In this study, we investigated the factors associated with timely infant immunization in 105 facilities in Kinshasa Province using a cohort of HIV-positive pregnant women and their infants. We observed similar immunization coverage for BCG (82.6%) and the first dose of HepB3 (68.4%) compared to the WHO/UNICEF estimates of national immunization coverage averaged from 2017-20 (84.5% and 71.3%, respectively) (20).
We employed the Andersen BHM to categorize factors that impacted vaccine uptake within three categories: predisposing characteristics (infant’s birth order, mother’s age, mother’s year of educational attainment, and her marital status), enabling factors (household wealth index and form of transportation to the facility), and external environment (general readiness, immunization readiness, type, ownership, and urban status of the facility where the mother-infant pair sought care). We observed a strong positive association of factors across all three categories with timely uptake of both vaccines, BCG and HepB3. Our analyses highlight the importance of considering the effect of determinants at different levels and time intervals. The implication is that policymakers in the DRC can be more focused on implementing vaccine uptake strategies depending on their target group. For example, we identified significant determinants of timely uptake of both BCG and HepB3 at the external environmental level, focusing on intervenable facility characteristics.
Higher general readiness scores and attending a religiously affiliated facility were independently associated with both BCG and HepB3 timely uptake, a finding that aligns with previous studies highlighting vaccine storage and stockout challenges in the DRC (13, 21). In prior work, lower general facility readiness scores have been reported as barriers to timely immunization (22). Therefore, new vaccine introduction strategies at the facility level should prioritize contributions to general readiness, such as workforce and operations, and immunization readiness, such as reliable availability of vaccines at each facility. Ensuring adequate supply prevents stockouts, which in turn prevents unnecessary/unproductive visits to a facility. This could also bring more equity as it reduces costs for families who live further away from the facility.
In addition, visits to public facilities were negatively and independently associated with timely uptake of BCG compared with visits to religiously-affiliated facilities. In contrast, visits to private, non-religious facilities were positively associated with timely uptake of both BCG and HepB3 than religious facilities. Previous studies on immunization uptake in SSA have cited a facility’s religious affiliation driving a mother’s choice to seek care there (13). In terms of private facilities, studies show that mothers visiting private facilities tend to be of higher income levels (23), a factor that was positively associated with timely vaccine uptake and may explain our findings. Future research could examine the reasons that vaccination rates are higher among those who select private facilities.
Within the category of predisposing characteristics, we found that the mother's educational attainment and age at the time of enrollment were positively associated with the timely uptake of vaccines, which indicated that knowledge, awareness, and experience with preventative care were vital for timely vaccine uptake. A mother's education level is an established predictor of infant immunization in low- and middle-income countries (15, 24–26) and education is a valuable solution to overcome challenges to vaccine uptake (27, 28). Mothers primarily receive healthcare information from two sources: health workers during ANC visits, and their families and communities. Previous studies have shown that knowledge about HBV risk and the vaccines’ protection is low among Congolese individuals (13, 29), with one study finding a basic knowledge of HBV among only 33.2% of healthcare workers(29) and another finding that only 31.2% (87/280) of pregnant knew how HBV was transmitted (30). To increase a mother’s knowledge, initiatives to disseminate vaccine knowledge could be targeted to the mother’s two primary sources of information, the facilities during ANC visits and their communities. In terms of the facility, training that includes vaccine information needs to be geared towards not only vaccine staff, but also the ANC staff who are disseminating information to pregnant women.(13) At the community level, approaches to address the knowledge barrier could be targeted towards the mother’s complete social network of family and friends (29).
Household wealth status was positively associated with timely BCG vaccination, consistent with other studies (15, 31, 32). A possible explanation for this observation is the cost of the vaccinations. Although vaccinations are technically free in the DRC, facilities often require vaccination fees for a vaccine card and well-baby consultation. In addition, indirect costs such as transportation and income loss may act as economic burdens obstructing vaccine uptake (15, 33). The economic burden was further substantiated by our finding that mothers were more likely to vaccinate their infants on time, or ever, if they lived within walking distance to the facility. This finding was confirmed by earlier studies in SSA that distance to facility, travel time, and need for transport were negatively associated with immunization uptake (15, 25); one such study found that traveling a distance of over 30 minutes by foot compared to a shorter distance reduced vaccine uptake by one-third (25). These implications for the enabling factors require policy intervention to reduce the economic burden of infant vaccines, such as transparency and standardization of vaccine costs across facilities and incentives for mothers living beyond walking distance from facilities.
A significant strength of this study was the aggregation of facility-level and individual-level longitudinal data across many facilities. Few studies have looked at a combination of the individual-level determinants of the mother-infant pair and the facility-level determinants of vaccine uptake. The study employs a unique approach to controlling for confounding by using data from both the supply (environment) and demand (mother-infant pair) side. This study's access to longitudinal panel data of over 2,000 women and inventory data about each of the study facilities allowed us to evaluate a comprehensive list of determinants across the BHM levels that determine timely uptake of vaccines. Beyond the unique challenge of administering the vaccine within 24 hours of birth, these determinants highlight the need for an implementation strategy to be rolled out alongside universal HepB. Previous studies demonstrate that timely uptake remains low in countries that have previously adopted HepB-BD because there is no clear guidance to overcome individual- and facility-level challenges (34, 35). Our study's main policy implication was to highlight the barriers to current BD vaccines – and HepB3 vaccine – in a context that strives to include the HepB-BD vaccine in its national immunization schedule. Policymakers may use these findings as evidence when developing a future implementation strategy streamlining all three BD vaccines – HepB-BD, BCG, and OPV0 – within the first 24 hours of life. Findings from this study can help national, sub-national, and facility-level stakeholders to strengthen the uptake of both BD vaccines and other available vaccines for infants across the DRC and SSA.
Despite the study's strengths, it was not without limitations. Our assessment leveraged sample participants from a cohort of women already enrolled in an HIV continuous quality improvement study, which impacted the generalizability of this study. However, vaccine uptake in this population was similar to the national average, and factors influencing vaccine uptake were aligned with other infant immunization studies in the Congo and elsewhere in SSA (15, 25–28, 31, 32, 36). The measure of the outcome variables, vaccine uptake status, and timing, also may have suffered from recall bias because of the contemporaneous approach of capturing vaccine dates during study interviews. The study staff reviewed infants' vaccine records to correct any errors in logging the vaccine dates to help alleviate any errors. Another limitation of this study is that we could not infer causal estimates and instead presented associations. In addition, only mothers recruited pre-delivery (approximately half of the sample) responded to a question regarding how many ANC visits they had attended. Finally, while the parent study did experience significant rates of LTFU, study staff followed up with respondents to understand reasons for LTFU and recapture some of the data otherwise lost. We were able to weigh LTFU within the 'never vaccinated' rate to represent more accurate rates of failure to vaccinate.