time.

The data collected were exported in Excel format and then in csv and the analysis were carried out with the R software version 4, which is free and open source [12] .

A descriptive analysis was carried out using frequencies with their standard deviation for the qualitative variables. For quantitative variables, basic parameters such as mean, minima, maxima, median and standard deviation were presented [13] .

A bivariate analysis between the use of the HPV vaccine in girls aged 9 to 14 as the dependent variable and the other factors studied was performed in R. The usual chi-square, t-test or ANOVA statistics were used, depending on the situation, to determine whether the variables studied influenced HPV vaccine use among 9- to 14-year-old girls in Langue de Barbarie. These tests were used to prepare the multivariate analysis by specifying the p-values of the different variables involved.

A top-down logistic regression strategy was used for the multivariate analysis. All variables with p-values less than 0.25 [14] [15] were retained in the initial model developed. A likelihood ratio test was performed manually to study the model parameters before selecting a final model. This process highlighted the factors associated with vaccination and the different degrees of association for each factor through the adjusted ORs.

The performance of the model was evaluated by calculating the Area under Curve (AUC) which is a good indicator of the quality of the final model.

Participation in the survey was free and voluntary for all respondents. Comprehensive information on the study was shared in advance of the questionnaire being administered and the interviews proceeded only once informed consent was obtained.

The research protocol was submitted to and approved by Senegal’s National Health Research Ethics Committee (CNERS) under number SEN23/84 on 25 October 2023.

A total of 799 people were interviewed in the various districts of the Langue de Barbarie. No cases of refusal were documented. The average age of the respondents was 35.67, with a standard deviation of 7.08 and range of 17 and 49. The median age was 36. Respondents were most often aged 30 - 40 (47.6%). Biological mothers were the main interviewees (82.7%). Educated participants accounted for just over half of the respondents (56.7%) and were mainly at primary school level (40.9%) ( Table 2 ).

More than half of the respondents (63.7%) had their daughters educated in French and 23.3% in Arabic; 6.6% had not attended school and 6.4% had dropped out.

The mean age of girls was 11.4 ± 1.7 years, with and a range of 9 to 14. The median age was 11 ( Table 3 ).

HPV vaccination coverage of the girls was 41.9%, of which 24.3% had received one dose (HPV1), 17.3% two doses (HPV2) and 0.3% three doses (HPV3).

A bivariate analysis between HPV vaccination status and the other factors studied was performed in R.

The analysis of socio-economic factors in relation to vaccination revealed a link between vaccination and factors such as place of residence, girls schooling, and the occupation of mothers or guardians ( Table 4 ).

Bivariate analysis performed with the variable related to knowledge of mothers and guardian revealed a link between vaccination and knowledge on cervical cancer, HPV and the strategy of vaccination ( Table 5 ).

*Link with HPV vaccination.

The analysis with factors related to women’s relationship to HPV and vaccine showed link between HPV vaccination and the mother or guardian knowing someone with gynaecological cancer, mother/guardian screened for cervical cancer, opinion on vaccination, acceptance of vaccination and willingness to recommend the vaccination to a third party ( Table 6 ).

Overall, bivariate analysis showed that vaccine uptake was related to the following variables: area of residence, education of girls, occupation of women, knowledge of cervical cancer, knowledge of HPV, knowing someone with gynecolical cancer, acceptance of the vaccine and recommendation. All these variables were included in the multivariate analysis to model women’s use of the HPV vaccine as a function of the variables studied.

The logistic regression used for the multivariate analysis highlighted the factors associated with vaccination and the different degrees of association for each factor through the adjusted Odds Ratio (ORs). Table 7 lists the factors associated with the use of the HPV vaccine in girls aged 9 to 14 in the LDB.

AUC = 81.4%.