Barriers to uptake of vaccination among communities in the city of Kigali, 2025

Abstract

Introduction

Immunization is a fundamental health component preventing an estimated 2.5 million deaths annually. Despite the documented successes of immunization, disparities of vaccine uptake exist in low- and middle-income countries and across their respective communities. This research assessed the barriers related to recall-based vaccination uptake among communities in Kigali city, the capital of Rwanda.

Methods

A mixed-methods study was implemented on a sample of 210 households and communities across the three districts of the City of Kigali. The quantitative study used a multistage sampling technique according to the WHO Immunization Coverage Survey to select the households from the districts of Kigali city. While the 3 focus group discussions and 6 interviews were implemented in communities not included in the quantitative survey. A thematic analysis was performed on the qualitative data. The quantitative data were analyzed using SPSS 23 and summarized numerically and graphically. Statistical tests performed for association were Chi-square tests and a binary logistic regression was used to establish the relationship between recall-based vaccine uptake and its associated factors. All statistical tests were considered significant when p < 0.05.

Results

Quantitative findings of this study reported that all 155 children ≤ 6 years were vaccinated. It also divided the barriers into four main categories: decision barriers, information barriers, accessibility barriers, and challenges in health facilities. A binary logistic regression predicted recall-based vaccine uptake based on 11 explanatory variables. The variables statistically significant were age and household size, with respective p < 0.001, and knowledge of vaccine-preventable diseases with a p-value of 0.038. Qualitatively, FGDs and interviews revealed fear of side effects, denial of vaccination in absence of husbands, long waiting time and poor communication from providers in facilities as factors that discourage attendance for immunization.

Conclusion

Barriers to recall-based categorization of vaccination frequency were mainly health system-related challenges, including insufficient supplies, costs for pregnant women for first-time vaccinations and denial vaccination in the absence of husbands. These findings call upon Rwandan health authorities and partners to address predictors of vaccination behavior associated with systemic challenges, intentional refusal and enhance immunization service delivery.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28521-8.

Introduction

Vaccination has been recognized as one of the most effective, equitable, and cost-efficient public health interventions globally. Optimal vaccination uptake is the complete and timely vaccination among eligible populations, in line with national and global targets [1]. Despite the global success, disparities in vaccination uptake persist between low and middle-income countries and among their communities. In 2018, middle-income countries recorded the highest number of unvaccinated children, equating to 70% [2, 3]. Global evidence reported in systematic reviews demonstrates that barriers to uptake are multifaceted. Service delivery factors, not exhaustive, include supplies, health workforce factors, urban dynamics and health care seeking factors as well as caregiver knowledge gaps, safety concerns, perceived risks, cultural beliefs, misinformation, financial and geographical accessibility [4, 5, 6, 7].

The 2023 World Health Organization Immunization Agenda reports that 20 million infants have not completed the full course of basic vaccines. More than 80% of them are unvaccinated – the “zero dose” children [2]. According to the WHO/UNICEF Estimates of National Immunization Coverage, 60% of these children are located in ten countries. These countries include Nigeria, India, Pakistan, Indonesia, Ethiopia, Philippines, Democratic Republic of Congo (DR Congo), Brazil, Angola and Vietnam [8]. Despite its benefits, various barriers were identified, Bangura et al. [9] divided them into three main categories. These include Parental/caretaker barriers, health system barriers and providers’ barriers. Parental/caretaker barriers were further divided into parental perception, maternal education and occupation, socioeconomic status and influence of male partner. Health system barriers include irregular supplies, transportation and limited human resources. Providers’ attitude toward mothers was reported to be a major immunization barrier [2, 8, 9].

Elsewhere in the literature, African authors address the influence of knowledge and awareness, but few explain the barriers to vaccine uptake. Sibanda et al. [10] identified misperceptions as associated factors for vaccine uptake and demonstrated high impact of prior vaccination behavior and having a university education. Miyahara et al. [11] addressed the birth-dose timeliness and identified predictors of vaccine uptake, which were maternal education and distance to facility. Sibanda et al. [10] as well as Miyahara et al. [11] identified additional predictors of vaccination uptake, including age, gender, education level, transportation difficulties, trust in local healthcare workers, vaccine misinformation exposure and distance to health care facilities. Namuyonga et al. [12] reported additional predictors, among which older age groups, females, and individuals with tertiary education were more likely to be vaccinated. Transportation emerged as one of the dominant predictors of vaccination.

