Objective: This systematic review synthesises international evidence on behavioural and programmatic factors leading to the success of child and youth school retention and completion in low -and middle-income countries (LMICs) to inform policy and social and behaviour change (SBC) interventions.

Data sources: The study follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines with the proposal registered on the International Prospective Register of Systematic Reviews (PROSPERO) database.

Search strategy: A systematic search was performed on PubMed, ERIC, Web of Science, ScienceDirect, and Google Scholar databases for articles published from Jan 2015 to July 2024. Study selection: Reviewers independently completed screening procedures and reviewed eligible publications focusing on LMIC policy frameworks, cultural, gender, socioeconomic, infrastructural, human resource and geographic factors impacting school drop-out and SBC interventions.

Data synthesis: SBC was rarely mentioned in intervention approaches, despite SBC activities being identified in programs targeting out of school children (OOSC), including training and capacity building, identification of key influencers, and the development of purposive SBC interventions. Facilitators to improved engagement and student retention included teacher training and capacity building, the use of structured syllabus, and targeting of participants from the education supply and demand sides. Cash transfer programs generally demonstrated positive impacts on student retention and reduced drop-out, including nutrition incentives in and out of schools.

Conclusion: Complementary multipronged and multilevel SBC approaches that create greater engagement and partnerships with parents, teachers and the communities they serve, are a likely to achieve the greatest traction in addressing challenges of OOSC.

PROSPERO registration ID: (anonymized for review)

• Out-of-school children • School dropout • School retention • Social and behavior change communication • Low-and middle-income countries

Tahir T, Shafiq K (2025) Social and Behaviour Change Interventions to Reduce School Drop-Out and Improve School Retention and Completion among School-Aged Children and Youth in Low -and Middle-Income Countries: A Systematic Review. J Behav 8(1): 1026.

Children dropping out of school before completion of primary or secondary studies is a problem for educationalists in many countries, but more so in low -and middle-income country (LMIC) contexts. UNESCO reports the number of out-of-school children (OOSC), has risen by 6 million since 2021 and now totals 250 million dropouts or non-attendance globally [1]. This accounts for 16% of children and youth in primary to upper secondary level not attending school, with 1 in 10 children worldwide not being in school at primary level with 122 million, or 48% of the out-of-school population being girls and young women [1]. Sub-Saharan Africa accounts for close to 30% of all OOSC globally with 1 in 5 African children not attending school (19.7%) and only around 50% of children attending upper secondary school [1,2]. Some of the key drivers of OOSC identified by UNICEF include household poverty, long distances to school, lack of tailored inclusive services for vulnerable children, including those with disabilities, and deeply rooted social norms that undervalue education. Furthermore, the lack of gender-sensitive/responsive school WASH (Water, Sanitation and Hygiene) facilities and menstrual and hygiene services have also been found to exacerbate school dropout, especially for girls [3]. Coupled with these challenges the COVID-19 pandemic has also reversed the gains made despite efforts made by government and development partners to establish and provide access to alternative learning pathways [4 ]. UNICEF continues to support African countries toward the improvement of educational and nutrition outcomes for children with projects in Tanzania, and other African countries. In Tanzania, the nation’s emphasis, like many other sub-Saharan African countries focusses on achieving educational goals, ranging from universal primary school enrollment to ensuring free and equal access at all levels. This commitment aligns with Sustainable Development Goal 4 (SDG 4), and Tanzania’s pledge to offer free, equitable and quality primary and secondary education to all children by 2030 through the Education Sector Development Plan [5 ]. Despite the implementation of the Fee-Free Basic Education (FFBE) policy in 2016, which significantly increased net enrollment rates, challenges persist. While primary school enrollment rose from 84% in 2016 to 95.3% in 2020, 3.6 million children of school-going-age remain out of school, with 37.5% having never attended. Dropout rates – particularly between standard 6 and 7 – pose a concern, with reported reasons including a perceived ‘lack of interest’ in continuing education, and costs such as school uniforms, learning materials, food, etc. remaining challenges to the very poor [2,6]. Other identified barriers include violence against adolescent girls [7], while education itself has also been identified as a protective factor against violence [8 ,9]. Implementation of FFBE has been seen to have a positive impact on enrollment rates, especially among vulnerable groups, thereby reducing the financial burden on families [10], However, challenges persist, including overcrowded classrooms, increased teacher workloads, and a decline in parental involvement, reflecting communication gaps and misunderstandings about the policy [7]. Attitudinal or perceptual determinants-particularly loss of interestsignificantly contributes to school dropouts, with studies indicating a greater impact on boys influenced by economic activities; family structure; peer pressure; early partner pregnancies; and early marriages [11,12]. Against this backdrop, the Government of Tanzania with support of UNICEF (Tanzania), and their partners, are looking to integrate a comprehensive, evidencebased Social and Behaviour Change (SBC) strategy to educational interventions, to address the issue of OOSC in Tanzania. SBC works to provide scientific insights to give communities more control over the decisions that affect their lives. This is best achieved through the recognition that changing knowledge isn’t enough to change behaviours with the approach focussing partnerships with families and community leaders on understanding their needs, identifying their strengths and lower barriers to positive change in the environments in which they act, making it easier for communities to adopt protective practices for their children [13]. Systematic reviews more often focus on creating a taxonomy of a body of research to advance a thesis and provide a foundation for future investigations [14]. However, systematic reviews can also provide valuable tools, coupled with a strong evidence base to inform the design and operationalising of educational and other health interventions that are in urgent need in LMICs. This systematic review explores the current literature to inform SBC interventions to enhance enrollment and retention rates with children, ultimately contributing to the overarching goal of achieving universal basic education in Tanzania and other LMICs, with a focus of the review being on developments since FFBE was implemented in 2016.

