Explanatory Factors of Vaccination Dropout Among Children Aged 12 to 23 Months in the Kikula Health Zone, Democratic Republic of Congo: Cross-Sectional Analytical Study

Introduction

Vaccination is recognized as one of the most effective public health strategies for reducing childhood morbidity and mortality. The World Health Organization (WHO) recommends that all children receive vaccines against tuberculosis, diphtheria, pertussis, tetanus, poliomyelitis, measles, hepatitis B, Haemophilus influenzae type b, pneumococcus, rubella, and yellow fever before their first birthday in endemic countries [1]. Despite these recommendations, global immunization coverage has declined in recent years. Between 2019 and 2021, coverage with the third dose of the diphtheria-tetanus-pertussis vaccine (DTP3) fell from 86% to 81%, leaving 25 million children without full protection, including 18.2 million “zero-dose” children [2]. Similar declines were observed for the first dose of measles-containing vaccine (MCV1), which dropped from 85% to 81% [2]. The COVID-19 pandemic contributed to the largest sustained backslide in routine immunization in 3 decades, disproportionately affecting low- and middle-income countries [3,4].

In the WHO African Region, progress has remained limited. DTP3 coverage increased only modestly from 70% to 74% between 2013 and 2019, while MCV1 coverage declined slightly from 70% to 69% [1]. Updated assessments from 2023 confirm that Africa continues to account for the highest proportion of zero-dose children globally, driven by persistent inequities, service disruptions, and health system weaknesses [5]. In 2021, an estimated 25 million African children were not vaccinated, including 734,000 in the Democratic Republic of Congo (DRC) [1].

In the DRC, the Expanded Programme on Immunization aims to achieve ≥90% coverage for all antigens. The national immunization schedule recommends Bacillus Calmette-Guérin (BCG) at birth; 3 doses of DTP-HepB-Hib at 6, 10, and 14 weeks; and oral polio vaccine, pneumococcal conjugate vaccine, rotavirus vaccine, and measles vaccine at 9 months. However, full immunization coverage remains below 50%. The 2017‐2018 Multiple Indicator Cluster Survey (MICS) reported that only 35% of children aged 12 to 23 months were fully vaccinated, while 20% had received no vaccines at all [6,7]. Financing delays, recurrent supply shortages, and geographic inequities are major contributors to low coverage [6,7]. Recent multicountry analyses in sub-Saharan Africa confirm that maternal education, socioeconomic status, and health system accessibility remain key determinants of incomplete immunization [8,9].

In the Haut-Katanga province, administrative data suggested that 88.3% of children aged 0 to 11 months were fully vaccinated in 2021. Yet, vaccine-preventable diseases persisted: 1347 measles cases (including 349 zero-dose children), 34 acute flaccid paralysis cases, 101 meningitis cases, 14 neonatal tetanus cases, and 74 cholera cases. The BCG-measles dropout rate reached 11.4%, representing 35,680 children who did not complete their vaccination schedule [10]. In the Kikula Health Zone, administrative coverage was reported at 113.7% in 2022, but independent surveys conducted between 2020 and 2023 revealed that only 26% of children aged 12 to 23 months were fully vaccinated, with no improvement over time [11]. The validated dropout rate in this study was 38.3%, replacing earlier contextual estimates of 26% from prior surveys. This discrepancy between administrative data and field-based findings highlights structural, social, and health system factors influencing vaccination dropout.

No prior study has specifically examined the determinants of vaccination dropout in the Kikula Health Zone, representing a critical evidence gap. Guided by the Andersen behavioral model of health service use [8] and the health belief model [9], which emphasize predisposing, enabling, and perception-related factors influencing health-seeking behavior, this study aimed to assess the prevalence and determinants of vaccination dropout between BCG and measles vaccines among children aged 12 to 23 months in the Kikula Health Zone, Likasi, DRC.

Methods

Study Design and Setting

We conducted an analytical cross-sectional study between April 22 and May 22, 2025, in the Kikula Health Zone, Likasi city, Haut-Katanga province, DRC. The zone includes urban and periurban health areas with heterogeneous access to immunization services. The Kikula Health Zone was purposively selected due to its high vaccination dropout rate and the documented discrepancy between administrative coverage (>100%) and survey-based estimates (<30%).

