Using qualitative systems mapping to analyze the linkages between the behavioral and social (BeSD) determinants of routine childhood immunization in LMICs Authors: Lalitha Vadrevu, Shradha S Parsekar, Monica Jain, Gunjan Taneja, Shekhar Menon Published: 2024 ABSTRACT AbstractBackgroundDesigning effective immunization programs requires a strong understanding of how the factors affecting vaccine uptake interplay. In this review, we analyze the relationships between the Behavioral and Social Drivers (BeSD) of routine immunization using qualitative systems mapping (QSM).MethodsIn this review, we analyzed 92 experimental and quasi-experimental impact evaluations (IEs) from 11 LMICs that were published between 2010 and 2020. Secondary literature on the study context or the intervention was also included. The WHO’s behavioral and social determinants (BeSD) framework was used to code the determinants identified in these IEs, and their relationships were mapped using qualitative systems mapping. We computed the in-degree (influenced by other determinants) and out-degree scores (influencing other determinants) to assess the extent of the influence of the BeSD determinants on one another.ResultsThe results identified that knowledge regarding immunization, trust in the health system and quality of immunization services, and community engagement by the health workers was influenced by several other determinants and had a high in-degree score. Caregivers perceptions of quality of immunization services, health provider availability, religious leaders, community engagement by health workers, and physical accessibility had a high out-degree score. We also identified two feedback loops between health provider availability and physical accessibility, and trust and perceptions of immunization quality.ConclusionsQSM analysis shows that the determinants of immunization uptake were interlinked with each other in complex ways. Our research identified BeSD drivers that affected multiple factors and can be viewed as key leverage points. Programs for improving vaccination uptake need to account for the strong role that caregivers’ experience of immunization services and their trust in the health system play in indirectly affecting immunization uptake. There is also a need to acknowledge that fear of vaccination is fear of the opportunity and monetary costs associated with vaccine side-effects. The results from this review can inform discussion and form the basis for context specific research on the factors affecting immunization.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12889-024-20850-w.Keywords: Routine immunization, Qualitative systems mapping, LMIC, Causal loop diagrams, QSM, CLD, Qualitative review, Behavioral and social determinants of vaccination, BeSD Introduction Despite improvements in the coverage of routine immunization services in LMICs, there are still large gaps in immunization uptake both within and across countries [1]. Recent estimates show that in 2023, 14.5 million infants did not receive an initial dose of DTP vaccine and an additional 6.5 million are partially vaccinated. Of the total 21 million, around 60% of these children belonged to 10 LMIC countries in Africa and South-East Asia [2]. There has been extensive research on the factors affecting the uptake of routine childhood immunization in the last few decades. Studies have shown that socioeconomic factors such as poor education levels and lower income affect caregivers’ ability to seek vaccination [3, 4]. Studies have also identified lack of awareness, religious beliefs, lack of trust in the health care provider, weak health systems, social context, lack of individual autonomy, and physical or financial inaccessibility as determinants of vaccination [5–9]. While underlining the influence of these factors on immunization uptake, researchers also widely agree that how these determinants interact with each other and collectively influence a caregiver’s decision to vaccinate their children is complex, time and context specific, and varies in directionality [7, 10]. For example, contrary to the common understanding that access to health providers is weaker in rural areas, in a study in China, caregivers in rural areas were found to have stronger relationships with health care providers than in urban areas that led to improved vaccination coverage [4]. According to a study exploring the determinants of DPT3 vaccination in Nigeria, increasing maternal age was not associated with better knowledge regarding vaccination because younger mothers had fewer children and fewer responsibilities, allowing them more contact with the health system than older mothers [11]. Maternal employment, which improved knowledge and access to vaccination services by increasing agency and financial independence for women in some cases, led to poorer familial support and greater opportunity costs for seeking vaccination and loss of wage days in others [12, 13]. There are various measures for identifying vaccination determinants, yet, understanding the complex relationship between the determinants of immunization has only recently gained attention [14, 15]. Systems thinking – based on the principles of complex adaptive systems –- is a tool that is useful in understanding the relationships between different actors/attributes, their interactions, and the processes through which they operate. It also helps in identifying feedback loops to identify leverage points and areas for intervention. Qualitative systems mapping (QSM) using causal loop diagrams (CLDs) is a popular tool in systems analysis [16, 17]. CLDs help visualize a central aspect of complex systems i.e., relationships [18]. CLDs depict the dynamics of the system and its individual actors using directional arrows and loops. The basic structural mechanism in CLDs consists of feedback loops of mutual causality known as causal loops [19]. CLDs have been used for comprehensively mapping health system actors/attributes and visualizing the relationships between them in the context of several public health issues, such as tobacco control [20], the COVID-19 pandemic [21], neonatal mortality [22], and health system strengthening [23]. There is extensive evidence on the determinants of vaccination from LMICs, as mentioned above, yet few studies exist that comprehensively map