Cohort retention in seasonal malaria chemoprevention: an analysis of magnitude and associated factors in Northern Bahr el Ghazal, South Sudan

Introduction

In 2025, South Sudan had an estimated 5.5 million malaria cases, up from 3.8 million cases reported in 2024 (UNICEF, 2025). Malaria remains the leading cause of morbidity and mortality in the country, with efforts to control it hampered by ongoing conflict, widespread poverty, and malnutrition (Androga, 2020; World Health Organization, 2022). Malaria is endemic in all regions of the country, with prevalence peaking by the end of the rainy season from September to November (Severe Malaria Observatory, 2026). Seasonality varies by geography; in the north of the country, total rainfall is lower than in the south and tends to last between five and six months (Republic of South Sudan, 2018; Androga, 2020). South Sudan faces one of the highest risks from the climate crisis, already experiencing increased flooding and shifts in rainfall seasonality, which are expected to further exacerbate malaria (UNICEF, 2021).

Seasonal malaria chemoprevention (SMC) involves the intermittent administration of sulfadoxine-pyrimethamine and amodiaquine (SPAQ) to prevent malaria infections in eligible children during the peak malaria transmission season (World Health Organization, 2012; WHO, 2012; WHO, 2021). SMC is typically delivered in areas where malaria transmission is highly seasonal across sub-Saharan Africa (World Health Organization, 2025). SMC was first introduced in South Sudan in 2022, and is delivered in five cycles, spaced 28 days apart, by community health workers through door-to-door household visits (Kweku et al., 2009; World Health Organization, 2023). Cohort retention refers to eligible children completing the full three-day SPAQ course across all planned cycles of the annual SMC round. Completion of the full course across all cycles is essential for maximizing the protective effect and impact of the intervention (Harrington et al., 2011). In 2023 SMC campaign, up to 20% of children enrolled in SMC completed all the planned 4 cycles (Malaria Consortium, 2024). Evidence in another study emphasized the importance of medication adherence (Dwajani et al., 2020). Despite the proven efficacy of SMC in preventing malaria, retaining children from the first cycle through to the last cycle remains challenging (Ayamba et al., 2025). Evidence from other settings where SMC has been implemented shows that missed doses or dropout during any monthly cycle can substantially reduce the effectiveness of the intervention, hence increasing the prospect of malaria outbreaks (Nikiema et al., 2022). A study in Ghana reported increased dropout of children enrolled into SMC; 7.43% between cycle 1 to 2, 6.24% cycle 2 to 3, and 12.34% between cycle 3 to 4 (Ayamba et al., 2025). The reasons for dropout were categorized into patient related, drug related and health system related factors. Inadequate caregiver awareness, stock shortages, population migration, and fatigue among health workers are among the factors that can affect cohort retention in SMC campaigns (Li et al., 2024). Therefore, addressing the systemic and community-level barriers that hinder consistent participation across all SMC rounds is crucial for enhancing cohort retention and ensuring medication availability.

While there is substantial evidence on the effectiveness, acceptability, coverage, and adherence to the three-day SPAQ course, there is limited evidence on cohort retention across cycles and its drivers. The study, therefore, aimed to assess cohort retention rate across SMC cycles during the 2024 round and identify factors associated with retention among eligible children in Northern Bahr el Ghazal, South Sudan.

Methods

Study design

This study used data collected during a household survey conducted to assess SMC coverage and quality in Aweil West and South counties in Northern Bahr el Ghazal state, at end of the 2024 SMC campaign. These two counties were selected as SMC was implemented there, targeting children aged 3–59 months.

Study setting

Northern Bahr el Ghazal state is located in the Northwest of South Sudan and comprises five counties that include Aweil South, East, West, North and Center. The state has a total area of 30,543 square kilometers (Figures 1A, B). Aweil Town is the state capital of Northern Bahr el Ghazal, and most of the state’s population comprises Dinka and Jurchol tribe members, with a minority of Luo tribe members, whose main sources of livelihood are agriculture and livestock.

The target population was children who received SPAQ and their caregivers in the two counties of Aweil South and Aweil West. Aweil West County population is estimated in 2017 as 236,402, while Aweil South County is 104,385 (City Population, 2025). According to the National Bureau of Statistics of South Sudan, 19% of the population in South Sudan is under five years old. The study respondents were the caregivers of SMC-eligible children.