Systematic Review question(s) To explore PI(E)COS criteria (Populations (P), Interventions/Exposures (I/E), Comparators (C),Outcomes (O), and Study designs or Settings (S) as follows:

RQ1: What SBC interventions have been the most successful in in LMICs to reduce child and youth school drop-out, and improve retention and completion?

RQ2: What are the determinants associated with the enrolment and retention of OOSC in primary and lower secondary school?

RQ3: How do social/ cultural norms, attitudes, perceptions, and social structures impact on school participation among primary and lower secondary schoolaged children?

RQ4: How effective are the existing policies and interventions in addressing the challenges faced by OOSC?

RQ5: What are the key challenges and opportunities associated with alternative education models for OOSC?

RQ6. Of the programs demonstrating positive outcomes how did their effectiveness vary by the type and scale ofthe interventions, and student sociodemographic and cultural backgrounds?

This systematic review was performed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The proposal was registered on the International Prospective Register of Systematic Reviews (PROSPERO) database before conducting the review. (Review ID: CRD42024549293).

Search Strategy

A systematic search was performed extensively on PubMed, Education Abstracts, ERIC, PsycINFO, Web of Science, EMBASE, ScienceDirect, and Google Scholar databases for articles published from Jan 2015 to July 2024. Articles published from 2015-2024 were considered given that 2015 was the year when many African countries commenced implementation of fee-free basic education (FFBE), as part of the “leaving no one behind” campaigns. In addition to the identified databases, reference lists of recent and relevant systematic reviews/meta-analyses and references cited in the relevant original articles were also screened for studies not found in the main search strategy.

The final search was conducted by combining individual search results and combining them with appropriate Boolean operators (“OR” and “AND”). Search results were imported into EndNote® version X7 software (Thomson Reuters, New York, USA) in EndNote library file (.enl) format and any duplicate studies were removed. URL to search strategy: Key words included: Social Behaviour Change AND School Drop-out OR Out-of-school children AND Basic Education AND Low -and Middle-Income Countries OR Developing Countries (see Appendix 1. Short Version of SRMA Search Terms/ Detailed Version of Search Terms).

Selection Protocol

Two reviewers (anonymized) conducted the search protocol independently with predefined search terms and strategies from databases of the published peerreviewed literature for potential inclusion in the review. ### meticulously searched for the literature while ### repeated the searches and both authors approved all studies that were included in the final synthesis. A third investigator ### settled any differences between the two investigators during the entire selection process. All review process.

Inclusion/Exclusion Criteria

Eligibility criteria were formulated as per PI(E)COS criteria (Populations, Interventions/Exposures, Comparators, Outcomes, and Study designs or Settings. The study population (P) will be deemed school-going children, and those dropped out of school children/youth residing in low-and middle-income countries. Interventions/ Exposures (I/E) will focus on social and behaviour change (SBC) interventions implemented within and outside of the school settings targeting the populations of interest, the scale of the interventions and the reach to program target groups. Comparators/Controls (C) will explore the different types of intervention approaches including interpersonal, community, social media and mass media approaches and randomized controlled trials (RCTs) which identify comparisons of changes with control groups.

Outcomes (O) will examine factors such as school retention rates, dropout from school, change in dropout rate, reenrolments and year of completion including changes in SBC indicators of program awareness, knowledge, attitude and perceptions, intentions and behavioural changes resulting from program awareness. Study designs or settings (S) will include experimental designs, inclusive of RCTs, cluster-controlled trials along with descriptive observational studies measuring program outcomes based on a pre-delivered interventions or programs. Settings for inclusion will include any studies or interventions conducted in LMIC contexts. Additionally, only peerreviewed journal articles, written in English, were to be included in the review. Studies published only from 2015-2024 will be included considering comprehensive SRMAs conducted examining school drop-out and retention interventions prior to this period, and the introduction of free universal primary education from 2015 in East African countries and other LMICs. Interventions based on educational programs in high income country settings will be excluded considering the different challenges faced by resource constrained LMIC contexts. Commentaries, editorials, letters to editors, correspondence, books, conference proceedings, and review studies will also be excluded.

Risk of Bias Assessment

In this systematic review, the risk of bias of the included studies was assessed using the Risk of Bias (ROB-2) tool for Cluster Randomized Controlled Trials [15]. This tool evaluates key domains for potential biases, randomization process, timing of identification and recruitment of individual participants in relation to timing of randomization, deviations from intended intervention, missing outcome data, measurement of the outcome and selection of reported results [14]. Each domain was judged as having a low, some concerns, high risk of bias, or unclear risk of bias. For studies which used observational design, the Risk of Bias in Non-Randomized Studies – of exposure (ROBINS-E) tool was used [16], ROBINS-E focusses on seven domains in terms of bias assessment including – bias due to confounding, measurement of exposure, selection of participants into the study or analysis, post-exposure interventions, missing data, measurement of outcome and selection of the reported results [17].