In East Africa, Rwanda is widely known for having strong political leadership, innovative delivery models, and high immunization coverage [13, 14]. Studies identified several cross-cutting themes. Despite high awareness of vaccines, knowledge gaps persist regarding vaccine schedules, side effects, and the purpose of newer vaccines. Trust in community health workers and national systems remains strong; however, inconsistency in messaging can undermine understanding and timely uptake. Economic and logistical constraints in households, including transport costs, waiting time, competing priorities, and household workload, continue to play a crucial role in delayed/missed vaccination. At the points of service, occasional stockouts and poor communication with healthcare providers reduced service reliability, and caregivers who encounter such obstacles may delay or abandon subsequent visits. Misinformation remains a significant barrier, particularly in urban and peri-urban settings, where digital information spreads quickly. Hence, community engagement strategies require continuous reinforcement to address evolving concerns and new vaccine introductions [15, 16, 17]. In this framework, despite the reported statistics of high child immunization in Rwanda of 95.5% [18], this research on barriers towards recall-based vaccination frequency among communities in Kigali city will also contribute to identifying the barriers, which apparently are unknown, to boost and improve the immunization coverage.

Methods

Study design

A mixed quantitative-qualitative community-based cross-sectional study was implemented in Kigali, the capital city of Rwanda.

Study area and population

Rwanda is a landlocked country with an area of 26,338 km²; it has five international borders (Burundi, the Democratic Republic of the Congo, Tanzania, and Uganda). Administratively, the country is divided into 5 provinces, the study area is the province of Kigali, which harbor 13.2% of the country’s population (1,745,555/13,246,394). The province of Kigali, with an estimated population of 1,745,555 (86.9% urban and 13.1% rural) comprises three districts (administrative level 2) and is subdivided into sectors (35), cells (161), and villages (1,061). To avoid selection bias, all three districts of Kigali Province, namely Gasabo, Nyarugenge and Kicukiro, were included in the study.

Sampling size and technique

Quantitative section

The sampling technique used was a modified WHO/EPI Immunization Cluster Surveys Sampling technique [1]. According to this modification and with respect to the procedure of WHO/EPI, Kigali City province is administratively divided in 3 districts. A district was considered as a cluster, and in each cluster, a spatial sampling technique was used to select 10 geographically representative sectors. Then in each of the selected sectors, as shown in Fig. 1 below, 7 households were randomly selected (3 clusters x 10 sectors x 7 households).

Fig. 1.

Household selection as designed and implemented by Amr et al

Qualitative section

Three Focus group discussions (FGDs) were conducted in each of the three clusters (districts), targeting mothers of children < 5, ranging from 8 to 11 participants per group. Six interviews were also implemented in the three clusters, targeting religious/church leaders and traditional healers. Two interviews were conducted per cluster. Both FGDs and interviews were purposively biased to enable the inclusion of only participants residing in rural settings.

Data collection

The data were collected through a standardized, pre-tested research tool developed by the researcher in English language and administered in Kinyarwanda for easy understanding of the participants. The interviewed person was either the head him/herself, the spouse of the head of the household, or an adult member of the household (≥ 18 years). The quantitative data collected were the characteristics of the study participants (age, gender, education level, profession, residence, position in the household, number of children), the knowledge about vaccination (previous health education, source of information, recommended vaccines in Rwanda), and the perception, challenges and barriers towards vaccination uptake (recall-based). The focus group discussion and interview data were recorded in communities not covered by the quantitative research in order to obtain rural community perspectives on the knowledge of vaccine-preventable diseases, awareness of recommended vaccines, missed vaccinations, perceptions and challenges towards vaccine uptake, challenges in facilities and motivating factors for taking vaccines. When, for each of the 11 themes, the answers of 2–3 participants do not add any new information, the research team considered that pursuing the discussion would result in collecting unnecessary data. A threshold for getting new information was set when two or more focus group participants provided new information. Otherwise, the answers were considered redundant and were labelled as data saturation in this qualitative research.

The quantitative data were computerized through a template elaborated through Google Form and analyzed through the Statistical Package for Social Sciences (SPSS version 23). Qualitative data were collected as audio through a recorder operated by a moderator and a notetaker who spoke the language of participants (Kinyarwanda). Profiles of participants were obtained through an adopted guide from Dalan Development Consultants Firm, Sierra Leone, Qualitative Research in 2016. The moderator and notetaker transcribed the recorded data in English language. This data was coded by four researchers and comparisons were done to assess consistency, and discrepancies were resolved through discussion and consensus. Data were then digitalized in Microsoft Excel for further analysis.

Data analysis

Qualitative data

The data collected consisted of six main probing questions on identifying the demographic profile of participants, their views about vaccination willingness, hesitancy, barriers, challenges and enablers for vaccine uptake. This data was collected in Kinyarwanda through a standardized tool and tape recordings which were transcribed in English and entered into Microsoft Excel. On entering the standardized tool information with the transcribed recordings, the data were triangulated to get a comprehensive view of each participant in the focus group. The researchers adopted a hybrid, deductive and inductive methodology to code the entered information. A theoretical framework and concepts were developed from the original probing questions by a data coding team. Upon review of the entered data, responses were compared against the framework to identify the new themes that emerged from the respondents. Therefore, the final themes were a combination of the initial framework and new themes identified by the respondents. They were namely Knowledge of recommended vaccines, Vaccine hesitancy, Challenges to accessing vaccines, Challenges faced by special groups, Missed vaccinations and Motivating factors to taking vaccines. Final thematic summaries were triangulated with quantitative findings to enrich interpretation.

