The study took place in the Atlantique department, specifically in the Allada-Toffo-Zè (ATZ) health zone, located in southern Benin. The health zone has an estimated population of about 740,000 inhabitants and comprises 33 districts (arrondissements) and 76 health facilities.

This was a descriptive, analytically oriented cross-sectional study, conducted from December 1, 2024, to June 28, 2025.

The study population consisted of mothers of children aged 6 - 18 months residing in the ATZ health zone, representing the study’s primary target. We excluded from the study individuals unable to answer questions clearly and precisely, those who did not give consent to participate, and incomplete forms (26).

The sampling method was a three-stage probability cluster design for the selection of mothers of children (respectively at the district [arrondissement], city ward [quartier], and household levels). As the health zone is subdivided into three (3) communes, a simple random sample of four (4) districts per commune was drawn. Within each district, a further simple random sample of three (3) city wards per district was drawn, yielding a total of 36 city wards. Within each ward, households were selected at random by bottle-spinning, with subsequent households then identified. The starting point in each ward was the ward center, identified by the local official who administratively heads the ward (the authority in charge of the ward, commonly called the “ward chief”), which corresponded to the central crossroads. From this crossroads, the direction of progression was determined by bottle-spinning, allowing the direction to be selected. Once the street was selected, the first household was chosen (a household being defined here as a group of persons, or a single person, sharing the same dwelling, generally eating together, and constituting the basic unit for assessing health risks, hygiene conditions, and access to care) by random drawing among the households of the compound. If a household was absent when the interviewer visited, the next household was chosen, with no subsequent return visit. In the absence of an eligible mother, the household was declared ineligible and replaced according to the same rule. Once all households along the selected direction had been covered, the next direction connected to the crossroads was selected, and so on until the desired cluster size was reached. In total, 630 mothers were approached; 604 questionnaires were retained after excluding 26 incomplete forms, corresponding to a response rate of 95.9% (604/630). Within each household, one mother of a child was randomly selected among those present in the household. The minimum sample size was determined using Schwartz’s formula: n = Zα² × (p·q/i²) × d, with Zα = 1.96 for α = 0.05, p the probability of occurrence of the event (p = 0.60) [9][10], q the probability of non-occurrence of the event (q = 1 − p = 0.40), i the precision, and d the design effect correction factor (i = 0.05 and d = 1.5). The minimum sample size was thus estimated at 554 mothers. In total, 604 mothers of children aged 6 to 18 months were selected.

Data were collected from mothers of children through interviewer-administered interviews. The interviewers comprised six (6) men and women, holders of degrees in health sciences, with at least two prior experiences in electronic data collection, good command of local dialects, and formal training. Quantitative data were collected using a structured, digitized questionnaire administered face-to-face. The questionnaire, based on the WHO’s BeSD framework [11], covered sociodemographic characteristics, views and feelings toward malaria and the MV, social norms and connections, motivation for vaccine uptake, practical issues related to the MV and the organization of vaccination services, and the child’s vaccination status verified against the vaccination record (carnet).