Study Population and Eligibility Criteria

The study population consisted of children aged 12 to 23 months and their mothers or primary caregivers.

Inclusion Criteria

The inclusion criteria were as follows: children aged 12 to 23 months at the time of the survey, residence in the health zone for 6 months or more, presence of the mother or primary caregiver, and provision of written informed consent.

Exclusion Criteria

The exclusion criteria were as follows: refusal to participate and children severely ill and unable to participate.

Determination of Vaccination Status

Children with vaccination cards had their immunization status verified directly by card inspection, and children without vaccination cards were also included in this study. For these cases, vaccination status was determined through caregiver recall and, when available, verification against health facility registers. This approach ensured that all eligible children were represented in the study while acknowledging the potential for misclassification bias.

Sampling Strategy

A three-stage cluster sampling technique was applied. First, health areas were selected using probability proportional to size. Second, streets or segments were randomly selected within each health area, with support from community health workers. Third, households were visited systematically (eg, every third household). If multiple eligible children were present, the youngest was selected to reduce recall bias. This approach ensured representativeness despite the absence of a complete sampling frame.

Sample Size

Sample size was calculated using the Schwartz formula for cross-sectional studies:

n=Z²× P(1−P)/d²

where Z=1.96 (95% confidence level), P=.26 (dropout prevalence from previous Kikula surveys), and Cohen d=0.05 (margin of error).

The initial estimate (n=296) was adjusted for a design effect of 1.5 and rounded to 300 to account for nonresponse. In total, 322 households were approached, and 300 households completed the survey (response rate 93.2%).

Recruitment Procedures

Prior to data collection, meetings were held with health zone authorities and community health workers to explain objectives and obtain authorization. Investigators introduced themselves at households, explained the study, verified eligibility, and obtained written consent. Noneligible or refusing households were skipped according to the sampling interval. Daily supervision ensured adherence to protocol and data quality.

Data Collection

Data were collected using a structured, interviewer-administered questionnaire covering maternal sociodemographic characteristics (eg, age, education, marital status, occupation, religion, and duration of residence), child characteristics (eg, age, sex, and birth order), vaccination status verified by card, maternal knowledge (eg, importance of vaccination, schedule, vaccine-preventable diseases, and adverse events following immunization), perceptions of services (eg, reception, waiting time, information, vaccine availability, and perceived cost), and health system factors (eg, distance to the facility, participation in campaigns, and possession of a vaccination card).

The questionnaire was pretested in a nonselected health area. Investigators received training on ethical procedures, standardized administration, and card verification. Daily debriefings corrected inconsistencies.

Variables

Dependent Variable

The dependent variable was vaccination dropout, defined as receipt of BCG but not the measles vaccine. The dropout rate was calculated using the following formula: dropout rate=(number of children who received BCG but not measles)/(number of children who received BCG)×100. The variable was coded as 1=dropout and 0=no dropout.

Independent Variables

The independent variables included the following: maternal sociodemographics (eg, age ≤25 y vs >25 y; education: none, primary, secondary, and higher; marital status; occupation; religion; and residence ≤5 y vs >5 y), child characteristics (eg, age, sex, and birth order), maternal knowledge (eg, importance, schedule, diseases, and adverse events), perceptions of services (eg, reception, waiting time, information, availability, and cost), and health system factors (eg, distance: <5 km, 5‐10 km, and >10 km; campaign exposure; and possession of a vaccination card).

Thresholds

Thresholds (eg, maternal age, distance, and residence duration) were based on regional literature and operational practice.

Operational Definitions of Key Variables

Use of child health services was defined as whether the child had ever been taken to a health facility for preventive or curative services (eg, growth monitoring, vaccination, and consultation) in the past 12 months. “Yes” indicated the child received at least one service, while “no” indicated the child had not received any service.