the linkages between these determinants to provide a macro picture of the key relationships between them. Current research on immunization using qualitative systems mapping focuses on individual country contexts and specific service delivery attributes. Studies using CLDs in immunization were conducted in Uganda [24], India [25], Malawi [26] Afghanistan [27], except for two studies from the LMIC context, in general, [28] and Africa [29]. These two studies also did not rely on any existing frameworks of the determinants of immunization except for the study in Afghanistan [27]. Specific attributes such as governance [25], health service responsiveness [26], impact of COVID-19 on routine immunization [29], immunization readiness [27] were analyzed in these studies. Through this research we aim to fill this gap in evidence by analyzing and mapping determinants of vaccination using the Behavioral and Social Drivers (BeSD) of Vaccination Framework in 11 countries in LMIC context. We analyzed the relationship between determinants affecting immunization using CLDs. Methods Criteria for including studies for this review We included experimental and quasi-experimental impact evaluations (IEs) identified from the 3ie’s Evidence Gap Map (EGM) and the review by the Yale Institute for Global Health [30, 31](Refer to Fig. 1 for details). The search for both these reviews was conducted in May and March of 2020, respectively. We selected impact evaluations from 11 countries to identify and analyze the factors contributing to the lack of vaccine uptake for routine child immunization. We selected countries using the following criteria: (a) the proportion of zero-dose children is large or has gone up and/or routine immunization rates are stagnating, declining or low. We used the data from WHO/UNICEF national immunization coverage estimates, 2021 to shortlist countries(Refer to supplementary File 1); (b) availability of a sizeable evidence base (at least 3–4 studies) for a given country. (c) Additionally, one country from each of the three regions (South and East Asia region, African region, Latin American region) with a higher routine child immunization coverage rate relative to the LMIC average was also selected. For Latin America region we had to select two because of low number of studies available. Study search and screening We selected IEs published between 2010 and 2020 from the following countries: India, Indonesia, Bangladesh, and Pakistan from the South and East Asia region; Ethiopia, Rwanda, and Nigeria from the African region; and Brazil, Nicaragua, Colombia, and Guatemala from the Latin American region. The IEs that were selected included any intervention that impacted routine immunization coverage of children aged 0–5 years. We also included interventions impacting intermediate outcomes through causal mechanisms, e.g., caregiver knowledge of childhood immunization. We identified supporting documents to the IEs by including any relevant publication pertaining to the study area discussing the intervention in question and being conducted in or referring to the year in which the intervention was implemented. We used clinical trial registries and Google Scholar to search for related studies. In addition, we also carried out targeted searches of the websites of implementing agencies and funders. We included the following types of supporting studies: feasibility studies, qualitative studies, formative or process evaluations, cross-sectional or observational studies, project documents or funder reports, and policy briefs. A total of 92 IEs were identified based on the above criteria along with 93 supplementary articles (refer to Supplementary file 2). Coding We did a deductive thematic analysis using the BeSD framework [32, 33]. Existing measures on the determinants to vaccination are mostly based on evidence from high-income countries and rely heavily on attitudinal factors without considering social processes and practical issues. To address these shortcomings, the WHO’s behavioral and social determinants (BeSD) of vaccination framework was developed based on discussions with key stakeholders from both high- and low-income countries with the aim of supporting the measurement of a wider range of drivers of vaccination [34]. The BeSD framework focuses on the beliefs and experiences of caregivers and includes themes that have traditionally not been measured comprehensively and have differed in quality, validity, conceptual clarity, and ease of making geographical comparisons. The framework includes four themes: what caregivers think and feel, social processes, motivation, and practical issues pertaining to caregivers’ access and knowledge of services as well as their delivery [32]. We used NVivo software (version 1.6.1) for qualitative synthesis. The coding framework consisted of the three main BeSD thematic areas and their respective subcategories that capture the determinants of childhood vaccination uptake from the caregivers’ perspective: (1) thinking and feeling – lack of knowledge regarding the importance or need for vaccination, fear of vaccination, lack of trust in the health care provider; (2) social processes – household norms, influence of religious leaders, neighbors and friends, providers’ engagement with the caregiver, and public mechanisms to encourage immunization; and (3) practical constraints – distance, cost, lack of awareness of vaccination schedule or place and time of vaccination, opportunity cost, and inability to bring the vaccination card. This includes caregiver experience with health care quality, which was further sub-coded into systemic challenges such as stockouts, provider availability, poor technical quality, poor provider behavior, long wait times, and refusal to open vials for immunization. We also coded the underlying health system factors, wherever available, from the healthcare providers’ perspectives, such as payment delays, lack of training, poor infrastructure, and healthcare workers’ attitudes about vaccination. While motivation is the fourth determinant under BeSD, we have not explicitly coded this factor as attitudinal, social, or practical constraints drive the motivation levels of caregivers – only the factors contributing to motivation were coded. We followed an iterative process to achieve coding consistency between