Two reviewers (##, ###) independently performed the risk of bias assessment for all included studies. Any discrepancies between the reviewers were resolved through discussion or by consulting a third reviewer (##) when necessary. The results of the risk of bias assessment are presented in Figures 1,2, which provide an overview of the risk of bias across studies for each domain.

Search Overview

The extensive database search yielded 3,573 publications. Twelve other studies were identified from searching other websites and citations. After removing 804 duplicate records and another 397 records marked as ineligible by automation tools or not retrieved, 2,372 titles and abstracts were reviewed, out of which 1,984 papers were irrelevant, 199 had no abstracts available, 120 were review articles, 16 were books, 10 were book sections/chapters, 19 were conference papers, 02 were generic articles, 01 full text couldn’t be retrieved and were thus excluded. The remaining 21 full texts were reviewed based on inclusion/exclusion criteria. Of these, 17 papers were retained for full text review. Another 15 articles were identified through a reference search of relevant articles and reviews; of these 4 full text articles were found eligible for inclusion. Therefore, a total of twenty-one (N=21) studies met the study eligibility criteria and were included in the final sample. The PRISMA flowchart in Figure 1. provides the details of inclusion and exclusion of studies.

Figure 1 PRISMA flowchart identifying and screening studies, eligible and selected studies (n=21).

Study characteristics

Of the 21 studies from January 2015-June 2024 selected for the final review, studies from 11 countries globally were identified. Most of the studies (n=10) were from the African region – Kenya (n=3): (Phillips-Howard P, Nyothach E, Kuile FO, et al.; 2016; KangwanaI B, Austrian K, Soler-Hampejsek E, et al.; 2022; Zulaika G, Bulbarelli M, Nyothach E, et al.; 2022); Malawi (n=2): (Kilburn K, Handa S, Angeles G, et al.; 2017; Hunsaker S.; 2018); Sierra Leone (n=2): (Mason M, Galloway D.; 2020; Smith WC.; 2022). Somalia (n=1): (Kipchumba E, Sulaiman M.;2017); Zambia (n=2): (Sampa M, Jacobs C, Musonda P.; 2018; Seidenfeld D, Prencipe L, Handa S.; 2015). Other study locations included India (n=4): (Prakash R, Beattie TS, Javalkar P, et al.; 2019; Amora YB, Dowden J, Borh KJ, et al.; 2020; Nedungadi P, Mulki K, Raman R.; 2016; Kookana RS, Maheshwari B, Dillon P, et al.; 2016); Latin America (n=3) Brazil: (Santos FM, Corseuil CHL.; 2022), Argentina/Brazil/Mexico: Cerutti P, Crivellaro E, Reyes G. et al.; 2019; Mexico: (Dustan A.; 2019); Philippines (1): (Grepon BG, Cepada CM.; 2021); Turkey (n=2): (Karaman 0.; 2022; Aygün AH, K?rdar MG, Koyuncu M, et al.; 2024); and Indonesia (n=1): (Setyadharma A.; 2018).

Study characteristics varied, with a half (n=10) of the studies focussing on evaluating data from high-quality RCTs with robust sample sizes for population level assessments of findings including, quasi-experimental and experimental designs or mixed methods approaches. Evaluation designs, longitudinal studies and time-series analysis accounted for the other half of the of studies reviewed (n=10). Five (n=5) of the total 20 studies, focussed on evaluation of conditional and unconditional cash transfers on student retention and drop-out. A summary of the final 21 studies selected, according to their type, and the calibre of the studies, is provided in Table 1.

Table 1: Summary of key characteristics of selected SBC interventions studies and outcomes addressing OOSC according to study type and author’s alphabetical listing. (n=21).

Meta-analysis assessment

Of the 21 studies included in the review, those that investigated the impact of cash transfers on various outcomes and or contained several parameters of interest relating to school completion/dropout were identified as having the best opportunities for meta-analyses. These nine (n=9) studies are listed as #4, 6, 7, 8, 13, 14, 15, 16, 21 in Table 2. However, following further review, a meta-analysis was deemed inadequate due to significant heterogeneity in study designs, reported outcomes, and data availability. Specifically, one study lacked estimable effect sizes, and two studies reported only regression coefficients without raw data. Additionally, two studies examined attendance within intervention sessions, diverging from the review’s focus. While three studies reported some relevant outcomes, this limited sample size (of estimates available in these studies) will render pooling of the data statistically which makes them insufficient for a meaningful metaanalysis.

Table 2: Evaluation Designs/Longitudinal Studies/Time-Series Analysis and Mixed Methods (n=10 studies)

Summary of key findings

Of note from the search of electronic databases was the fact that social and behaviour change (SBC) was rarely if ever mentioned in intervention approaches, despite a number of SBC activities being highlighted in programs targeting OOSC including, training and capacity building, identification of key influences of teachers, parents and peers, the development of purposive training interventions, structured syllabus, and the targeting of a number of participants from the education supply and demand sides. (Mason M, Galloway D.; 2020). Additionally, SBC messages to achieve desired program outcomes, and support policy initiatives to incentivise or “Nudge”[17], vulnerable groups toward greater engagement and retention in schools were also highlighted (Santos FM, Corseuil CHL; 2022; Nedungadi P, Mulki K, Raman R.; 2016).