Recourse to child health service was defined as the caregiver’s reported use of formal health services for the child (eg, vaccination sessions, consultations, and campaigns). “Yes” indicated that the caregiver reported attending at least one vaccination session or campaign, while “no” indicated that the caregiver reported never attending.

Possession of a vaccination card was defined as whether the caregiver could present the child’s vaccination card at the time of the survey. “Yes” indicated that the card was available and verified, while “no” indicated that the card was absent, and vaccination status was determined by recall and/or registers.

Data Management

Incomplete questionnaires and cases without vaccination cards were excluded. Data were entered into Epi Info (version 7.2.3.1; Centers for Disease Control and Prevention), cleaned, and then exported to SPSS (version 26; IBM Corp). A partial double-entry procedure was performed to check consistency. No imputation was applied.

Statistical Analysis

Data analysis was performed using descriptive, bivariate, and multivariate analyses:

• Descriptive analysis—frequencies and percentages were calculated for categorical variables, while means or medians were calculated for continuous variables.
• Bivariate analysis—chi-square or Fisher exact tests were used. Variables with P<.20 were retained for multivariate analysis.
• Multivariate analysis—logistic regression was performed to estimate adjusted odds ratios (AORs) with 95% CIs. Robust SEs were used to account for clustering. Multicollinearity was assessed using variance inflation factors. Statistical significance was set at P<.05.

Extremely high AORs (eg, >30,000) were interpreted cautiously, acknowledging that they likely reflect sparse data or near-perfect separation rather than precise effect sizes.

Ethical Considerations

Ethics approval was obtained from the Ethics Committee of the School of Public Health, University of Kinshasa (ESP/CE/099/2023). Written informed consent was obtained from all participants. Data were anonymized, coded, and securely stored. No financial incentives were provided to participants, in accordance with national guidelines. The study results represented conservative estimates due to the exclusion of children without documented vaccination records.

Results

Overview

Among 300 children aged 12 to 23 months included in the survey, 115 (38.3%) had dropped out between BCG and measles vaccination, while 185 (61.7%) completed the schedule (Table 1).

Most respondents were mothers (295/300, 98.3%). Children were evenly distributed by sex (male: n=153, 51% and female: n=147, 49%). Most mothers were aged 25 years or younger (n=247, 82.3%). Education levels were no schooling (n=7, 2.3%), primary (n=70, 23.3%), secondary (n=207, 69.0%), and higher or university (n=16, 5.3%). Most mothers were housewives (n=230, 76.7%) and married (n=259, 86.3%). Protestant affiliation was predominant (n=164, 54.7%). Duration of residence exceeded 5 years for 174 (58%) mothers, while 126 (42%) had lived in the area for 5 years or less (Table 2).

Vaccination was considered “very important” by 103 (34.3%) of 300 mothers and “important” by 178 (59.3%) mothers. Only 8 (2.7%) mothers considered vaccination “not important.” More than half (n=169, 56.3%) of the mothers did not know their child’s vaccination status. Knowledge of adverse events following immunization was reported by 158 (52.7%) mothers, while 142 (47.3%) were unaware. Fear of side effects was expressed by 148 (49.3%) mothers. Nearly all households (n=289, 96.3%) reported low income. Concerns about vaccine safety for the child were expressed by 195 (65%) mothers (Table 3).

Reception at health facilities was rated “good” by 294 (98%) of 300 respondents. Waiting times were less than 1 hour for 163 (54.3%) respondents, 1 to 2 hours for 108 (36%), and more than 2 hours for 29 (9.7%). Information about vaccines was received by 234 (78%) respondents. Distance to facilities was less than 5 km for 223 (74.3%) respondents, 5 to 10 km for 54 (18%), and more than 10 km for 23 (7.7%). Most respondents (n=271, 90.3%) reported having received information about vaccination campaigns (Table 4).

Bivariate Analysis

Dropout was significantly associated with the following: duration of residence of 5 years or less (odds ratio [OR] 1.95, 95% CI 1.20‐3.72; P=.007), nonuse of child health services (OR 4.62, 95% CI 2.20‐9.01; P<.001), absence of a vaccination card (OR 44.19, 95% CI 18.98‐102.91; P<.001), and lower maternal education (no schooling: OR 42.0, 95% CI 3.16‐556.5; P=.001 and primary education: OR 7.41, 95% CI 1.56‐35.1; P=.001; Table 5).