Quantitative data

Descriptive statistics summarized quantitative data numerically in terms of mean, standard deviation and median, and graphically in form of frequency tables and graphs. Associations between variables were assessed through Chi-Square tests, correlation, and ANOVA where appropriate. A binary logistic regression analysis was performed to establish the relationship between recall-based vaccine uptake and its associated factors. The outcome variable was recall-based vaccine uptake, which was created after cross-tabulation of the household vaccination status per year with the household age profile. The household vaccination status was recorded into 13 categories according to the year vaccination was reported to have been received. Whereas the household age profile was categorized into 3 groups, which were namely ≤ 6 years, 7–18 years and > 18 years. The result of the cross-tabulation was further regrouped into three categories and labelled as recall-based vaccine uptake (theoretical variable which conveys a theoretical concept of recalled vaccine frequency), as demonstrated in Fig. 2. The three categories were distributed as follows: those reported to be vaccinated in 1 year were coded as 1; they represent 38% of the study population (80/210). The second category was those reported to be vaccinated in 2 consecutive years and coded 2 (57%, 119/210). The third category coded 3 were those reported to be vaccinated over > 2 consecutive years (5%, 11/210). Furthermore, the theoretical concept variables of recalled vaccine frequency were transformed into an epidemiological variable to enable quantifying statistically the recall-based vaccine uptake. Therefore, the second and third categories were combined into one variable named reported vaccination in ≥ 2 consecutive years and coded 1. The final dichotomous variable of recall-based vaccine uptake was therefore coded as 0 for those who reported vaccination in 1 year, and 1 for those reported vaccination in ≥ 2 consecutive years. This definition of recall-based vaccine uptake is not an operational definition, as it was self-reported data from adult household members. Therefore, its operational use cannot allow the estimation of immunization completeness or timeliness. It provides the overall status/magnitude of vaccine uptake in the study population. Therefore, it is subject to vaccine uptake recall bias, and the transition across age groups cannot be precisely estimated. Table 5 displays the 11 predictors of recall-based vaccine uptake in the study population. All statistical tests were considered significant with p-value of < 0.05.

Fig. 2.

Formulation and grouping of the recall-based vaccine uptake as designed by Amr et al

Table 5.

Binary logistic regression predicting recall-based vaccine uptake based on 11 explanatory variables (n = 210)

Variables							95% CI
	B	S.E.	Wald	df	Sig.	OR	Lower	Upper
Age	-0.157	0.036	19.171	1	0.000	0.855	0.797	0.917
HH Census	1.187	0.215	30.567	1	0.000	3.277	2.151	4.991
Distance to nearest facility	-0.315	0.305	1.068	1	0.301	0.730	0.402	1.326
Educational level	-0.255	0.337	0.573	1	0.449	0.775	0.400	1.500
Gender	-0.186	0.389	0.228	1	0.633	0.831	0.388	1.780
Safety of vaccines	-0.092	0.181	0.259	1	0.611	0.912	0.640	1.299
Self-rated knowledge	-0.283	0.425	0.445	1	0.505	0.753	0.328	1.731
Knowledge of VPD*	1.084	0.522	4.310	1	0.038	2.955	1.062	8.220
Transportation challenges	1.038	0.627	2.735	1	0.098	2.823	0.825	9.655
Means of transportation	-0.089	0.236	0.143	1	0.705	0.915	0.576	1.453
Received info. From school	0.102	0.411	0.062	1	0.804	1.107	0.495	2.477

Bold values indicate statistically significant coefficients (p < 0.05) 

*Vaccine-preventable diseases

Results

Characteristics of the study participants

A total of 210 participants were enrolled from the three districts of Kigali city, as described in Table 1. Participants were mainly adult household members (66.2%, 139/210), with females representing 72.9% (153/210) and males representing 27.1% (57/210). The age ranged from 19 to 57 years, with a mean age of 34 (95% [CI: 33.1–34.9], n = 210). Adults of reproductive and childbearing age of 25–44 years were 88.1% (185/210) of the participants. The average household size was 3.5 members ± 1.1. Approximately, 96% (96.7%, 201/210) were employed, and those involved in business were predominant (35.9%) as displayed in Table 1.

Table 1.