Facilitators or enablers toward improved engagement and student retention in schools included a number of approaches. Multipronged and multisectoral approaches were highlighted as having greater opportunities for positive change over singularly pronged initiatives. These included other SBC activities such as teacher training and capacity building, and the use of Interventions that involved parents as well as students (Grepon BG, Cepada CM.; 2021). Other incentives to improve student retention for vulnerable groups of pregnant and lactating mothers included recommendations for day-care support in schools where CCTPs may not have significant impact to retention on some cohorts (Santos FM, Corseuil CHL.; 2022). Furthermore, implementation of Girl Friendly School (GFS) policies and programs were also identified as reducing the levels of absenteeism and school drop-out amongst highly vulnerable girls, although the authors noted more studies on approaches and efficacy of these approaches may be needed (Kipchumba E, Sulaiman M.; 2017).

Other gender focussed incentives to increase retention and reduce drop-out included the availability of sanitary products (Phillips-Howard P, Nyothach E, Kuile FO, et al.; 2016) although, adequate training was recommended and needed to be provided with the availability of a range of products. Other perspectives from the demand side identified that capacity building of parents through engagement and training was also found to provide positive results, as well as actively including parents and teachers in schools to address absenteeism and drop-out, in particular with problem children (Karaman 0.; 2022).

Unconditional and Conditional Cash Transfer programs (UCTPs/CCTPs) featured prominently with studies, with eight (n=8) UCTPs/CCTP studies identified which matched the search criteria, with findings found to have generally demonstrated positive impacts on student retention and reduced drop-outs (KangwanaI B, Austrian K, SolerHampejsek E, et al.; 2022; Setyadharma A.; 2018; Kilburn K, Handa S, Angeles G, et al.; 2017), and this includes retention of vulnerable groups (Sampa M, Jacobs C, Musonda P.; 2018) and very young children in households receiving the benefits (Seidenfeld D, Prencipe L, Handa S.; 2015). However, the impact of cash transfers on student retention may be limited in hard-to-reach vulnerable groups residing in remote rural areas (Prakash R, Beattie TS, Javalkar P, et al.; 2019).

Additionally, income shocks were found to be a significant predictor of school continuity (Cerutti P, Crivellaro E, Reyes G. et al. (2019) suggesting that CCTPs may positively impact on school retention. However, the findings on CCTPs/UCTPs are also mixed with indications that liquidity constraints may not be key driver of high school dropout in some urban settings (Dustan A.; 2019) while other studies identify socio-cultural factors including suspicions and negative perceptions of schemes that are perceived as equitable (Banda E, Svanemyr J, Sandøy IF, et al.; 2019). A number of barriers to school engagement and student retention were identified through the review. Environmental factors such as sanitation and water scarcity in dry regions have been identified as contributing to school absenteeism with girls affected the most (Kookana RS, Maheshwari B, Dillon P, et al.; 2016), while poverty, geographic locations, including distances to schools, teacher quality, the lack of material resources and emotional support, physical and sexual abuse, early pregnancy and poor perceptions of the benefits of schooling were also cited as barriers to reducing OOSC rates in a number of studies (Mason M, Galloway D.; 2020; Sandøy IF, et al.; 2019).

Prospects were also identified for existing young adult school dropouts to renew their study through selfdirected home learning (Musita R, Ogange BO, Lugendo D.; 2019) although practical approaches to re-entry of previous dropouts to secondary schools may need to be considered through specifically designed low-cost courses. Scholarship programs were also found to have positive impacts on school retention and completion especially when they cover boarding school fees (Hunsaker S.; 2018). However, scholarships may also foster resentment from other students or face considerable resource constraints for replication at scale. A number of other approaches were also identified from the studies selected in supporting longer term school retention of primary and secondary school students. The use of digital technologies and mobile apps were found to enhance the learning process for teachers and students and reduce levels of drop-out or improve re-entry of students through novel approaches. (Nedungadi P, Mulki K, Raman R.; 2016; Musita R, Ogange BO, Lugendo D.; 2018).

Learnings from the recent COVID-19 pandemic has also led to recommendations that long-term considerations should include sustainable planning that looks beyond the reopening of educational institutions including comprehensive financial support packages for groups most likely to be affected by pandemic threats and other school disruptions as these can have a negative impact on educational outcomes, with those impacts being highest among most vulnerable youth (Smith WC.; 2022; Zulaika G, Bulbarelli M, Nyothach E, et al.; 2022).

