Who are the hard-to-reach population? Spatial analysis from a cross-sectional study on the access to bed net and malaria vaccine in the Lake Victoria Region, Kenya

doi:10.21203/rs.3.rs-5332173/v1

Download PDF

Research Article

Who are the hard-to-reach population? Spatial analysis from a cross-sectional study on the access to bed net and malaria vaccine in the Lake Victoria Region, Kenya

https://doi.org/10.21203/rs.3.rs-5332173/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Long-lasting insecticidal nets (LLIN) and vaccines are effective malaria control tools. However, inadequate uptake has been reported in countries where both interventions are available. To maximize the impact these tools provide, it is crucial to identify populations that are not being reached and the barriers to uptake.

Methods

In a cross-sectional study conducted in April 2024 in Kanyamwa Kologi Ward in Homa Bay County, Kenya, 4,662 households in 58 randomely selected villages were visited and interviewed. We examined the proportions of households that: 1) received at least one new LLIN within the previous five months (net distribution), 2) reported all children used LLIN (net usage), 3) reported at least one child had received one dose of the RTS,S vaccine (vaccine uptake), and 4) reported all children had received four doses of the vaccine (vaccine completion). Bayesian spatial autoregression analyses were used to identify the association and its credible intervals (CrI) between the household-level characteristics and the four outcomes.

Results

The overall uptake proportions were 89.9% for net distribution, 84.4% for net usage, 88.2% for vaccine uptake, and 53.7% for vaccine completion. All four outcomes showed geographical heterogeneity with significant (p < 0.05) Moran's I. Households headed by adults of >40 years had higher odds of having received a new LLIN (OR=1.85, 95% CrI 0.93–4.61), having one child who had received one vaccine dose (OR=1.83, 0.69–4.66), and having all children fully vaccinated (OR=2.36, 1.09–5.46), but lower odds of net usage by all children (OR=0.62, 0.40–0.96). Households with five or more children had higher odds of having received a new LLIN (OR=5.51, 1.69–53.8) but lower odds of net usage by all children (OR=0.24, 0.14–0.38) and having all children fully vaccinated (OR=0.20, 0.04–0.61). Distance to the nearest health center was negatively associated with all outcomes. Household wealth was positively associated with all outcomes.

Conclusion

Uptake of LLIN and malaria vaccine in Homa Bay County, Kenya varied by geography and household characteristics. These findings suggest the need for targeted, outcome-specific interventions to enhance overall public health policy uptake in malaria prevention.

Malaria vaccine

LLIN

Vaccine uptake

Mass net distribution

RTS

Kenya

spatial analysis

access

The current global malaria situation is concerning. The progress in the reduction of malaria cases and deaths has stalled since 2015, with an estimated 249 million global cases and 610,000 deaths in 2022¹. Furthermore, Plasmodium falciparum partially resistant to artemisinin and its derivatives used in first-line treatments for malaria, was reported in sub-Saharan Africa (SSA) in 2021², and is becoming a global threat³. Ensuring the universal coverage of existing interventions in targeted populations is a pillar in the global strategy for malaria⁴.

Long-lasting insecticidal net (LLIN) is the most widespread intervention of malaria control and has contributed to the remarkable reduction of malaria cases and deaths since the early 2000s⁵. The most widely used insecticides in SSA countries were pyrethroid-based, but the emergence and spread of insecticide resistance in mosquito vectors has become problematic across the region. In response, the World Health Organization (WHO) issued a conditional recommendation in 2017 for the use of PBO-pyrethroid nets as a strategy to manage insecticide resistance⁶. Several cluster randomized trials (CRTs)^7,8 showed that PBO-pyrethroid nets have greater epidemiological and entomological efficacy compared to pyrethroid-only LLINs in areas of high pyrethroid resistance.

Two safe and protective malaria vaccines, RTS,S/AS01 and R21/Matrix-M have been registered and licensed in several African countries over the last two years. In October 2021, the WHO recommended that children in moderate-to-high malaria transmission settings in SSA receive the RTS,S/AS01 vaccine. RTS,S/AS01 vaccine showed a 36.3% clinical malaria reduction among children aged 5–17 months in its phase 3 trial⁹, and 13% malaria-related child mortality reduction after nearly four years of pilot administration in Kenya, Malawi, and Ghana¹⁰. More recently, in December 2023, the WHO prequalified the R21/Matrix-M vaccine¹¹. In its phase 3 trial, the R21/Matrix-M vaccine showed a 72% clinical malaria reduction among children aged 5–36 months in Burkina Faso¹².