Characteristics of study participants (n = 210)

Variable	N	%	Variable	N	%
Age in years (95%CI)	34 (CI: 33.1–34.9)		Marital status (n = 210)
18–24 years	5	2.4	Divorced	7	3.3
25–34 years	122	58.1	Married	162	77.1
35–44 years	63	30.0	Single	31	14.8
45–54 years	18	8.6	Widowed	10	4.8
≥ 55 years	2	1.0	Medical insurance (n = 210)
Gender (n = 210)			Private	75	35.7
Female	153	72.9	Public	135	64.3
Male	57	27.1	Educational level (n = 210)
District (n = 210)			Primary	14	6.7
Gasabo	70	33.3	Secondary	137	65.2
Kicukiro	70	33.3	University	56	26.7
Nyarugenge	70	33.3	Postgraduate	3	1.4
Household size*	3.5 ± 1.1 (1–6)		Occupation (n = 210)
Category in household (n = 210)			Businessman/woman	73	34.8
Adult member of HH	139	66.2	Farmer	5	2.4
Head of HH	68	32.4	Government employee	25	11.9
Relative/visitor	3	1.4	Healthcare personnel	2	1.0
Religion (n = 210)			Private institute	46	21.9
Christian	175	83.3	Skilled laborer	27	12.9
Muslim	31	14.8	Unemployed	7	3.3
None	4	1.9	Unskilled laborer	25	11.9
			Chronic condition (n = 210)
			No	192	91.4
			Yes	18	8.6

*Data presented as Mean ± SD (range) for continuous variable

Bold entries indicate subheadings denoting distinct categories within the table for formatting purposes only. It does not indicate statistical significance

Knowledge, attitudes, and practices of vaccination uptake

Vaccination history and knowledge patterns

Vaccination history and practices of participants varied significantly by household structure. Households with children ≤ 6 years represent 73.8% (155/210), children 7–11 were 6.7% (14/210), and with children ≥ 12 years were 19.5% (41/210). Among these categories, all 155 children aged ≤ 6 years were reported to benefit from vaccines. Vaccination rates were high during the COVID-19 pandemic years, with 31% coverage in 2020–2022. Vaccine knowledge among respondents identified 73.8% (155/210) moderate knowledge, 22.9% (48/210) good knowledge and 3.3% (7/210) poor. Attitudes of participants show that 99.5% believe in the effectiveness of vaccines, with 96.7% rating their knowledge as good/moderate (Table 2).

Table 2.

Vaccination history and practices

Variable	N	%	Variable	N	%
HH Census			Respondent vaccination status (n = 210)
Total population	724		2025	7	3.3
Mean	3.45 (95% CI: 3.29–3.61)		2024	26	12.4
Min	1		2023	20	9.5
Max	6		2020–2022	157	74.8
Total population*	469		Type of vaccination received*
Mean	2.23 (95% CI: 2.13–2.33)		Childhood vaccination	1	0.5%
Min	1		Pregnancy vaccination	106	50.5%
Max	4		Elderly vaccination (> 45)	1	0.5%
HH Vaccination History			After-exposure vaccination	0	0.0%
Total vaccinated*	460		Occasional vaccination	102	48.6%
Mean	2.19 (95% CI: 2.08–2.30)		Vaccine knowledge rate (n = 210)
Min	1		Good	48	22.9
Max	4		Moderate	155	73.8
HH Vaccination Year (n = 210)			Poor	7	3.3
2020–2022	65	31.0	Vaccines prevent diseases (n = 210)
2023	53	25.3	Yes	209	99.5
2024	45	21.4	No	1	0.5
2025	47	22.4

*Variable calculated by age categories of household members

Bold entries indicate subheadings denoting distinct categories within the table for formatting purposes only. It does not indicate statistical significance

Despite the positive practices revealed during household surveys, FGDs, and interviews showed different responses on missed vaccinations.

“However, I personally believe that using natural plants can help prevent diseases, and I think this approach is undervalued in the current system. I do not trust vaccines. Our ancestors were treated effectively with plants, and I find modern vaccines costly and less reliable compared to traditional remedies”. (Traditional healer, KII, Nyarugenge)

The majority of respondents on FGDs had different justifications such as conflicting work schedules, accessibility challenges and absence of family support. A significant finding among the discussions of Gasabo sector was the insufficient supplies of vaccines and lack of knowledge on working days and hours.

“Sometimes the health workers send you back because the number of vaccines they have prepared is less compared to the number of people who came, and sometimes you go and find that they didn’t work on that day because it’s a day off, or you were late and they send you back.” (Mothers, FGD, Gasabo) In addition to the knowledge gaps revealed, a significant number of respondents expressed prioritization of infant vaccines until 15 months of age. The trend of vaccination history of households showed a variation with the emergence of epidemics, as shown in Fig. 3. It declined from 2020 to 2022 to 2024 with 31% (65/210) and 21.4% (45/210) respectively.

Fig. 3.