Risk of bias assessment

Figure 1 provides assessment of potential biases in multiple Cluster RCTs including in this systematic review. The assessment of bias in the RCTs cluster included in this systematic review revealed varying levels of concern across different domains. Several studies, such as those by Prakash et al., Kilburn et al., and Seidenfeld et al., demonstrated a low risk of bias across most domains, indicating a robust methodological approach. In contrast, studies like Karaman et al. and Amora et al. were identified with a high overall risk of bias, primarily due to concerns in critical areas such as randomization and blinding procedures. Other studies, including Philips-Howard et al., Mason et al., Kangwana et al., and Sampa et al., exhibited some concerns in multiple domains, suggesting potential limitations in study design that may influence the reliability of their findings. Overall, the risk of bias assessment indicates a heterogeneous quality of studies, with a mix of low-risk and high-risk biases (Figure 2).

Figure 2 Risk of Bias Assessment for Article Domains D1-D5.

The assessment of bias in the observational studies of exposure included in this systematic review highlights a spectrum of methodological quality across the studies (Figure 2). Several studies, such as Grepon et al., Cerruti et al., Santos et al., Smith et al., and Hunasker et al., were identified as having a high overall risk of bias. These studies exhibited consistent concerns across multiple domains, particularly in randomization (D1), measurement of outcomes (D4), and overall assessment (D7), which may affect the reliability of their findings. Conversely, studies like Zulaika et al., and Aygun et al., demonstrated a low overall risk of bias, with minimal concerns noted in most domains, suggesting a more robust methodological approach. However, a number of studies, including Dustan A., Setyadharma et al., Nedungadi et al., Kipchumba et al., and Kookana et al., were found to have some concerns overall, indicating potential limitations that could introduce bias into their findings (Figure 3).

Figure 3 Risk of Bias Assessment for Article Domains D1-D7

The synthesis of findings from this systematic review identified that SBC is currently not part of the lexicon for educational policy making or interventions. This highlights the importance of greater integration of SBC approaches within the educational sector to address OOSC challenges in LMIC contexts. As such, there is a need to strengthen the discussion on the theoretical underpinnings of SBC and how it could be better integrated into educational interventions which to date, are not evident in the international literature, despite SBC demonstrating successes in other child health and welfare issues [18,19]. The review findings also point to the fact that the greatest benefits may be achieved through SBC targeted interventions to high-risk groups including low-income, rural populations, girls, and previous school drop-outs, with these findings reflecting previous research in the area [1,2].

Furthermore, the synthesis of findings indicates that solutions to the problems of OOSC are complex and will not be achieved through singularly focussed programs and activities, but rather through the adoption of more strategic approaches. In the first instance, this may require a more comprehensive understanding of the scope of the problems as well as greater understanding of the current programs, practices, and resources targeted to addressing the problems. Also identified was the importance of developing community-driven plans to address school dropout including interventions that accept community resource opportunities and limitations. Last, evident from the large scale RCTs and longitudinal studies, was the need to continue with ongoing assessment, including monitoring, evaluation, learning and adaptation (MELA) for SBC activities to ensure that community initiatives have the desired behavioural impact. Paying greater attention to warning signs of dropout problems, getting a deeper understanding of which students are most likely to drop out, and identifying the early-warning signs to alert teachers, school staff, and parents to the need for intervention were factors highlighted for consideration prior to the development of community led interventions.

Of the findings in relation to the eight (n=8) UCTP/ CCTP studies, it is apparent, given the weak to moderate evidence of success, that more strategic implementation and promotion of CTPs should be considered. Despite the mixed findings of this review, metanalyses conducted of CCTPs, have identified that they have demonstrated strong support for all schooling outcomes in impact, transfereffectiveness, and cost-effectiveness estimates for LMIC contexts at a global level [20]. This may incorporate a more family/student-centred funding model, given the understanding that it will cost more to educate vulnerable students, including those living in poverty, those living with disabilities, or facing rural disadvantage or gender inequalities. This also recognises the need for a cash transfer funding model in areas of need, to be adjusted to a funding allocation based on the socio-demographics of individual students, their schools and communities, which more closely aligns funding to specific student needs. CCTPs which included food supplementation within school settings was a feature of school programs in Southern Asia, although, it was not evident in the African region.

School meals programs have been found to “NUDGE” students to attend school as well as providing incentives for parents to send their children to school [21,22]. Although, initially funded by international donors such as the World Food Programme (WFP), transitioning toward full Government ownership and implementation has provided these programs with a level of sustainability. The programs can have 3-fold benefits in disadvantaged rural areas including increased attendance, retention and primary school completion, improved consumption of nutrient-dense foods with high-risk children; and increased market participation of smallholder farmers through the provision of quality and diversified products [16,17]. Regulations and policy, including policies for Universal Primary Education (1977) and Fee-Free Basic Education (FFBE; 2016) can also have significant impact on school engagement and retention of vulnerable students. Although, this issue was not included in the literature review inclusion criteria, with policy approaches and appropriate laws being reviewed in the African context and seen as important to maintain school attendance and manage learner absenteeism [23]. Stigma and discrimination emanating from student low socio-economic status was also highlighted in numerous studies in the review.