Despite the availability of these effective malaria public health measures, inadequate uptake, commonly set at < 80%, has been reported in many countries^13,14. To fully utilize these tools, it is crucial to identify populations that are not being reached and to gather more evidence on the barriers to uptake across various regions in SSA. Previous studies have shown the relationship between certain factors, such as socioeconomic status and health service factors, and the uptake of LLINs and vaccines^15–18. However, these variables are often geographically correlated, making it better to explicitly consider the spatial distribution to avoid biased estimates of association¹⁹. Employing spatial analysis to identify specific factors associated with the uptake of LLINs and vaccines can contribute new evidence to this field of study.

Mass net distributions are conducted in Kenya every three years; the most recent campaign was officially launched in November 2023²⁰. In the campaign, PBO-pyrethroid nets were distributed free-of-charge to the community to target universal coverage. Meanwhile, the pilot RTS,S vaccination programs were initiated in eight counties in the western region of Kenya in 2019²¹. One of the counties is Homa Bay County, where we conducted a baseline survey for a cluster-randomized trial to evaluate the protective efficacy of a novel vector control tool, Olyset® Plus ceiling nets in April 2024²². Here, leveraging the available cross-sectional data, we aimed to describe the uptake of mass net distribution and malaria vaccine implementation in the Lake Victoria region of Kenya in addition to investigating the characteristics of households that do not adopt these public health measures by using spatial regression analysis. By examining both LLIN and vaccine uptake simultaneously in a spatial context, this study provides a unique opportunity to identify common and divergent factors influencing these crucial malaria control interventions.

Study area

The study area is Kanyamwa Kologi Ward in Ndhiwa Sub County, Homa Bay County. The details of the area are described elsewhere²². Briefly, the ward has one level-four hospital and seven health centers with an approximate population of 33,000. Agriculture is the primary economic activity, with sugarcane as a main commercial crop. Residents also keep animals such as cattle, sheep, goats, and poultry²³. The area generally experiences a long rainy season from March to June and a short one from October to December. In the study area, the latest mass net distribution was conducted in November and December 2023. In addition, RTS,S/AS01 vaccination was introduce to the area in 2019. Children aged 6 to < 12 months were eligible for the first dose. The second and third doses were to be administered at least one month apart, while the fourth dose was scheduled to be given as soon as possible after the child reached 24 months of age, with an upper age limit of 3 years²⁴.

Study design and data collection

We conducted a cross-sectional malaria survey from April 8 to May 3, 2024, as a baseline survey for the cluster randomized trial to evaluate the efficacy of ceiling nets in Kanyamwa Kologi Ward²². Among 58 randomly selected villages, survey teams consisting of community health promoters (CHPs) and laboratory technicians visited all house structures and administered a questionnaire to the mother or the head of the household. The questionnaire included demographics of all household members including name, age, and sex, information about household structures, assets, bed net possession and usage, and malaria vaccination history of children. We modified the questionnaire used in the 2020 Kenya Malaria Indicator Survey (KMIS). Questionnaire responses were collected by CHPs using the Research Electronic Data Capture (REDCap) software on electronic tablets²⁵.

Descriptive analysis

We examined four binary variables at the household level: 1) whether the household received at least one new PBO-pyrethroid net through the mass distribution campaign in 2023 (net distribution), 2) whether all children under 15 years old in the household used a bed net (regardless of whether it was a PBO-net or other LLINs) on the night before the survey (net usage), 3) whether at least one child aged 2–5 years in the household received a single dose of the malaria vaccine (vaccine uptake), 4) whether all children aged 2–5 years in the household received all four doses of the malaria vaccine (vaccine completion). We calculated the proportion of each variable using different denominators. For net distribution, we included all households with and without children; for net usage, we only included households with children under 15 years; for vaccine uptake and completion, we only included households with children aged 2–5 years. As explanatory variables, we targeted four categorical variables: 1) the mean age of the mother/caregiver/household head (15–24, 25–40, and ≥ 41), 2) the number of children under 15 years old in the household (≤ 2, 3–4, and ≥ 5), 3) distance to the nearest health center (≤ 1 km, ≤ 2 km, and > 2km), 4) the quantile of wealth index (low, middle, and high). The caregiver was defined as the main person responsible for the care of children in a household. The wealth index was calculated according to the Demographic and Health Survey (DHS) guidelines²⁶.