The dependent variable of the study was the child’s adoption of the malaria vaccine (MV), defined as actual uptake or non-uptake of the malaria vaccine in accordance with the current vaccination schedule, verified from the vaccination record. Benin’s MV dosing schedule comprises four doses, administered at 6, 7, 9, and 18 months respectively. A child was considered up to date for their age if, at the time of the survey, they had received all doses for which the eligibility age had already been reached (for example, an 8-month-old child was expected to have received doses 1 and 2, without the absence of dose 3, due at 9 months of age and therefore not yet due, being counted as a vaccination gap). The independent variables were grouped according to the dimensions of the WHO’s Behavioral and Social Drivers of Vaccination (BeSD) framework. Sociodemographic characteristics included the child’s age, the child’s sex, the mother’s age, education level, marital status, religious affiliation, occupation, monthly income, and the distance between the home and the health center. Variables related to knowledge of malaria covered knowledge of the disease, its causes, its clinical signs, its progression to severe forms or death, as well as knowledge of seasonal malaria chemoprevention in children. Variables related to views and feelings toward malaria and the malaria vaccine included the perceived risk of malaria occurrence in the child, the perceived risk of progression to a severe form or death, the perceived financial and social impact of malaria, trust in childhood vaccines, awareness of the existence of the malaria vaccine, knowledge of its dosing schedule, trust in the MV, and perceptions of its effectiveness, safety, and importance for the child’s health. Variables related to social norms and connections included perceived community endorsement of the MV, exposure to community discussions about the vaccine, the opinion of close parents and friends regarding the MV, the mother’s decision-making autonomy, the person authorizing vaccination when decision-making depends on someone else, and perceived equality of access to MV information between men and women. Motivation variables mainly comprised the intention to have the child vaccinated against malaria. Variables related to practical issues and the organization of vaccination services included the perceived financial ease of obtaining the vaccine (overall direct and indirect costs of vaccination for the child, including transportation, meals, any incidental costs, and the opportunity cost linked to interrupting an income-generating activity on the day of vaccination), ease of access to vaccination services, availability of transportation, evaluation of the quality of vaccination services, evaluation of the quality of reception at health centers, and whether or not the mother had received information or advice about the MV from a health worker.

Data were analyzed using R software, version 4.3.3. Quantitative variables were described using their median and interquartile range. Qualitative variables were summarized using frequencies and proportions. Associations between qualitative variables and malaria vaccine uptake were assessed using the Rao-Scott adjusted χ² test, suited to complex sampling designs, via the svychisq function of the R survey package. When cell counts were insufficient for certain categories, these were merged to satisfy analysis conditions. Variables showing an association with a p-value < 0.20 by the Rao-Scott adjusted χ² test were retained as candidates for the multivariable model. Factors associated with malaria vaccine uptake were identified using binary logistic regression accounting for the sampling design. The survey design was specified using the R survey package, in accordance with the three-stage probability cluster sampling design. Results are presented as adjusted odds ratios (aOR) with their 95% confidence intervals (95% CI). The statistical significance threshold was set at p < 0.05. Missing data were handled by excluding incomplete observations, with no imputation method applied.

Data collection was preceded by the required administrative authorizations: a research authorization from the Department of Public Health of the Faculty of Health Sciences at the University of Abomey-Calavi (No. 165/FSS/DSP/SCOM of December 12, 2024), followed by a survey authorization from the Departmental Health Directorate of the Atlantique department (No. 097/25/MS/SGM/DDS ATL/SPAF/SA of February 10, 2025). In addition, written informed consent was obtained from each participant. Data confidentiality and participant anonymity were guaranteed throughout the study.

The median age of selected children was 12 months (IQR: 9 - 18 months), ranging from 6 to 18 months, with 57.1% aged 12 to 18 months. The majority of children (50.8%) were female. The median age of surveyed mothers was 26 years (IQR: 22 - 30 years), ranging from 16 to 45 years, calculated on 572 valid observations (age could not be recorded for 32 of the 604 mothers: 5.3%). The 26 - 35 age bracket was the most represented (47.4%). Most mothers were in a cohabiting or common-law union (69.3%) and of Christian faith (86.9%). Overall education level was low: 57.4% had no formal schooling. Regarding income, 96.9% of mothers reported very low income (0 - 50,000 FCFA/month), working mainly in the informal sector (petty traders/vendors 36.8%, homemakers 25.0%). Nearly all (77.5%) lived less than 5 km from a health center. Regarding association with vaccine uptake, these variables generally showed no statistically significant relationship with vaccine uptake: neither the child’s age (F = 3.70; p = 0.063), nor sex (F = 0.34; p = 0.707), nor the mother’s age (F = 0.04; p = 0.963), marital status (F = 0.61; p = 0.518), religion (F = 0.47; p = 0.683), education level (F = 0.18; p = 0.940), income level (F = 2.58; p = 0.117), or occupation (F = 1.45; p = 0.204) (Table 1).