However, the impact of free uniform programs on OOSC rates was not evident. Other studies have also emphasised a note of caution in considering the negative impact of approaches such as uniform programs designed to build retention and improve school performance with some students [24], with findings that they may also lead to increased drop-out with higher-risk groups if the programs are not implemented in an equitable manner, with barriers to engagement also reported in other studies with pregnant mothers (Santos FM, Corseuil CHL.; 2022). A number of review interventions focussed on capacity building of both education supply and demand-side participants, acknowledging the importance of cultivating relationships with teachers, trusted adults and peers, and the iterative nature of approaches [25], through advisory groups —small groups of students that come together with teachers to create an in-school family of sorts. These advisories, which meet during the school day, may provide a structured way of building skills and enabling those supporting relationships to grow and thrive.

Enhancing family/friend relationships has also been seen as a factor which may assist in providing a more protective environment, particularly for girl students, to reduce violence by teachers [8,9]. From the supply side of teachers and school administrators, findings indicate that a range of capacity building approaches may be required to ensure learning is relevant to the needs of students, as boredom and disengagement have been identified in the review as some of the reasons students stop attending class and drop-out of school, with these issues also evident in other literature on student engagement and retention [26]. From the demand side of problem students, personalising learning opportunities to create small learning communities, for groups with special needs was another strategy to reduce the dropout rate. In this regard, the review identified that the demands of working and/or family responsibilities make it untenable for high-risk students to attend school during traditional schedules with similar challenges to identifying flexible learning opportunities for working students [27]. This may require more flexible approaches to the school syllabus through the development of sandwich programs identified in the review as having mixed benefits as well as weekend or evening classes, to provide at-risk student flexibility in scheduling.

For vulnerable girls’, retention approaches could also include the greater provision of water, sanitation and hygiene (WASH) personal hygiene products and facilities as well as the eligibility or provision of subsidies for childcare, to allow girls and their partners to continue with their education [28]. The findings on the prospects of improved rigour within schools also indicate that raising the academic bar may reduce dropout rates through improved teacher, student engagement while supporting students desire to achieve higher levels of success, and also prepare them to graduate with improved options. This also reinforces the need to target interventions to teachers to build their skills and confidence in providing optimal teaching opportunities as well as reinforcing strict adherence of ethical conduct.

SBC Intervention outcomes

Systematic review findings against the six research questions are as follows:

RQ1: What SBC interventions have been the most successful in in low -and middle-income countries to reduce child and youth school drop-out, and improve retention and completion?

The evidence indicates that multipronged SBC approaches can be the most successful with activities that include capacity building and training of supply side actors—teachers, principals and school administrators— and demand side actors—students, parents, families, peers and other community influencers, combined with advocacy for policy initiatives to create educational environments which are “more conducive to student retention” through progressive and flexible syllabus, safe and stimulating learning environments with adequate WASH resources. A focus on this range of activities, may be the most effective combination of interventions to elicit social and behavioural change toward improved student uptake, retention and future educational outcomes. However, despite the evidence on the efficacy of a combination of interventions adopting both “global and targeted strategies” impact on OOSC in developing countries, scoping reviews recommend further research to be conducted in different countries to confirm the effectiveness of the range of interventions, given the differing cultural contexts [29].

RQ2: What are the determinants associated with the enrolment and retention of OOSC in primary and lower secondary school?

The predominant range of social and behavioural determinants identified from the review included socioeconomic, geographic and behavioural factors that influenced enrolment and retention of children in schools in LMICs. Income disparities which are higher in rural and remote rural communities require children to generate income for their families, rather than participate in school. Resource constraints including WASH resources are also a significant determinant with low-income groups and have a higher impact on girls than boys. From a behavioural perspective, poor attitudes and negative social norms of low-income rural parents and children, in relation to the quality of schooling or the longer-term benefits of education persist, particularly with higher-risk groups. In line with WASH initiatives, girl friendly schools (GFS) are seen as a possible tool to increase girl’s school attendance The studies confirm earlier studies in Tanzania and the East Africa region on the challenges to retention with specific at-risk groups [30].

RQ3: How do social/ cultural norms, attitudes, perceptions, and social structures impact school participation among primary and lower secondary schoolaged children?

Negative social norms toward education are often established because of the growing inequality and dispossession of vulnerable groups and are prominent in many rural and urban informal settlements in LMICs. Education is not yet established as a positive social norm in many communities and hence inadequate sanctions are attached to it. Feelings of hopelessness and helplessness may pervade in a number of communities with negative social norms toward education, with a variety of studies identifying lower levels of “perceived behavioural control” —poor attitudes and self-efficacy perceptions to affect positive change—with participants and community groups. As a result, greater engagement and support through SBC is required to build skills and confidence of community members with the understanding that early education of children is a basic human right that offers many longer-term benefits for families. Additionally, improving knowledge about sexual and reproductive health may reduce the high level of pregnancies identified in areas of disadvantage and improve retention rates of girl students [11].

RQ4: How effective are the existing policies and interventions in addressing the challenges faced by OOSC?

A range of policies and interventions following fee-free-basic-education (FFBE) have contributed to improving school attendance and completion. One of the primary areas of intervention to address school retention and engagement apparent in the review were evaluations of UCTPs and CCTPs with the programs generally demonstrating some improvements in school attendance and engagement. However, non-significant findings in differences between intervention and control groups also indicate that more culturally nuanced approaches to UCTPs/CCTPs may need to be piloted and carefully evaluated given the lack of achievement in intended results identified in some studies. This includes CTPs that provide NUDGES through nutrition components such as “home-based lunch provision”, that contrary to program objectives, contributed to increased absenteeism, lateness, transfers and attrition, thus reducing child retention capacity of schools [31].