Multivariate Analysis

After adjustment, only possession of a vaccination card remained a strong independent predictor of dropout: children without a vaccination card had an AOR of 30.9 (95% CI 11.6‐82.0; P<.001).

Extremely high ORs and wide CIs should be interpreted with caution. These values likely reflect sparse data or near-perfect separation rather than precise effect sizes.

Discussion

Principal Findings

This analytic cross-sectional study of 300 mother-child pairs in the Kikula Health Zone revealed a validated vaccination dropout rate of 38.3% between BCG and measles vaccines. Possession of a child health or vaccination card was the strongest independent predictor of dropout, with children lacking cards having nearly 31-fold higher odds of incomplete immunization. Other factors, such as shorter duration of residence, lower maternal education, and nonuse of child health services, were significant in bivariate analysis but lost significance in multivariate models, likely due to collinearity and small subgroup sizes. These findings highlight the critical role of vaccination documentation and maternal knowledge in sustaining immunization coverage.

Interpretation in Light of Theoretical Frameworks

The results can be interpreted through the Andersen behavioral model of health service use [8] and the health belief model [9]. Predisposing factors such as maternal age, education, and marital status influenced vaccination behavior, although their effects diminished after adjustment. Enabling factors such as possession of a vaccination card and use of child health services facilitated completion of the schedule. Perceptions such as maternal fear of side effects and lack of knowledge about vaccination status contributed to dropout. These frameworks underscore that vaccination behavior is shaped not only by access but also by perceptions and enabling resources.

Perceptions—maternal fear of side effects and lack of knowledge about vaccination status contributed to dropout.

Comparison With Previous Studies

Our findings are consistent with national surveys (MICS 2017‐2018) reporting low full immunization coverage (<40%) and high dropout rates in the DRC [6,7]. These findings highlight the critical role of vaccination documentation and maternal knowledge in sustaining immunization coverage. Similar associations between maternal education, vaccination documentation, and service use have been reported in other sub-Saharan African settings [5]. However, the magnitude of the effect of card possession in this study (AOR 30.9, 95% CI 11.6‐82.0) is unusually high, reflecting the centrality of documentation in this context. Extremely high ORs and wide CIs should be interpreted with caution, as they likely reflect sparse data or near-perfect separation rather than precise effect sizes [12].

Implications for Programmatic Action

The discrepancy between administrative coverage (>100%) and survey-based findings (38.3% dropout) highlights systemic weaknesses in data quality [10,11]. Inflated administrative reports may delay corrective measures and misguide resource allocation. Programmatic priorities should include strengthening card management and documentation systems, enhancing maternal education and awareness of vaccination schedules, deploying mobile vaccination units to reach underserved households, improving communication strategies to address fears of side effects, and integrating dropout monitoring into routine supervision to ensure timely corrective action.

These measures are essential to reduce dropout and improve equity in immunization coverage in the DRC.

Limitations

This study has several limitations. First, the inclusion of children without vaccination cards, whose immunization status was determined through caregiver recall and health facility registers, may have introduced misclassification bias. Therefore, the results should be interpreted with caution, although this approach allowed for a more comprehensive assessment of dropout determinants. Second, the cross-sectional design precludes causal inference. Third, some variables lost statistical significance in multivariate analysis due to collinearity and small subgroup sizes, which limits precision. Fourth, the study was conducted in a single health zone, which may restrict generalizability to other settings. Despite these limitations, the sample size (N=300) was adequate to detect significant associations, and the findings provide valuable insights for programmatic action.

Conclusions

Vaccination dropout between BCG and measles remains high in the Kikula Health Zone, driven primarily by lack of documentation and maternal knowledge gaps. Addressing these barriers through improved card management, maternal education, and targeted outreach is critical to strengthening immunization coverage. The findings underscore the need for accurate data, integrated monitoring, and context-adapted interventions to reduce dropout and achieve equitable immunization in the DRC.