Sample size estimation and sampling procedure

A multistage sampling technique was used, as described in detail elsewhere (Malaria Consortium, 2025). The survey sample size was determined in accordance with Malaria Consortium guidelines to provide accurate and representative estimates of SMC coverage at the county level. A total of 1,500 households were surveyed across 75 clusters in the two countries (40 clusters in Aweil West and 35 in Aweil South), with probability proportional to size within each cluster. This clustered sampling method was selected to optimize logistics while ensuring statistical validity. The sample size was based on an assumed SMC coverage of 75–80%, a 95% confidence level, and a ±5% margin of error. An intra-cluster correlation coefficient of 0.2 was used to account for response similarity within clusters, yielding a design effect of 4.8. The size was increased by 5% to account for potential non-responses. The total population was 83,037. A sample of 1300 eligible children will be required to power the survey to estimate SMC coverage and other key indicators for eligible children (3–59 months), with a 95% confidence level and a margin of error of 5%. We used SurveyCTO to randomly select one child from the listed children per household to whom all survey questions were administered.

Dataset and data collection

The data used were collected during the end-of-round survey. During the survey, a structured questionnaire was designed and uploaded to smartphones and tablets. The research assistants were trained by the research team in the use of the SurveyCTO Collect in data collection and Good Clinical Practice. Research assistants were deployed to administer the questionnaire utilizing a face-to-face interviewer-administered approach. A trained interviewer guided the respondent through the questionnaire, ensuring accurate and complete responses. The questionnaire was administered to parents or caregivers of SMC-eligible children in the selected households to ensure data quality and integrity. Research assistants recruited were native of the communities and well conversant in the local dialect (Dinka), which they used to translate the English questionnaire during the interview.

The research team supervised data collection to ensure that all data collectors conducted interviews in their assigned clusters. This was done to ensure that the data collection process was carried out per set standards.

Caregivers were asked whether their children had received SMC medicines during the most recent cycle (cycle 5) and the preceding cycles (cycles 1-4). For children who received the first dose, caregivers were further asked whether it was administered as directly observed therapy (DOT) by a community health worker and whether the child subsequently completed the full three-day course of SMC medicines for each cycle.

The survey also gathered information on caregiver awareness, knowledge, and beliefs regarding SMC. Sociodemographic information included child-level characteristics [age and sex], caregiver characteristics [age, gender, education level, employment status, and health-seeking behaviour for febrile childhood illnesses], and household-level factors, including place of residence [urban or rural], mosquito net ownership and use, and indoor residual spraying.

Study variables

The outcome variable was cohort retention, which was defined as a child being enrolled in SMC from cycle 1 and continuing to complete all the remaining cycles and adhering to a 3-day course of SMC medicines per cycle. Cohort retention was a binary outcome (“Yes” if the child was retained; “No” if not). The independent variables considered were the child’s age and sex, caregiver’s age and sex, level of education, knowledge of the importance of SMC, confidence in the effectiveness of SMC medicine, practice of nomadism, and occupation.

Data analysis

Statistical analysis was carried out using Stata 19. Descriptive statistics were used to characterize the study population and assess key attributes within the sample, and frequency tables were generated for variables such as age, gender, caregiver education level, caregivers’ knowledge of the purpose of SMC, caregivers’ confidence in the effectiveness of SMC, nomadism practice by caregivers and caregivers’ status of residence in the area at the start of the campaign in June 2024. Results were reported as frequencies, percentages (Musau, 2023), and 95% confidence intervals (Musau, 2023). The characteristics of the participants were compared using the Chi-square test.

Regression analysis was conducted using the modified Poisson model with robust standard errors since our outcome was highly prevalent. While logistic regression is commonly applied to binary outcomes, the resulting odds ratios tend to inflate the association when the outcome is highly prevalent. Therefore, a modified Poisson approach was used to provide more interpretable prevalence ratio estimates (Barros and Hirakata, 2003).

Bivariable analyses were used to estimate unadjusted (crude) associations between cohort retention and potential factors associated with it. This was followed by multivariable regression modeling to estimate adjusted associations. Variables with a p-value of 0.2 or less at bivariate analysis were considered. Variables such as caregiver sex and age were selected a priority for inclusion in the multivariable model based on their established association with retention, regardless of statistical significance.

The model was constructed using a stepwise logical modeling approach, where one variable at a time was added to the model until model parsimony was achieved. To account for the hierarchical structure of the data and potential correlation of observations within clusters, a mixed-effects modified Poisson model was fitted with random intercepts at both cluster and county levels. The intra-cluster correlation coefficient was calculated to quantify the extent of the between group variability. The likelihood ratio test was used to compare models with and without clustering. Model performance was compared using the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) (Liu et al., 2023).