Conversely, the inclusion of school food supplementation, as a component of CCTs may be more successful than UCTPs delivered at household level. Findings from other studies also indicate that CCTPs that include nutrition requirements may improve student biometric measures while improving child school attendance [32]. Given that income shocks were found to be a significant predictor of school continuity and retention, CTPs for families with children at high-risk should be considered through targeted interventions. Alternatively, CCTPs for families with school age children should include a condition that the benefits will only be provided following confirmation of regular attendance to schools, given that school retention is the objective.

RQ5: What are the key challenges and opportunities associated with alternative education models for OOSC?

The key challenges from a LMIC context relate to human and financial resource constraints to implement multipronged SBC approaches and trialling of alternative educational models and the considerable scale of the interventions required, given the considerable challenges. What is evident is that a number of alternative educational models identified from the international literature such as ODeL secondary school re-entry programs (Kenya), Girl Friendly Schools (Somalia), Continuing Professional Development and Learning (CPDL) capacity building approaches (Sierra Leone), Accountability Theory approaches to support orphaned children, and Sandwich Programs (South Africa), Designed Adjustment Programs (DAP) for Children in problem preschool children (Turkey), and the greater adoption of mobile technology in educational settings (India), need to be considered and co-designed with the communities facing the greatest challenges. Opportunities arise with more strategic planning and formative research in the early stages of SBC program development to support culturally relevant, alternative educational models that create greater engagement of local communities in the development of these models. This will ensure a buy-in and ownership of the education interventions to ensure sustainability in the longer term.

RQ6. Of the programs demonstrating positive outcomes how did their effectiveness vary by the type and scale of the interventions, and student sociodemographic and cultural backgrounds?

Learnings from the interventions and studies assessed through the review as well as evidence from earlier studies [33], identify the need for more systematic and strategic SBC approaches to build knowledge, change attitudes, perceptions and social norms toward the benefits of school learning environments for students and their families. The variable impact of some programs including cash transfers, educational models and programs to re-engage dropouts indicates that socio-economic, demographic and cultural factors, implementation approaches, the scale of the interventions and other factors have played a part in the success or otherwise of the interventions. As such, SBC can play a more significant role in developing more culturally nuanced, educational approaches, including the piloting, evaluation and scaling-up of successful approaches in the African context.

Limitations

The heterogeneity of the outcome measures, the relatively small number of empirical studies, and insufficient data reporting only specific educational outcomes did not allow for a quantitative synthesis for this review in the form of a meta-analysis. All experimental studies had some concerns in the risk of bias assessment due to the inherent nature of the SBC interventions embedded within broader educational or economic cash transfer programs to support OOSC, which may have affected the research findings. Nonetheless, these studies scored higher in other domains to minimize any potential high risk of bias. Some studies were of shorter duration, and as such, the limited period of the interventions may have led to insufficient findings in educational outcome assessments of children. Studies were primarily conducted in the African, South-Asian and South American regions and as such, they may be generalized to these regions. Furthermore, the differences in findings between multicountry settings highlights the fact that local contextual and cultural factors might determine the outcomes, and thus the tailoring of SBC approaches, tools and activities should be considered. Additionally, the value, conditionality, and distribution patterns of cash transfers in the identified studies were variable, constituting a potential source of bias in evaluating outcomes.

they serve are also likely to achieve the greatest traction over time in addressing educational challenges of OOSC. Practical recommendations include the implementation of comprehensive SBC strategies that involve schools, families, and communities, and engaging teachers, parents, community leaders, and students in capacity-building efforts and collaborative initiatives to create a more conducive learning environment, especially for vulnerable groups.

Next, strategic implementation of CCTs should also be considered through linking CCTs to regular school attendance as combining cash transfers with at-school nutritional support has been shown to be effective in retaining students, particularly among low-income and rural populations. Last, strengthening the policy implementation framework is crucial to effectively address foundational education challenges.

Policies must not only promote access but also embed SBC approaches to tackle social norms, attitudes and behaviours that contribute to school dropout and absenteeism. By recognising the role of SBC in influencing positive educational outcomes, policymakers can create more inclusive, adaptive, and communitydriven interventions.

All data relevant to the study are included in the article or uploaded as supplementary information.