Knowledge of malaria among mothers was nearly universal (99.3%). The majority of mothers (94.4%) correctly attributed the cause to mosquito bites, although a minority cited alternative causes such as sun exposure (12.7%) or spiritual causes (0.2%). Nearly all mothers (96.9%) knew the signs (at least two (2) clinical signs) of the disease, and 68.9% were aware that the disease could progress to severity.

Table 1. Sociodemographic characteristics of surveyed mothers and children in the Allada-Toffo-Zè health zone, 2025.

*32 missing values for mother’s age, corresponding to mothers whose age could not be determined.

Mothers’ knowledge of malaria was not significantly associated with MV uptake (Rao-Scott F = 1.47; p = 0.234). Knowledge of the disease’s signs was likewise not associated with uptake (F = 2.49; p = 0.124). By contrast, knowledge of malaria chemoprevention as a preventive measure was significantly associated with uptake, but in a counter-intuitive direction: mothers who did not spontaneously mention it had a higher uptake rate (78.2% versus 50.0%; F = 21.17; p < 0.001), likely reflecting confusion between chemoprevention and vaccination in the responses. Awareness of the MV’s existence remained the knowledge factor most strongly associated with uptake (87.6% versus 31.1%; F = 143.42; p < 0.001), as was knowledge of the vaccine’s dosing frequency (F = 11.14; p < 0.001) (Table 2).

Table 2. Association between knowledge of malaria/the malaria vaccine and MV uptake among children aged 6 to 18 months.

The proportion of mothers who had their child vaccinated with the MV was higher among those who perceived a substantial risk of malaria occurrence in their child; 80.5% of mothers rating this risk as very high had vaccinated their child, versus 72.7% among those perceiving a high risk and 71.9% among those citing a moderate risk (F = 4.43; p = 0.006). A similar trend was observed for perceived severity of malaria. Uptake concerned 75.4% of mothers perceiving a high risk of progression to a severe form or death (F = 4.47; p = 0.003). In financial terms, 79.3% of mothers (among the 597 who answered this question) perceived a substantial financial and social impact of malaria; however, this did not reach the significance threshold (F = 1.60; p = 0.182) (Table 3).

Table 3. Association between perceived malaria risk in the child and malaria vaccine uptake.

*n = 597 due to 7 missing responses.

Trust in childhood vaccines in general was significantly associated with malaria vaccine uptake (F = 7.46; p < 0.001); 80.1% of mothers with very high trust in childhood vaccines had their child vaccinated against malaria. Trust specifically in the malaria vaccine was likewise strongly associated with uptake (F = 30.56; p < 0.001). The uptake rate reached 84.6% among mothers with very high trust in the MV and 82.1% among those with high trust. Perceived effectiveness of the malaria vaccine emerged as one of the variables most strongly associated with MV uptake (F = 72.86; p < 0.001). Among mothers who considered the vaccine very effective, 94.5% had their child vaccinated. A similar association was observed for perceived safety of the vaccine (F = 56.35; p < 0.001). Mothers who rated the vaccine as “safe” or “very safe” had uptake rates of 92.6% and 91.4% respectively.

Perceived importance and benefits of the malaria vaccine were also strongly associated with uptake (F = 26.58; p < 0.001). Vaccination concerned 89.9% of mothers who rated the vaccine as very important and 81.3% of those who rated it as important (Table 4).

Table 4. Association between mothers’ trust, perceptions, and attitudes toward vaccines and the malaria vaccine, and malaria vaccine uptake in the Allada-Toffo-Zè health zone, 2025.