The associations were reported as prevalence ratios, with their corresponding 95% confidence intervals (CIs), with statistical significance set at p<0.05.

Results

Characteristics of survey respondents

The demographic characteristics of the caregivers (respondents), and eligible children are summarized in Table 1. Among the eligible children in the survey, 46.8% were male and 53.2% were female. Most eligible children, 93.8%, fall within the age range of 12–59 months, whereas infants aged 3–12 months make up 6.2% of the group. Many respondents were female caregivers, making up 83.1%. Most caregivers were young, with the largest proportion, 38.7%, aged 20–29 years, followed by those aged 30–39 years, 33.5%. A significant proportion (66.9%) had no formal education. Additionally, over half (65.5%) were engaged in agricultural work, while nearly one in five caregivers (22.9%) were unemployed.

Table 1. Characteristics of study participants involved in the survey, 2024, South Sudan.

Fewer than 5% (4.1%) of the respondents were non-residents in areas where SMC implementation has been in effect since June 2024.

Extent of SMC cohort retention and associated factors among eligible children

Of the 1,500 eligible children sampled, 66.9% (95% CI: 64.4 – 69.2) were retained. Prior to the regression analysis, the ICC was estimated at 0.58, indicating that 58% of the variability in retention was attributable to differences between clusters. Additionally, the likelihood ratio test confirmed that the model that accounted for clusters was a better fit for the data than the model without clustering (P<0.001).

Table 2 presents the results of the bivariable and multivariable analysis of factors associated with retention of SMC-eligible children. At bivariable analysis, the variables: caregivers involvement in agricultural work (uPR, 2.09: 95%CI, 1.63-2.69: P = 0.001), or paid work (uPR, 3.20: 95% CI, 2.11-4.86: P = 0.001); confidence in SMC medicine effectiveness (uPR, 27.8: 95%CI, 8.49-91.3: P = 0.001); knowledge of SMC importance (uPR, 8.75: 95%CI,5.33-14.4: P = 0.001); and residence in an area since June 2024 (uPR, 2.31: 95%CI, 1.38-3.87: P = 0.001), showed significant positive associations with retention. On the other hand, factors that did not show significant association with retention included caregivers’ gender (uPR, 0.86: 95%CI, 0.65-1.14: P = 0.294), caregivers aged 20–29 years (uPR, 0.81: 95%CI,0.56-1.14: P=0.210); child gender (uPR, 1.06: 95%CI, 0.85-1.31: P = 0.628); child age 3<12 months (uPR, 0.89: 95%CI, 0.58-1.39: P = 0.619); and no education (uPR, 1.46: 95%CI, 0.94-2.25: P = 0.089).

Table 2. Unadjusted and adjusted prevalence ratios for retention in SMC and associated factors.

uPR, Unadjusted prevalence ratio; aPR, Adjusted prevalence ratio.

Bold means statistically significant association values.

In the multivariable analysis, the prevalence of retention among children whose caregivers had paid work was 28% higher than the prevalence of retention among children with unemployed caregivers, while holding other factors constant (aPR,1.28: 95%CI, 1.01-1.62: P = 0.040). However, the prevalence of retention among children whose caretakers did agricultural work was not statistically significantly different from that of children whose caretakers were unemployed (aPR, 1.16; 95% CI, 0.94-1.44; P = 0.162). Retention was positively associated with children whose caretakers had confidence in effectiveness of SMC medicines (aPR, 4.85, 95% CI: 1.25–18.9: P = 0.023), and those whose caregivers had knowledge of the purpose of SMC (aPR,2.04: 95% CI: 1.34–3.11 Compared with the children aged 12–59 months, the prevalence of retention was not significantly different from that of children aged 3<12 months (aPR, 0.91: 95% CI: 0.79–1.04: P = 0.161), When compared to children with caregivers aged less than 20 years, the retention among children with caregivers’ aged 20–29 years (aPR, 0.93: 95% CI: 0.85–1.02: P = 0.144), 30–39 years (aPR, 0.99: 95%CI, 0.91-1.08: P = 0.825), 40–49 years (aPR,0.98: 95%CI,0.86-1.11: P = 0.715) 50–59 years (uPR,0.92:95%CI,: 0.78–1.09, P = 0.335) and above 60 years (aPR,0.90: 95%CI,0.66-1.23: P = 0.514) was not significantly any different. In line with education level, the children whose caregivers had no education (aPR, 1.13: 95% CI, 0.96–1.34, P = 0.143) were not significantly different from those with caregivers who had secondary or higher education.