1. UNESCO. 250 million children out-of-school: What you need to know about UNESCO’s latest education data. 2023.
2. UNESCO. New estimation confirms out-of-school population is growing in sub-Saharan Africa: Factsheet. 2022; 48.
3. Rugimbana CE, Mwansa P. Mwila School Dropout among Girls: Reasons and Measures in Community Secondary Schools in Temeke Municipality, Tanzania. Asian J Educat Soc Stu. 2023; 43: 47-61.
4. Shayo EH, Mubyazi G, Barongo V. Experiences of teachers and students on school closures and its consequences during the COVID-19 pandemic in Nyarugusu refugee camp, Tanzania. PLOS ONE. 2024.
5. Ministry of Education Science and Technology. Education Sector Development Plan (ESDP) 2021/22 2025/26. 2022.
6. UNICEF Office of Research (INNOCENTI). Food Systems for Childrenand Adolescents Working Together to Secure Nutritious Diets. 2018.
7. Nkuba M, Hermenau K, Goessmann K, Hecker T. Reducing violence by teachers using the preventative intervention Interaction Competencies with Children for Teachers (ICC-T): A cluster randomized controlled trial at public secondary schools in Tanzania. PLoS ONE. 2018; 13: e0201362.
8. McEvoy D, Brannigan R, Cooke L. Risk and protective factors forself-harm in adolescents and young adults: An umbrella review of systematic reviews. J Psychiatr Res. 2023; 168: 353-380.
9. Evans DK, Hares S. Adolescent Girls’ Safety In and Out of School: Evidence on Physical and Sexual Violence from across Sub-Saharan Africa. Center for Global Development, Working Paper. 2020.
10. Shukia D. Fee Free Basic Education: Worth Rethinking. Huria J. 2020;27.
11.  Moshi FV, Tilisho O. The magnitude of teenage pregnancy and its associated factors among teenagers in Dodoma Tanzania: a community-based analytical cross-sectional study. Reprod Health. 2023; 20: 28.
12. Ouma DH, Ting Z, Pesha JC. Analysis of the Socio-Economic Factors That Contribute to Children School Dropout in Artisanal Small-Scale Gold Mining Communities of Tanzania. J Education Practice. 2017; 8: 71-78.
13. UNICEF. Social and behaviour change. 2024.
14. Patricia A. Alexander Methodological Guidance Paper: The Art and Science of Quality Systematic Reviews. Rev Educat Res. 2020; 90: 6-23.
15. Sterne JAC, Savovi? J, Page M, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMI. 2019; 366: l4898.
16.   Julian PT, Higgins RL, Morgan A. A tool to assess risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E). Environ Int. 2024; 186: 108602.
17.   Thaler R. Nudge: Improving Decisions about Health, Wealth, and Happiness. USA: Yale University Press. 2008; 45: 700-701.
18. Mahumud RA, Uprety S, Wali N. The effectiveness of interventions on nutrition social behaviour change communication in improving child nutritional status within the first 1000 days: Evidence from a systematic review and meta-analysis. Matern Child Nutr. 2022; 18: e13286.
19.    Briscoe C, Aboud F. Behaviour change communication targeting four health behaviours in developing countries: A review of change techniques. Soc Sci Med. 2012; 75: 612-621.
20. García S, Saavedra JE. Educational Impacts and Cost-Effectiveness of Conditional Cash Transfer Programs in Developing Countries: A Meta-Analysis. Rev Edu Res. 2017: 87: 921-965.
21. Hossain MS. School Meals and Child Outcomes in Bangladesh. BUFT J Busines Econ. 2021; 2: 274-287.
22. Shair W, Hussain S, Idrees M. Social Safety Net Programs and FoodInsecurity in Pakistan. 2023.
23. Coetzee S, Venter R. South African law and policy regulating learner absenteeism at public schools: Supporting an ecosystemic management approach. South Afr J Educat. 2016; 36: 1-11.
24. Sabic-El-Rayessa A, Mansurb NM, Batkhuyagc B. School uniform policy’s adverse impact on equity and access to schooling. Compare. 2019; 50: 1-18.
25. Griffiths AJ, Alsip J, Hart SR. Together We Can Do So Much: A Systematic Review and Conceptual Framework of Collaboration in Schools. Can J School Psychol. 2021; 36: 59-85.
26. Macklem GL. Boredom in the Classroom: Addressing StudentMotivation, Self-Regulation, and Engagement in Learning. 2015; 88.
27. Carnevale AP, Smith N. Balancing Work and Learning: Implications for low-income students. Georgetown University Center on Education and the Workforce. Washington, USA. 2018.
28. Black SE, Devereux PJ, Løken KV, Salvanes KG. Care or Cash? The Effect of Child Care Subsidies on Student Performance. Review Econ Statist. 2014; 96: 824-837.
29. Eslamian A, Arshi M, Tahmasebi S, Alipour F, Rafiey H. Scoping Review on Interventions, Actions, and Policies Affecting Return to School and Preventing School Dropout in Primary School. Int J Prev Med. 2023; 14: 92.
30. Ng’ondi NB. Socio-demographic and service provision characteristics associated with primary school attendance among the Most Vulnerable Children in Tanzania, Children and Youth Services Review. 2012; 34: 2255-2262.
31. Abonyo N, Onderi H, Ayodo TMO. The effects of home-based lunch provision on child retention in primary schools in Mbita Sub-county Kenya. J Edu Res Behav Sci. 2014; 3: 330-342.
32. Andersen CT, Reynolds SA, Behrman JR, Crookston BT, Dearden KA, Escobal J, et al. Participation in the Juntos Conditional Cash Transfer Program in Peru Is Associated with Changes in Child Anthropometric Status but Not Language Development or School Achievement. J Nutr. 2015; 145; 2396-2405.
33. Smink J, Schargel FP. Helping Students Graduate: A StrategicApproach to Dropout Prevention. Eye on Education; NY. 2004; 49.