the coders. Two researchers coded the studies included in this QSM analysis independently using NVivo software. We developed a comparison table containing the researchers’ coding to identify the differences in coding and to standardize the operational definitions of the codes used. This process continued until there were no major differences in the coding performed by the researchers. Systems analysis – causal loop diagrams While CLDs can be mapped using both primary and secondary data, we used secondary data i.e., published impact evaluations and their supporting documents, to map the linkages between BeSD barriers [18]. We coded and analyzed the relationships between different determinants by developing co-occurrence tables and relationship charts in NVivo. We mapped a determinant as co-occurring with another determinant if it was cited in the same study context using coding matrices. We coded a determinant as related to another using relationship charts if it was explicitly identified in the background/context, from the study data—specifically qualitative or process evaluation data—or through the author’s conclusions. Both quantitative and qualitative information was included. The identified relationships were mapped into a causal loop diagram (CLDs) using Vensim software. We calculated the in-degree or the number of factors influencing a certain variable and the out-degree or the number of factors influenced by that variable in the system. We identified factors with a high out-degree score as being key leverage points for improving vaccination uptake. Determinants affecting 3 or more BeSD drivers were identified as drivers that could be used as leverage points (a determinant can affect one factor on the minimum). The causal loop diagrams are depicted with factors affecting immunization in circles connected by arrows showing the directionality of the relationships. The red arrows with a ‘-‘ sign indicates an inverse relationship where an increase in one factor has a negative influence on the related factor, while the green arrows with a ‘+’ sign indicates a direct relationship where an increase in one factor has a positive influence on the following factor and a decrease will also lead to a decrease. We used dotted lines to depict relationships with variables or uncertain directions. Feedback loops that fed into each other were identified as reinforcing loops and marked as ‘(R)’ in the figures. These loops between variables are self-enforcing and show that an increase in one variable tends to increase another variable which in turn can increase the first variable. Results We mapped the BeSD determinants that influenced and were influenced by other determinants in the LMIC context using their in-degree and out-degree scores (Refer to Table 1 for details). The results show that determinants pertaining to knowledge, fear of vaccination, religious and personal beliefs, and trust in the health system were influenced by many determinants and were associated with the uptake of services. Restrictive household norms, opportunity cost, and caregivers’ experience of quality of services were also affected by several determinants and affected uptake. Studies also showed either quantitatively or qualitatively that lack of knowledge regarding vaccination, lack of trust, fear of vaccination, opportunity cost, and caregivers’ experience on immunization services was associated with whether they seek vaccination for their children. Table 1In-degree and out-degrees of the determinants of vaccination uptake in LMICsBeSD DeterminantIn-degree (factors influencing a variable)Out-degree (factors influenced by a variable)Thinking and FeelingLack of knowledge regarding immunization50Fear of vaccination and its side-effects21Restrictive religious or personal beliefs21Trust in the Health System54Social ProcessesInfluence of traditional or religious leaders05Neighbors and Friends Support11Knowledge of family members on immunization02Restrictive household norms and decision making32Community engagement by the health workers44Information and education to mothers on Immunization11Social groups and public mechanisms to encourage vaccination12Practical ConstraintsAwareness of place and time of vaccination30Awareness of vaccination schedule30Cost of transportation10Opportunity Cost31Physical accessibility26Inability to produce vaccination card01Caregiver experience – Quality of Immunization services93Caregiver experience – Healthcare provider availability24Caregiver experience – Stockouts11Caregiver experience – Healthcare Infrastructure01Provider experience – Training and Supervisions01Provider experience – Payment delays01Contextual FactorsPolitical Stability04Migration02Socio-economic background01Misinformation14 In-degree and out-degrees of the determinants of vaccination uptake in LMICs Determinants with a relatively higher out-degree scores (out-degree score of 3 or more) i.e., influenced many other determinants were, trust in the health system, caregivers’ experience of quality of care, influence of traditional and religious leaders, physical accessibility, health worker engagement with the community, and health provider availability. The details of the studies showing these relationships have been provided in the supplementary materials. (refer to Supplementary File 3). We divided the results section into attitudinal and social determinants to include determinants under the thinking and feeling and social processes categories in BeSD, and practical determinants of vaccination to include practical constraints. We described the various factors affecting these determinants in addition to the key determinants that affect multiple determinants of vaccination. Attitudinal factors and social processes affecting immunization Awareness of the need for or importance of immunization, caregivers’ trust in the health system, restrictive household norms, fear of vaccination and restrictive religious or personal beliefs were found to affect vaccination uptake and were influenced by many other determinants. Awareness of the need for or importance of immunization A lack of knowledge regarding the need for or importance of immunization is a critical challenge and was associated with immunization uptake. Social groups and public mechanisms that encourage vaccination, the extent of knowledge