Discussion

The findings reveal that of the 1,500 eligible children studied, 66.9% (95% CI: 64.4–69.2) received SMC medicines across all five cycles. This indicates a first-to-fifth-SMC dose dropout rate of 33.1%. Notably, this study finds that caregivers’ confidence in the effectiveness of SMC medicines, their knowledge of the purpose of SMC, and children whose caregivers have paid work are the main factors associated with SMC cohort retention among eligible children. Thus, the factors associated with SMC cohort retention is vital for shaping SMC distribution and uptake, as highlighted in this study, may be summarized as behavioral, economic, and informational. This finding matches similar studies in sub-Saharan Africa, which also show that confidence in antimalarial drugs improves SMC and related efforts (Antwi et al., 2016; Ogbulafor et al., 2023; Sitoe et al., 2025). This trust largely comes from understanding why SMC is given and the importance of completing all doses, which boosts caregiver confidence and acceptance (Okot et al., 2025). This fits with the Health Belief Model, which suggests that perceived benefits affect how people follow health advice. As previous research indicates, caregivers’ trust in the SMC’s effectiveness increases the likelihood of accepting the medication and ensuring the children complete the multi-day doses rather than only receiving the drugs (Ogbulafor et al., 2023).This trust also helps to ensure SMC is taken consistently, even when there are minor side effects (Ayuiel et al., 2026). Therefore, building trust through community involvement, clear communication about side effects, and support from trusted local leaders, especially in rural areas like the study site, can significantly improve acceptance.

The findings also revealed that caregivers’ knowledge of or understanding of SMC’s purpose was associated with cohort retention. This is consistent with findings from malaria control programs across sub-Saharan Africa (Dhiman et al., 2023; Diarra et al., 2025; Sitoe et al., 2025). Awareness of SMC tends to improve SMC uptake and program reach, since informed caregivers know its purpose, trust its benefits, and will ensure that children receive the treatment when it is offered (Ayamba et al., 2025). These results imply that health educational efforts clarifying SMC’s mechanisms and benefits could greatly increase uptake and help ensure that caregivers have their children receive all scheduled SMC cycles. This is vital, especially in rural and underserved populations, where knowledge of SMC’s role in preventing childhood malaria morbidity and mortality may be limited.

The findings further revealed that children of caregivers with paid work were more associated with being retained in the SMC distribution cycle. Regular work routines often organize daily activities. This setup helps door-to-door SMC distributors find caregivers and children at home during certain hours; as such, paid workers can easily plan to stay at home for this activity without worrying about missed wages (Lasmi et al., 2024). In the study area, people with regular paid jobs tend to attract more relatives to households. This often includes an adult who can stay home and care for the child on SMC days. Furthermore, caregivers with regular jobs are more likely to keep children at home and less mobile during the rainy season, when SMC is distributed (Genadek and Hill, 2017). Nomadic parents or caregivers frequently migrate in search of farming opportunities or other jobs when their children are due for SMC distribution. In contrast, parents with paid employment are less likely to migrate, and their children are unlikely to miss scheduled cycles due to their non-nomadic lifestyle (Bliznashka et al., 2023). Paid employment is also likely to bring additional income, which helps with access to communication tools that track health campaign schedules (Tariku et al., 2025). Moreover, salaried caregivers tend to have broader social networks, more confidence, and greater exposure to public health messages, including SMC. This can help them influence their acceptance of the drug, hence ensuring their children’s adherence (Komuhangi et al., 2022). This calls for flexible SMC distribution schedules for families, considering their work and caregivers’ ability to negotiate work and stay home, to ensure children receive scheduled SMC cycles.

Sociodemographic variables of caregivers (gender, age, education, occupation) and children (age, gender) showed no significant association with SMC cohort retention. This contradicts earlier findings in Nigeria, Burkina Faso, Chad, and Togo, where caregivers aged 50–55 were more likely to ensure their children received SMC. Past studies have also noted that societal norms and gender roles, especially male dominance in decision-making, restrict children’s access to SMC retention (Ibinaiye et al., 2023). In the Aweil study settings, such factors may not have been associated barriers to SMC cohort retention, due to a uniform study population with shared cultural and societal norms (Baluku et al., 2024; Yihune Teshale et al., 2025). Besides, confidence in SMC’s effectiveness and knowledge about SMC may have had a greater influence on cohort retention, also overriding other factors. In addition, the door-to-door approach used to deliver SMC also helps overcome other factors, such as distance and physical barriers.