Perceived community endorsement of the malaria vaccine was strongly associated with MV uptake (F = 29.05; p < 0.001); 94.7% of mothers perceiving very strong community endorsement of the vaccine had their child vaccinated. Exposure to community discussions about the malaria vaccine was likewise associated with uptake (F = 21.71; p < 0.001). Among mothers exposed to discussions about the MV within their community, 89.6% had their child vaccinated, versus 67.7% among those who had not participated in or been exposed to such discussions. Similarly, a favorable opinion of close family and friends toward the MV was significantly associated with uptake (F = 36.20; p < 0.001); 74.8% of mothers benefiting from a favorable opinion from those around them had their child vaccinated, versus only 27.6% among those reporting an unfavorable opinion. Regarding the mother’s decision-making autonomy, it was not significantly associated with malaria vaccine uptake (F = 0.41; p = 0.524). The uptake rate was 73.6% among mothers reporting that they made vaccination decisions for their child themselves, versus 69.8% among those depending on another person for this decision. However, among mothers lacking decision-making autonomy, the person granting vaccination permission was significantly associated with uptake (F = 6.68; p = 0.015). When permission came from the respondent’s own mother, 83.8% of children were vaccinated, versus only 60.4% when this decision depended on the head of household. Perceived equal access to vaccine information between men and women was significantly associated with MV uptake (F = 5.20; p = 0.029). Mothers perceiving equal access to information had an uptake rate of 77.8%, versus 63.7% among those perceiving inequality. The intention to vaccinate the child remained, by far, the determinant most strongly associated with actual uptake (F = 122.42; p < 0.001). 80.0% of mothers expressing an intention to vaccinate their child actually did so, versus only 9.4% among those with no such intention, showing a direct influence. Regarding the determinants favorable to vaccination intention, all modalities presented as multiple-choice options and tested were significantly associated with vaccine uptake, with p-values of 0.000, indicating very strong statistical significance. Leading these, health professionals’ recommendations (88.1%) and family/friends’ opinions (87.9%) emerged as the strongest drivers. These were followed by quality of reception (87.5%), clear information (72.7%), vaccine accessibility (72.7%), and financial support or free vaccine provision (57.1%) (Table 5).

Table 5. Association between social norms, community influences, and malaria vaccine uptake among children aged 6 to 18 months in the Allada-Toffo-Zè health zone, 2025.

The uptake rate reached 92.6% among mothers rating financial access to the vaccine as very easy, versus 78.0% among those rating it as easy (F = 2.60; p = 0.043). Ease of access to vaccination services was likewise associated with uptake (F = 2.72; p = 0.043). Mothers rating access as very easy had an uptake rate of 89.3%, versus 59.1% among those rating access as not easy at all. In addition, availability of transportation to the vaccination center was not significantly associated with uptake (F = 2.36; p = 0.134), although the uptake rate was slightly higher among mothers with access to transportation (81.7%) than among those without (70.5%). Evaluation of the quality of services provided at vaccination centers was likewise associated with uptake (F = 4.45; p = 0.005). The uptake rate was 85.1% among mothers rating services as very good and 74.3% among those rating them as good. Evaluation of the quality of reception was also significantly associated with MV uptake (F = 5.73; p < 0.001). Mothers who rated reception as very good had an uptake rate of 87.8%, versus 72.4% among those rating it as fairly good and only 50.0% among those rating it as poor. Having received advice or information from a health worker regarding the malaria vaccine was strongly associated with uptake (F = 38.70; p < 0.001). Among mothers who had received information or advice about the MV, 89.0% had their child vaccinated, versus only 57.5% among those who received no information (Table 6).

Table 6. Association between practical issues related to access to vaccination services and malaria vaccine uptake among children aged 6 to 18 months in the Allada-Toffo-Zè health zone, 2025.

Mothers who reported an intention to have their child vaccinated had approximately 6.6 times higher odds of uptake than those without this intention (aOR = 6.69; 95% CI: [2.18 - 20.54]; p = 0.002). Compared with mothers holding an unfavorable perception (not very effective, not effective at all) of MV effectiveness, those holding a favorable perception (very effective, moderately effective) had more than 9 times higher odds of having their child vaccinated (aOR = 9.46 [2.92 - 30.62]; p = 0.001). Mothers perceiving strong (a great deal, a lot) community endorsement of the MV had more than 2.6 times higher odds of vaccine uptake (aOR = 2.62; 95% CI: [1.37 - 5.00]; p = 0.006) than mothers of children perceiving low endorsement (moderately, a little, not at all) or who did not know how to assess it.

A favorable opinion of close parents or friends regarding the child’s vaccination was associated with an uptake probability multiplied by nearly 4 (aOR = 3.84; 95% CI: [1.63 - 9.04]; p = 0.004) (Table 7).