of family members, neighbors’ and friends’ support for vaccination, caregivers’ education levels, and the extent of counseling that they receive from health workers regarding the need to immunize their children, affected their understanding of the importance of immunization (See Fig. 1). Social groups and societal mechanisms Caregivers’ social networks of neighbors and friends influenced their awareness levels, as these social groups act as information sources for caregivers. Knowledge and acceptance of immunization among friends and neighbors, including on social media, improved the knowledge levels. We found that social groups increased the interaction of women with other community members through group meetings and systems of collective decision making, in addition to providing encouragement to seek vaccination. Social groups such as self-help groups, women’s assemblies, village clubs, and community health committees provide avenues for sharing experience and knowledge. These groups were also capable of working with and influencing the levels of engagement of community health workers by improving links and interactions with the community. Knowledge of family members and friends regarding immunization We also found that the knowledge of immediate family members, such as in-laws and older children, can positively influence caregivers’ knowledge of immunization and its health benefits. Mothers-in-law were found to have a strong role, as they also influence maternal and child health decision-making. Health worker counselling on immunization We found that the extent of interaction between community health workers and caregivers positively influenced the extent of education and counseling those caregivers received. This included the frequency of home visits or the extent of engagement during the home visits. Education levels Poor education and literacy levels were found to be negatively associated with knowledge regarding immunization. It has also been shown that poor knowledge levels are influenced by fear of vaccination. Caregivers trust in the health system We found that caregivers’ trust in health system was influenced by three factors: healthcare provider availability, caregivers’ experience of immunization services, misinformation prevalent in the community and migration. Trust in the system influenced the healthcare-seeking behaviors of the caregivers. Healthcare provider availability Frequent change and unavailability of healthcare providers created a sense of uncertainty that prevented caregivers from taking their children for subsequent vaccination. This finding underlines the importance of familiarity with providers and stable strong relationships with them to foster trust in the health services. In rural and remote regions where the retention of nurses and health workers continues to be a persistent challenge as caregivers faced the cost of visiting the health center multiple times for vaccination. Misinformation and influence of traditional religious leaders Resistance from religious leaders in the community for immunization services due to religious beliefs or misconceptions, negatively influenced caregivers’ trust in vaccination. However, religious leaders in another context facilitated immunization uptake by increasing awareness on immunization days and its benefits. Misinformation that immunization services are the conspiracy of foreign governments and NGOs and that they can cause sterility were also some of the factors that influence communities’ trust in immunization and health providers. Caregivers’ experience with immunization services We found that caregivers’ experiences regarding the skill and quality with which immunization services were provided affected their trustworthiness. Caregivers who felt that the frontline worker was vaccinating the child ‘harshly’ attributed the side effects of immunization to it. Caregivers also equated boils or rashes after vaccination with poor skills of the providers. They were found to report issues such as poor prescription of drugs, poor communication and infrequent visits by health workers which dissuaded them from taking their children for vaccination. A lack of trust among caregivers also discouraged healthcare workers from providing vaccinations. In addition to the above determinants, migration was another factor that was associated with poor trust levels in the migrant community owing to the lack of community and health system linkages. Household norms, fear of vaccination, and religious and personal beliefs We found that decision-making systems at the household level determine whether the mother takes the child to vaccination and was linked with fears of vaccination and religious or personal beliefs. Household norms and decisions were influenced by the social networks of friends and neighbors and by prevalent misinformation in the community (See Fig. 1). Fear of vaccination is affected by the social price of immunization and opportunity and monetary costs Husbands and mothers-in-law are key decision makers in the household, and review shows that even if caregivers wanted to immunize their children, they did not take them for immunization if their husbands disapproved of it. The price for immunizing the child without the approval of the husband was found to be very steep in some cases. Studies have shown that women expressed strong emotional responses to being asked to immunize their child because of the fear of an altercation with their husbands. Additionally, some expressed concerns that taking the child for immunization could even lead to their husbands divorcing them. In addition, female caregivers as well as their husbands were worried about the opportunity costs of losing a days worth of wages or of not fulfilling household duties if their child had side effects. Fear of vaccination was also influenced by monetary costs of treatment of side-effects of vaccination, particularly by the husbands. Household norms also prevented health workers from engaging with caregivers and their families. Healthcare workers were prevented from counseling other male members of the household regarding the benefits of immunization, as it was not appropriate for men of the household to engage with women healthcare workers. This had a trickledown effect, as the women’s decisions were determined by their