SMC is a highly cost-effective intervention when coverage and adherence rates are maximized. Strong cohort retention strategies during SMC campaigns, such as follow-up by community health workers, use of mobile technology for reminders, and caregiver education, have been found to improve health outcomes in Burkina Faso, Mali, and Nigeria (Cairns et al., 2020; Gilmartin et al., 2021; Ibinaiye et al., 2023; Kajubi et al., 2023; Ogbulafor et al., 2023). In South Sudan, given the constrained healthcare infrastructure and logistical challenges, achieving high SMC cohort retention requires tailored, context-specific interventions (Doocy et al., 2022). These strategies might involve delivering SMC through combined approaches, such as child immunization or nutrition programs, and encouraging greater local ownership of health initiatives (Ward et al., 2019; Littmann et al., 2024). Efficient documentation and assessment of these approaches can provide key insights, helping expand SMC in remote settings and ensuring that all children receive the full preventive benefits of the intervention.

The study had strengths, including robust probabilistic methods, a sufficiently large sample size, and multistage sampling, which improved representativeness and reduced bias. It also used solid statistical techniques to account for clustering and ensure accurate estimates.

However, some limitations should be acknowledged. The only approved version of the questionnaire was in English therefore, the research assistants agreed during training on how to translate it into the local Dinka language for respondents who did not understand English. This limitation was minimized through recruitment of research assistants who were fluent in both English and Dinka. However, the translation by research assistants might have not been accurate, which could have affected accuracy of the data collected.

Additionally, being a cross-sectional survey, which relies on participants’ self-reporting, it may be prone to recall issues, social desirability bias, or inaccurate responses.

Overall, caregivers’ confidence in SMC’s protective effects, caregivers with paid work, and caregivers’ understanding of the purpose of SMC are the key factors associated with children’s retention in SMC distribution in Aweil South and West counties of South Sudan. Addressing gaps in trust and awareness among caregivers is vital for improving SMC coverage and lowering malaria-related illness and death in children. Community sensitization on the effectiveness of SMC and engagement should be central to strategies aimed at improving retention in SMC cycles as a malaria prevention approach. SMC distribution campaigns should include targeted support for caregivers who are unemployed or working informally. This support aims to prevent their children from dropping out, especially those involved in nomadic activities and agriculture in this community. Outreach strategies must also be designed to fit the needs of caregivers with strict or demanding work hours. Future research can also provide deeper insights into the factors affecting children’s retention in SMC programs in other settings with different sociodemographic and geographical settings as well as with different delivery approaches.

Conclusions

Caregivers’ confidence in the protective effect of SMC and their understanding of the intervention mainly drive cohort retention among children in Aweil South and West. Addressing gaps in trust and awareness among caregivers is crucial for improving SMC coverage and reducing malaria morbidity and mortality among children. Thus, community sensitization and engagement efforts should be a vital consideration when designing strategies to maximize the effectiveness and impact of SMC as a malaria prevention strategy. Further studies may help provide deeper insights into the factors influencing cohort retention of children in SMC programs.

Correction note

A correction has been made to this article. Details can be found at: 10.3389/fmala.2026.1999637.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmala.2026.1913848/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Andre Lin Ouedraogo, Bill and Melinda Gates Foundation, United States

Reviewed by: Joël Ouoba, Health Ministry, Burkina Faso

Donald Hessou-Djossou, Universite Nationale des Sciences Technologies Ingenierie et Mathematiques, Benin

Footnotes

SPAQ, Sulfadoxine-pyrimethamine plus amodiaquine; SMC, Seasonal Malaria Chemoprevention; ICC, Intracluster correlation coefficient; aPR, Adjusted Prevalence Ratio; uPR, Unadjusted Prevalence Ratio; DOT, Directly Observed Therapy; CI, Confidence Interval; MOH, Ministry of Health; RERB, Research Ethics Review Board; NMCP, National malaria control program; NMOH, National Ministry of Health; SMOH, State Ministry of Health; WHO, World Health Organisation; UNICEF, United Nations Children’s Fund; Ref., Reference (category).