Trend of recall-based vaccination uptake from 2020 to 2025

Knowledge of vaccine-preventable diseases (VDP)

Most of the participants were able to identify, HPV (100%, 210/210), Polio (99.5%, 209/210), Hepatitis B (99.5%, 209/210), Pneumococcus infection (98.6%, 201/210), Tetanus (96.7%, 203/210), and Measles (94.8%, 199/210) as vaccine-prevented diseases (Supplement 1). These findings were supported by FGDs and interviews that revealed general awareness on vaccinations recommended for children, where a very high number of respondents were able to indicate the various vaccination schedules.

“Participant 1: I know of the vaccines given to children. Participant 2: I know the measles vaccine, TB and tetanus for children.” (Mothers, FGD, Gasabo).

“There are vaccines that are given for the newborn babies, at one and a half months, 2 and a half months, 3 and a half months. There is TB vaccine, pneumonia. As children get older, are given diarrheal vaccine, meningitis, tetanus.” (Religious leader, KII, Kicukiro).

Attitudes and practices toward vaccination uptake

Pain was the predominant concern for 84.3% (177/210) of the participants. Approximately, 58% (58.6%, 123/210) were concerned about allergic reactions, 54.3% (114/210) about fever and 18.1% (38/210) had no concerns or fears. Participants reported that information on vaccination was predominantly obtained through mass media (96.7%, 203/210), healthcare workers (95.2%, 200/210) and friends, co-workers or neighbors (94.8%, 199/210). Secondary sources of information were reported to be schools and social media. 10% (10%, 21/210) of the participant received the information from religious and traditional leaders. Information provided by health care workers included side effects (100%, 210/210), when to get vaccinated (97.1%, 204/210), where to get vaccinated (96.7%, 203/210), recommended vaccines (95.7%, 201/210), and instructions on vaccination card use (95.2%, 200/210) (Supplement 2). FGDs and interviews revealed variation in the perception and willingness toward vaccination.

While majority recognized vaccines as preventive measures, some showed concerns and were inclined towards the use of herbal remedies.

“I’m less confident about children’s vaccines because they are not produced locally, and I sometimes worry that foreign vaccines could contain unknown substances. Vaccines are important, and I understand their value. However, in our church, many people still hesitate because they believe vaccines may contain hidden symbols or technologies related to the devil.” (Church leader, KII, Nyarugenge).

Barriers to vaccination uptake as reported by participants

The barriers towards vaccination were categorized into decision-related barriers, information barriers, accessibility barriers, and challenges in health facilities, as described in Table 3. Fear of side effects dominated decision-related barriers, representing 91.4% (192/210) of the participants, while fear of injection represented 90% (189/210). Efficacy doubts such as misconceptions and disbeliefs were minimal (13.3%, 28/210). Religious and traditional beliefs were barriers to 3.3% (7/210), while family influence was not a reported as a barrier. Accessibility challenges were divided into transport cost/time (10.9%, 23/210) and cost of vaccines (0.5%, 1/210). Majority (93.8%, 197/210) of the participants obtained their vaccination in health care centers. 86% (86.2%, 181/210) of participants have vaccination facilities within 1–5 km, and more than half (71.4%, 150/210) use motorcycle to reach their destination. Long waiting time (82.4%, 173/210) was the main health facility challenge reported. FGDs and interviews revealed that religious beliefs, misinformation, and service-related challenges were the base of vaccine hesitancy. Herbal medicine is believed by many, as reported by traditional healers and religious leaders, to treat and prevent diseases. Discussions have revealed that participants’ beliefs are majorly influenced by the church and fear of side effects/ poor medical service discourage vaccination.

Table 3.

Barriers towards vaccination uptake (n = 210)

Barriers	Number	%*	Barriers	Number	%*
Decision-related barriers (n = 210)			Distance to the nearest facility (n = 210)
Fear of side effects	192	91.4	< 1 km	16	7.6
Fear of injection	189	90.0	> 5 km	13	6.2
Efficacy doubts	28	13.3	1–5 km	181	86.2
Religious/Traditional beliefs	7	3.3	Means of transportation (n = 210)
Family members influence	0	0	Motorcycle	150	71.4
Access to services (n = 210)			Walking	31	14.8
Transportation (cost & time)	23	10.9	Car (taxi or bus)	25	11.9
Cost of vaccines	1	0.5	Bicycle	4	1.9
Where to get vaccinated (n = 210)			Facility challenges (n = 210)
Health center	197	93.8	Long waiting time	173	82.4
Private clinic	13	6.2	Getting a day off for vaccination	1	0.5
			Expensive vaccine	0	0
			Unavailability of vaccines	0	0

*In this table, for the six barriers, their respective percentages were calculated as the number who reported present/yes, divided by 210 and multiplied by 100

Bold entries indicate subheadings denoting distinct categories within the table for formatting purposes only. It does not indicate statistical significance

“Most people I receive with hesitation about vaccines are those with some side effects like skin rashes asking if I have a drug that can replace vaccine but I didn’t encourage them to not get vaccinated, instead I help them with these manifestations.” (Traditional healer, KII, Gasabo).