husbands’. The caregivers were also fearful of vaccination if the religious leaders did not support it or if they suspected that it was harmful. Personal or religious beliefs are influenced by social norms, religious leaders, and prevalent misinformation Restrictive social norms prevented caregivers from taking their children for immunization. Studies show that the perception of vaccines were harmful or cultural beliefs that women should not step out of the house before forty days after childbirth prevented caregivers from taking their children for immunization. These negative beliefs were exacerbated by the views held by religious leaders in the community. Physical accessibility The analysis showed that physical accessibility in terms of access to all-weather roads and migration affected community engagement by the health workers. Inadequate access to all-weather roads and distance of the households from the health center were found to potentially undermine the extent of health workers’ engagement with the community and, by extension, and awareness generation activities. Populations that migrated from other regions were found to have poor engagement with health care providers relative to others and were further socially and geographically marginalized due to a lack of vaccination cards and social networks. It was also found that the extent of interactions of caregivers with other community members depended on their access to all-weather roads, which is indicative of the role that physical mobility plays in accessing social networks. Attitudinal and social drivers for immunization affecting many BeSD determinants Our analysis showed that Influence of religious leaders, community engagement by health workers, and caregivers’ trust on the health system affected multiple determinants of immunization. Practical and service delivery constraints to immunization The QSM of the determinants identified under practical constraints and health system challenges showed that caregivers’ experience of quality immunization services, awareness of place, time and schedule of vaccination, and opportunity cost were important determinants of vaccination and was associated with vaccination uptake (see Fig. 2). Two reinforcing feedback loops emerged from the review -- one between experience of quality immunization services and trust, and the second between physical accessibility and provider availability. Qualitative systems mapping-linkages between attitudinal and social barriers Caregivers’ experience with immunization services We found that while it seems intuitive to assume that an improvement in caregivers’ experience of quality immunization services will lead to an increase in the number of women taking their children for immunization, both extremely high and extremely low demand for vaccination can affect caregiver experience of quality. For example, included studies have shown that frontline workers and nurses cancel immunization camps and outreach activities when very few women are present. They also noted that if the catchment area for the facility was large and if there were many women in proportion to the existing health providers, planning outreach activities, immunization camps, and conducting home visits was a challenge, leaving many hard-to-reach areas untouched. This also led to the mother being asked to come the next day. This, in turn, had a negative effect on whether caregivers experienced quality immunization services (See Fig. 3). Qualitative systems mapping-linkages between practical constraints and service delivery issues In this analysis, we found that caregivers identified effective communication, frequent visitation, polite behavior, and skillful administration of vaccines that do not cause boils or rashes as markers of quality. Caregivers’ experience with immunization services was found to be a byproduct of nine factors. We found it to be positively influenced by provider engagement with the caregiver through home visits, provider training, supportive supervision, human resource availability, political stability, physical accessibility, trust in the health provider, and the caregiver’s ability to produce vaccination cards. Payment delays faced by health providers and stockouts negatively affect performance, which in turn affects caregivers’ perceptions of quality. In addition, caregivers’ perceptions of immunization quality influenced their awareness of the vaccination schedule, the place and duration of vaccination, and their trust in the healthcare system. Health system factors We found that poor infrastructure, payment delays, inadequate training, lack of supportive supervision, stockouts and human resource unavailability influenced whether providers conducted immunization campaigns and effectively covered all eligible households during home visits in hard-to-reach areas. Immunization requires at least four to five visits by the caregiver to the immunization center, and absenteeism was found to affect the caregiver’s perception of quality. Payment delays and lack of supportive supervision for many health workers acted as a serious deterrents that affects their motivation levels, forcing them to perform multiple jobs or seek alternate employment. This affected providers’ engagement with caregivers and household members. Poor infrastructure, such as a dysfunctional cold chain, impaired their ability to conduct camps. The experience of quality vaccination services was also dependent on caregivers’ understanding of the roles of the health providers and workers. This knowledge provided clear expectations regarding the responsibilities of health workers and nurses implementing immunization services. A lack of clarity on how their duties differed created misplaced expectations that negatively affected caregivers’ experience of quality. Other programs and political stability We found that political stability and other concurrent programs had the potential to affect quality. For example, political instability had a serious effect on the payment and remuneration of healthcare providers, as did the recruitment of new providers. This also resulted in halting of immunization services. Political turmoil prevented health workers from conducting home visits that are crucial for generating awareness and providing reminders of upcoming vaccination appointments. Supplementary immunization activities such as the polio campaign had both positive and negative effects on immunization. They were found to lead to the diversion of funds and human resources, leaving fewer providers providing routine immunization services in some contexts while also facilitating the uptake of routine immunization services in others. Physical accessibility The ability of the caregivers to geographically reach the health center of the immunization camp was a strong determinant of perceived quality. Caregivers did not hold positive views of immunization services if they were far away. It is worth noting that physical accessibility and provider availability influenced one another and have a reinforcing feedback loop. Providers often did not cater to populations living in remote areas for home visits or camps, leading to absenteeism. Long distances to the health center meant increased costs of transportation or increased opportunity costs due to long travel times. Physical accessibility was also found to be correlated with caregivers’ ability to bring immunization cards to the center, which was crucial for determining what vaccines their child needs and, eventually, the caregivers’ perceptions of immunization quality. Trust that caregivers have on the provider had a reinforcing effect on their perceptions of quality which in turn increased the levels of trust. Awareness of the vaccination schedule and the place and duration of vaccination We found that caregivers’ experience with immunization quality, trust in the health system, and religious influence affected awareness of the place, time, and schedule of vaccination. Both the experience of immunization quality and trust influenced providers’ engagement with the community, which in turn determined caregivers’ awareness levels. Additionally, support from religious leaders has been found to strengthen systems of reminder and follow-up for vaccination. Practical drivers of immunization affecting many BeSD determinants We found that physical accessibility, caregivers experience of quality health services, and health provider availability affected many determinants of vaccination. Discussion Our results agree with existing QSM literature on the importance of trust in the health system, physical accessibility, mother’s educational status, role of familial and social support, husbands support, misinformation, and religious beliefs on vaccination uptake [24, 27, 29]. Additionally, our analysis finds that while some determinants like religious leaders negatively impact immunization uptake, they were and could be leveraged to facilitate access and improve awareness on immunization. Our research also identified drivers that can be viewed as key leverage points that need attention, for example, healthcare provider availability, perceived quality, trust in the health system, physical accessibility, influence of religious leaders, and community engagement by the health workers. These factors affect various other determinants and interventions that focus on these can have a trickle down affect on the entire system. We also found that the caregivers’ perception of immunization service delivery affected immunization uptake while also being influenced by many health system factors and affected many other drivers of vaccination like trust, awareness, and uptake. Perceived quality of immunization services was also identified as being influenced by vaccine service capacity and technical quality in a QSM study in Afghanistan [27]. Our analysis identified two feedback loops that reinforce each other – between healthcare provider availability and physical accessibility, and perceptions of immunization quality and trust. It is worth noting that these determinants of vaccination need to be considered together in planning an intervention given the mutual influence they have on each other. Strengths and weaknesses There are several strengths and limitations to the review. We used qualitative systems mapping to analyze published literature. It is among the few reviews attempting to analyze the relationships between determinants to vaccination using an established framework like the BeSD. By analyzing impact evaluations from 11 LMIC countries and visualizing relationships between vaccination determinants through CLDs we hope to provide leverage points for implementation. Our review serves as a template of the determinants and their relationships which can be used to inform implementation research done in various local contexts. Our review is an addition to the small but increasing volume of evidence on the relationship between immunization determinants using complex adaptive system tools from Uganda [24], India [25], Malawi [26] Afghanistan [27], Africa [29] and the LMIC context [28]. It also covers a large evidence base on immunization interventions from 11 LMIC countries in South and Southeast Asia, Latin America, and Africa and draws from the literature on the evaluation of immunization programs. While we have included secondary literature pertaining to these programs, the review does not include stand-alone descriptive studies of a quantitative or qualitative nature that are not related to the programs. It is also worth noting that certain multidimensional themes, such as trust and understanding of the importance of vaccination, were difficult to define [35]. Consequently, many themes pertaining to caregiver perceptions of health services and trust were found to coexist. The analysis also showed that social groups have the potential to influence and empower mothers to seek vaccination. However, the review does not address the details of the nature, type, strength, or weakness of these social networks or their implications for immunization uptake. In summary, we present the results of this review as a macro picture of the relationships between the factors prevalent in LMIC contexts. The results from this qualitative system mapping exercise have not yet been supplemented with the experiences of practitioners and implementers in the field of immunization. However, in the context of immunization, the results from this review can serve as a starting point for discussion with both the providers and the recipients of routine vaccination services to better understand the dynamics of the factors in their local context [17]. Research using causal loop diagrams has been used along with feedback from other stakeholders to develop tailored interventions due to the iterative and reflexive nature of the systems thinking approach [25]. CLDs have been used to develop a shared understanding among members of a community regarding the factors determining a certain health issue [20, 36]. For example, researchers in Lebanon used group model building workshops with 15 members in the community where they were asked to identify factors affecting mental health and develop a preliminary concept model describing the dynamics of mental health condition and experiences to ensuring wellbeing [37]. The CLDs from this review can be contextualized using feedback from local stakeholders and used for designing tailored interventions. The results of this review have the following implications for policy and practice. I)Social groups and community mechanisms can be leveraged to improve awareness and attitudes related to immunization: There is a plethora of evidence on the impact of community engagement and mobilization on improving immunization outcomes [38]. We found that community groups can serve as information beds for caregivers and family members regarding immunization. Community mobilization efforts that leverage existing social groups in the community to discuss health and immunization issues can have a significant impact on caregivers’ and family members’ awareness levels and fears about immunization. Social groups and community mechanisms can be leveraged to improve awareness and attitudes related to immunization: There is a plethora of evidence on the impact of community engagement and mobilization on improving immunization outcomes [38]. We found that community groups can serve as information beds for caregivers and family members regarding immunization. Community mobilization efforts that leverage existing social groups in the community to discuss health and immunization issues can have a significant impact on caregivers’ and family members’ awareness levels and fears about immunization. II)Setting expectations and communicating the role of providers in immunization can enable building trust: Trust in the health care system has been a widely researched topic [39]. Our analysis agrees with the literature that shows that trust is influenced by the caregiver’s experience of healthcare, underlining the importance of communication and engagement of the health system with the communities [28]. Behavior change communication is a crucial aspect of outreach in the community. While training and supportive supervision are important for ensuring that health workers have the knowledge and skill to educate and counsel caregivers regarding immunization, we also found that a lack of clarity regarding the role that community health workers play, as opposed to the nurses and doctors who immunize children. Poor engagement with the community due to excessive workload or inaccessibility can affect caregivers’ trust in the health provider. Ensuring that the roles of the key groups and members in the community are clarified can help in building trust. Setting expectations and communicating the role of providers in immunization can enable building trust: Trust in the health care system has been a widely researched topic [39]. Our analysis agrees with the literature that shows that trust is influenced by the caregiver’s experience of healthcare, underlining the importance of communication and engagement of the health system with the communities [28]. Behavior change communication is a crucial aspect of outreach in the community. While training and supportive supervision are important for ensuring that health workers have the knowledge and skill to educate and counsel caregivers regarding immunization, we also found that a lack of clarity regarding the role that community health workers play, as opposed to the nurses and doctors who immunize children. Poor engagement with the community due to excessive workload or inaccessibility can affect caregivers’ trust in the health provider. Ensuring that the roles of the key groups and members in the community are clarified can help in building trust. III) Strengthening health system factors holistically can affect immunization quality and build trust: Our review clearly shows that the caregiver experience of quality of immunization affects trust in the provider—a finding that resonates with other reviews on the subject [28]. Caregivers’ experience with immunization depends heavily on all the health system building blocks. Strengthening the health system can improve not only the quality but also the trust levels and linkages between the community and health workers. Given the concurrent influence of these factors on quality, there is a serious need to take a comprehensive approach that aims to simultaneously strengthen these aspects of the health system. Programs that focus only on building the capacity of health workers or only on building infrastructure and supplies run the risk of failure, as other factors can negatively affect the experience of caregivers. Strengthening health system factors holistically can affect immunization quality and build trust: Our review clearly shows that the caregiver experience of quality of immunization affects trust in the provider—a finding that resonates with other reviews on the subject [28]. Caregivers’ experience with immunization depends heavily on all the health system building blocks. Strengthening the health system can improve not only the quality but also the trust levels and linkages between the community and health workers. Given the concurrent influence of these factors on quality, there is a serious need to take a comprehensive approach that aims to simultaneously strengthen these aspects of the health system. Programs that focus only on building the capacity of health workers or only on building infrastructure and supplies run the risk of failure, as other factors can negatively affect the experience of caregivers. IV)Mitigating the indirect costs associated with immunization can improve its uptake: The benefits of immunization can fade in comparison to the high indirect costs from vaccination for some caregivers. While many programs exist that include outreach and counseling to other household members and