“Sometimes nurses don’t give good services, they talk to us in a bad way, so you decide not to go back next time”. (Mothers, FGD, Kicukiro)

The main challenges to vaccination reported across the districts revolve around access, vaccination timing, and service delivery.

“One major challenge is that vaccination is only offered on specific days. If those days coincide with work or income-generating activities, people may miss their chance. Extending vaccination days would help a lot.” (Church leader, KII, Nyarugenge).

Economic and communication barriers such as transport, long distance to health centers, charges, crowding, poor communication from nurses and long waiting times, especially for pregnant women, were frequently mentioned. Some facilities require the presence of a husband or charge fees for vaccination, hence discouraging attendance.

“Participant 1: They charge a lot for pregnant women at the first encounter. Participant 2: Some facilities require the presence of a husband for first-time vaccination during pregnancy, which is difficult if the husband isn’t available.” (Mothers, FGD, Gasabo).

“Sometimes when you go and wait for nurses, you get hungry because of waiting for a long time, and when you go to find something to eat and come back, you find that they have arrived and they send you home saying you’re late, and they don’t care about explanations.” (Mothers, FGD, Nyarugenge).

Urban residents reported fewer access issues, in contrast other district residents advocated for improving accessibility, affordability and respectful service to increase vaccination uptake. Vulnerable groups including pregnant women, older adults and new mothers face unique barriers such as weakness after childbirth, shame from unofficial pregnancies and distance to health centers. Fetal complication due to vaccine was among belief reported by participants.

“Older people and pregnant women have difficulty moving” and that “first-trimester vaccination can take the whole day”.(Mothers, FGD, Gasabo).

“The challenges I saw are sometimes due to unofficial/unplanned pregnancy. This makes them feel ashamed in society and in turn not going to the health center. Another issue may be accessibility to the health care facility (Long distance). This makes them skip vaccines until the time of giving the birth.” (Religious leader, KII, Gasabo).

“Pregnant women, infants, and older people are different from other people. Some women think that when someone is pregnant, she is not able to be vaccinated because they think that these vaccines can cause many issues to her baby. Other people fear the side effects of the vaccine and then they deny going for vaccination.” (Religious leader, KII, Kicukiro).

Empowerment of community health workers due to their good leadership to deliver vaccination to vulnerable groups was a strategy suggested by respondents. The fear of epidemics such as COVID-19 was recognized by participants in improving awareness of vaccination.

Relationship between awareness, barriers, and vaccination uptake

The outcome variable recall-based vaccine uptake, coded as 1 when reported vaccination in one year, 2 when reported vaccination in 2 consecutive years, and 3 when reported vaccination in > 2 consecutive years, was regrouped as a dichotomous variable, with values of 0 for reported vaccination in one year and 1 for reported vaccination in ≥ 2 consecutive years.

A Binary logistic regression analysis was performed to predict the outcome variable based on 11 predictors. The model fitting information assessed that the regression equation fit the observed data well, with the Nagelkerke r² of 0.340 measuring the proportion of variance in the dependent variable explained by the 11 indicators and a correct classification of 76.7%. The normal value of r² ranges between 0 and 100%, with a higher r² indicating a better fit. This shows that our model fits the equation moderately up to 34.0%. The likelihood test revealed a Chi-Square value of 60.357, with a p < 0.001, indicating that despite the model moderately fitting the equation, the 11 predictors selected are significantly representative of the data. Table 4 below indicates moderate model fitting with an overall value of 34.0%.

Table 4.

Model summary and Pseudo R

-2 Log likelihood	Cox & Snell R Square	Nagelkerke R Square
218.745a	0.250	0.340

Three factors, as shown in Table 5 below, were statistically significant predictors. Age and household size with p-values of < 0.001 and knowledge of vaccine-preventable diseases p = 0.038. A factor, transportation challenges, with an OR = 2.823 (95% [CI: 0.825–9.655], p = 0.098), appeared to be a potential contributor and should not be retained as an explanatory variable due to its wide confidence interval.

One explanatory variable contributed > 1 time, and six variables contributed < 1 time to explaining the occurrence of recall-based vaccine uptake. In overall, the binary logistic regression revealed that age, household size (HH Census) and the knowledge of vaccine-preventable diseases were the statistically significant contributors.

Discussion

This study revealed the patterns and barriers affecting recall-based vaccination uptake in Kigali city. The barriers were grouped into four major barriers affecting communities. They were decision barriers, information barriers, accessibility, and health facility challenges. Decision-related barriers were the most frequently reported and largely modifiable; they include fear of side effects, fear of injection, efficacy doubts, religious and traditional beliefs. Information barriers encompassed patient- healthcare workers communication on side effects, when and where to get vaccinated, recommended vaccine, utilization of vaccination card and children’s growth pattern. Accessibility barriers were linked to transportation cost and time, location of vaccination centre, distance to the nearest facility and means of transport. Health facility challenges were long waiting time, absences from work for vaccination, expensive vaccines, and shortage of vaccines.

The identified decision-related barriers such as sociocultural beliefs and efficacy doubts align with the existing evidence on vaccine hesitancy. The refusal of vaccination by traditional healers and religious leaders observed in this study mirrors the findings of Iwu et al. [19], who reported the prioritization of herbal treatment and demonstrated the communities’ trust in traditional remedies. In the literature, studies established the significant influence of religious leaders and traditional healers, thus advocating for critical integration to enhance vaccine legitimacy and acceptance [19, 20]. Facility challenges identified in this study were vaccine stockouts, lack of information on service days and long waiting hours, as published by Bangura et al. [9], who described them as health system barriers that hinder immunization service delivery. This study findings pointed out distance of facilities as a challenge. It also stressed the importance of the attitude of health workers and their communication with healthcare beneficiaries as critical factors for vaccine uptake. Knowledge of vaccine-preventable diseases was generally high for the core childhood vaccines under the Expanded Programme on Immunization (EPI) in Rwanda, such as Measles, Polio, Tetanus, and Hepatitis B, reflecting strong national investment in vaccination education. Reports from Rwanda Biomedical Center and UNICEF documented the gender differences in vaccination knowledge, with females demonstrating higher levels of awareness [13, 14]. In the findings of this study, this was reflected and could be explained by the women-led caregiving roles within households. Secondary education or higher was reported by majority of the participants and was associated with increased vaccine awareness. These findings align with Trepanowski and Drazkowski [21] who pointed out the association between education and vaccination uptake.

Regarding the trend of reported recall-based vaccine uptake, from 2020 to 2025, maybe related to the past experience of COVID-19 outbreak and, later on, the recent outbreaks of Monkeypox and Marburg disease (no recommended vaccines yet available), which the country experienced subsequently in 2024 and 2025. This may be supported by the qualitative research across the replies of the religious leaders, who mostly recognized that vaccination in case of epidemics is a governmental decision, and ethically, all citizens should comply.

This study identified three statistically significant predictors of recall-based vaccine uptake, which were age (OR = 0.855, 95% [CI: 0.797–0.917], p = < 0.001), household size (OR = 3.277, 95% [CI: 2.151–4.991], p = < 0.001), and knowledge of vaccine-preventable diseases (OR = 2.955, 95% [CI: 1.062–8.22], p = 0.038). The regression analysis findings revealed that seven other predictors explained recall-based vaccine uptake by < 2 times. The contribution of transportation challenges of > 2 times (OR = 2.823, 95% [CI: 0.825–9.655], p = 0.098) was not considered due to its wide confidence interval. Namuyonga et al. [12] reported predictors of vaccine uptake, which were older age groups (OR = 2.94, p < 0.001), females (OR = 1.63, p = 0.001), individuals with tertiary education (OR = 4.26, p < 0.001) which more likely to be vaccinated and absence of transportation difficulties (OR = 5.97, p < 0.001).

The predominance of modifiable decision-related barriers emphasize the importance of strengthening immunization communication strategies within the national policies of Rwanda. Health promotion packages should include culturally tailored communication tools, unified information, in addition to appropriately trained healthcare workers in delivery of messages, counselling and interpersonal communication. This will efficiently address fears of injections, side effects and reinforce trust in vaccines. Engagement and integration with traditional religious systems within communities would further boost uptake by aligning beliefs with health outcomes and goals. The role of schools as the primary source of information is paramount, hence necessitating the evaluation of the quality of information delivered to ensure informed confidence in vaccines rather than hesitancy and uncertainty. Furthermore, as recommended by many respondents in themes of motivating factors toward vaccination, addressing facility barriers such as waiting time, vaccine availability, and flexible waiting hours could improve uptake among special groups such as pregnant women and older populations.