husbands, our review shows that objections to vaccination largely stem from fears of the indirect costs that the family will incur. It is essential to anticipate the indirect costs to take the child for vaccination or for treatment of side effects. Identification and mitigation of these indirect costs through monetary incentives, transport services, or planning immunization sessions at a time that is most convenient to caregivers will be important in highlighting the benefits of vaccination against the indirect costs that they might incur. Mitigating the indirect costs associated with immunization can improve its uptake: The benefits of immunization can fade in comparison to the high indirect costs from vaccination for some caregivers. While many programs exist that include outreach and counseling to other household members and husbands, our review shows that objections to vaccination largely stem from fears of the indirect costs that the family will incur. It is essential to anticipate the indirect costs to take the child for vaccination or for treatment of side effects. Identification and mitigation of these indirect costs through monetary incentives, transport services, or planning immunization sessions at a time that is most convenient to caregivers will be important in highlighting the benefits of vaccination against the indirect costs that they might incur. V)Practitioners need to plan for potential high variability in demand for vaccination for ensuring maintenance of high-quality immunization services. There are extensive guidelines from national governments and international organizations on planning equitable and quality immunization services [40]. The decision on the number of immunization sessions is determined by the size of the population served by the health center and the human resources available. We find that the experience of quality health services positively drives the number of caregivers seeking immunization. However, large increases or decreases in the number of caregivers seeking immunization have the potential to disrupt services. Health workers are overburdened with work, leading them to ask caregivers to return for vaccination the next day, which negatively affects their perceptions of quality. On the other hand, too few women can result in the postponement of immunization camps, as it is not economical to open multidose vials for a few people. Program managers need to consider the variability in demand and understand its repercussions on perceptions of healthcare workers and quality while planning immunization services. Practitioners need to plan for potential high variability in demand for vaccination for ensuring maintenance of high-quality immunization services. There are extensive guidelines from national governments and international organizations on planning equitable and quality immunization services [40]. The decision on the number of immunization sessions is determined by the size of the population served by the health center and the human resources available. We find that the experience of quality health services positively drives the number of caregivers seeking immunization. However, large increases or decreases in the number of caregivers seeking immunization have the potential to disrupt services. Health workers are overburdened with work, leading them to ask caregivers to return for vaccination the next day, which negatively affects their perceptions of quality. On the other hand, too few women can result in the postponement of immunization camps, as it is not economical to open multidose vials for a few people. Program managers need to consider the variability in demand and understand its repercussions on perceptions of healthcare workers and quality while planning immunization services. Conclusion Our review using qualitative systems mapping underlines the complex interplay and influence that immunization determinants have on each other. Immunization programs in many LMIC countries rely heavily on grassroots-level community health workers who are tasked with providing services until the last mile. Community health workers need to leverage the social groups and networks within these communities to improve awareness and attitudes toward vaccination. In addition, comprehensive strategies that build strong health systems with robust funding, infrastructure, and supplies are critical for the delivery of quality health services, as are provider motivation and retention. We also found that opportunity costs drove fears regarding vaccination. Trust in the health system emerged as a latent factor that influenced awareness levels in the community. The results from this QSM highlight the interconnectedness of many of these determinants and point to the need for context specific, comprehensive programs to improve vaccination in LMIC contexts. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Acknowledgements We would like to thank the participants of the stakeholder engagement meeting in January 2023 for their significant feedback. We would like to thank Prof. Maren Duvendack, University of East Anglia, United Kingdom, for her input. We would also like to thank Sebastian Martinez at 3ie for his overall guidance and support of the review. We sincerely thank Avantika Bagai for her contributions during the initial stages of this review. Abbreviations Behavioral and social drivers Causal loop diagram Evidence gap map Impact evaluations Low- and middle-income countries Qualitative systems mapping Authors’ contributions LV and SP conducted initial qualitative coding of the studies included in the review. LV mapped the relationships and developed causal loop diagrams. MJ provided overall guidance during the analysis phase. The first draft of the manuscript was written by LV. SP, MJ, GT, and SM provided review comments and finalized the manuscript. Funding This review was supported by funding from the Bill and Melinda Gates Foundation (BMGF) through INV-030670. Data availability Data is provided within the manuscript or supplementary information files. The list of studies included in the review have been detailed. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Data Availability Statement Data is provided within the manuscript or supplementary information files. The list of studies included in the review have been detailed.