Limitations

A limitation of this study was related to the mixed-methods design, which included a quantitative cross-sectional community-based survey of 210 adult participants reporting immunization uptake in their households. In a country like Rwanda, where population data are available at all levels (administrative level 1/Province to administrative level 5/Village), a multistage sampling technique could be more appropriate for implementation, as it will enable the combination, according to the administrative levels, of various probability sampling techniques. This may end up providing a more representative sample size, with its associated degree of accuracy and the power of the sample size calculation. However, the method used was a modified WHO/EPI survey method, which was more appropriate for a rapid survey, and in the best case, would provide a sample size ranging from 300 to 400 participants. Therefore, most often, it should be restricted to descriptive statistics and inferential statistics limited to associations rather than demonstrating a relation-based. We acknowledge the limitations for generalization in future studies. A second crucial limitation of the mixed methods quantitative-qualitative design, in which the qualitative study was limited only to the rural areas and not executed at the same time as the quantitative study, but performed back-to-back. However, as the qualitative study provides information from the rural community perspective, this design missed the information from the urban communities survey, with a selection bias. In such mixed-methods, both quantitative and qualitative studies should be implemented in both rural and urban populations to strengthen the triangulation, which is one of the robust methods to assess the validity and reliability (worthiness) of qualitative findings.

Moreover, the qualitative study was limited to a few focus group discussions and interviews. Hence, the quantitative study findings were not linked with the qualitative study. The quantitative survey, which was the driving method in this mixed study, has a limitation as the answers provided by the household members were not tested through the Cronbach’s alpha test for reliability and validity. Furthermore, the thematic analysis has not been performed through a software for thematic analysis (e.g. NVivo). It was an expert opinion which led to the grouping of the qualitative findings in themes.

Another limitation was related to the research survey template, which was a Google form which aggregated the variables with multiple answers. Therefore, the data curation and compilation had been hectic and may introduce bias in defining some key predictors.

Additional crucial limitation of this study was the outcome variable called recall-based vaccine uptake, while it is not the standard definition of vaccine uptake as full immunization status, age-appropriate vaccination and perception proportion could not be derived. Therefore, future research should use the standard definition of vaccine uptake. Nevertheless, this research, being a mixed-methods study, can be considered as an explorative study to better understand the barriers related to vaccination uptake. Future research can employ a mixed-methods quantitative-qualitative design to be implemented in the same communities to ensure triangulation and to facilitate the link between quantitative and qualitative data for estimating vaccination coverage.

Conclusion

This study addressed the barriers to vaccination, beyond early childhood, related to decision-making, communication, and the health systems. These findings call upon Rwandan health authorities and its partners to sustain and enhance vaccination uptake through targeted communication strategies, improvements in service delivery and community engagement.

Abbreviations

EPI

Expanded Programme on Immunization

FGD

Focus Group Discussions

HCW

Healthcare Workers

HH

Households

KII

Key Informant Interviews

SPSS

Statistical Package for Social Sciences

VPD

Vaccine-preventable Diseases

WHO

World Health Organization

Authors’ contributions

**AMH** ; developed and implemented the research proposal, participate in the data analysis, draft the initial manuscript and read the final manuscript prior to submission.**AM** ; supervised the development of the research proposal, coordinated the implementation of the research, guided the data curation and evaluation, read the initial manuscript and proof read the final manuscript prior to submission.**HM** ; edited the manuscript, read the final manuscript prior to submission.**HM; ** participated in the development of the research tools, appraised the results presentation, read the initial and final manuscript.**MN; ** guided the data evaluation and proof read the final manuscript prior to submission.**MH** ; authorized the implementation of the research proposal and proof read the final manuscript prior to submission.**FY** ; supervised the development of the research proposal and its implementation, guided the qualitative survey, reviewed the focus group discussion instruments, critically analyzed the results, read the initial and final manuscript, proof read the final manuscript prior to submission.All the authors have read the final manuscript prior to submission.

Funding

All the costs related to this research were funded by Amr Magdi Hassan in the framework of the partial fulfillment of his MBBS graduation thesis.

Data availability

All the data related to and supporting the findings of this research is included in the manuscript and additional data were attached as supplement.

Declarations

Ethics approval and consent to participate

Ethical approval was sought from the Faculty of Medicine Review Committee and the University of Medical Sciences and Technology (UMST) Kigali – Institutional Review Board (IRB), which is an approved extension of the Rwanda National Ethics Committee (RNEC). The ethical approval power to UMST-IRB was granted in accordance with the Helsinki Declaration, the World Medical Association International Code of Medical Ethics, and the American Medical Association Guidelines and Protocol (1957, revised in 2021). These guidelines were followed for granting ethical clearance for the implementation of this study. Ethical approval to access communities was also sought from the Mayor of the Province of Kigali City and the District Executive Administrators of Gasabo, Kicukiro, and Nyarugenge. Approval from the District Executive Administrator grants access to the sectors, cells, and villages, which include the households.

The study implemented a population-based survey that did not use any invasive procedures and was implemented with respect to all community-based research ethics codes, professionalism, confidentiality, privacy, and free participation, without coercion. Ethical approval to access communities was granted from District Administration Offices. Informed consent was obtained from participants verbally during house-to-house surveys, focus group discussions and informant interviews. The consent was recorded by the interviewer in the interviewer-administered questionnaire. Participants during FGDs and interviews were registered with unique identifier codes developed from their initials.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.