Statistical analysis

We assessed spatial correlation in each of the four variables using Moran’s I. Then, we modeled covariates with a logistic regression model with a conditional autoregressive (CAR) Bayes regression model for each of the four binary outcomes by following equations:

$$\:\text{l}\text{o}\text{g}\left(\frac{{p}_{i}}{1-{p}_{i}}\right)\:=\:{\varvec{x}}_{i}^{T}\varvec{\beta\:}+{z}_{i}$$

where $\:{p}_{i}$ is the probability of the outcome of household $\:i$, $\:{\varvec{x}}_{i}^{\:}$ is the covariate vector including a) the mean age of mother, caregiver, and head of the household, b) the number of children per household, c) distance to the nearest health center, and d) wealth index. These four variables were classified into three or four categories. $\:{z}_{i}$ is spatially structured random effects of the Leroux model, which assumes the (conditional) normal distribution N(), as follows:

$$\:{z}_{i}|{z}_{i\ne\:j}\:\sim\:N(\frac{\rho\:\sum\:_{j}{w}_{ij}}{\rho\:\sum\:_{j}{w}_{ij}+1-\rho\:},\frac{{\tau\:}^{2}}{\rho\:\sum\:_{j}{w}_{ij}+1-\rho\:})$$

where $\:{w}_{ij}$ is the element of the spatial weights matrix. We defined $\:{w}_{ij}$ as 1 if the distance between ith and jth households is within 1 km and 0 otherwise. We estimated parameters $\:(\varvec{\beta\:}$, $\:\rho\:$, $\:\tau\:$) by using the Markov Chain Monte Carlo approach with four chains and 20,000 iterations including 500 burn-in periods by CARBayes package in R. The odds ratio (OR) of the probability of each outcome was estimated by $\:\text{exp}\left(\widehat{{\beta\:}_{\:}}\right)$ with a 95% credible interval (95% CrI). Spatial autocorrelation is controlled by the parameters of ρ and τ. In addition, we evaluated the model performance and fitting by using the Deviance Information Criterion (DIC) and Widely Applicable Information Criterion (WAIC) for both types of CAR models and generalized linear models (GLMs) that do not incorporate a spatial weight matrix.

Ethics declarations

Ethical approval

s were obtained from the Mount Kenya University Intituitional Scientific Ethics Review Committee (MKU-ISERC) (approval number: 2565) and the Ethics Committee in Osaka Metropolitan University (approval number: 2024-068). Printed consent forms were read and explained to every house structure head/caregiver and their signatures captured electronically through electronic tablets during the survey.

Uptake of the new LLINs and malaria vaccines

We collected data from 4,662 households in the study area; 31 were outside of Kanyamwa Kologi Ward and excluded. Among the remaining 4,631 households, 3,434 had at least one child under 15 years and 1,930 had at least one child aged 2–5 years. The overall uptake proportions for each outcome were 89.9% for net distribution, 84.4% for net usage, 88.2% for vaccine uptake, and 53.7% for full vaccination (Fig. 1).

The uptake proportions for each dose of the malaria vaccine, net distribution, and net usage, stratified by the responsible seven health centers, are shown in Fig. 2 and Suppelementary Table 1. All health centers achieved over 80% coverage for at least the first dose of the malaria vaccine. However, the coverage declined after dose 2, with only three health centers maintaining coverage above 80% for dose 3. By dose 4, none of the health centers had coverage above 80%; the uptake proportions varied from 44.5–78.4%. For net distribution, all health centers achieved 80% coverage, although net usage by children under 15 years old varied by area from 76.6–88.7% in the different health centers. Generally, vaccine coverage was associated with LLIN distribution and usage, though the association was not consistent across all health centers. For instance, HC_3, which had the highest vaccination coverage, also had the highest net usage but ranked second to last in LLIN distribution (Fig. 2).

Table 1

Odds ratios (95% CrI) of the association between uptake of interventions and household-level attributes.
	Net		Vaccine
	distribution	usage	uptake	completion
Mean age of mother/care giver/head
15–24	ref	ref	ref	ref
25–40	1.65 (0.88–3.59)	0.98 (0.64–1.53)	2.37 (1.09–7.53)	3.31 (1.13–9.81)
≥41	1.85 (0.93–4.61)	0.62 (0.40–0.96)	1.83 (0.69–4.66)	2.36 (1.09–5.46)
No. of children per HH
0–2	ref	ref	ref	ref
3–4	1.78 (1.12–5.60)	0.43 (0.30–0.58)	0.83 (0.40–1.45)	0.43 (0.21–0.80)
≥5	5.51 (1.69–53.8)	0.24 (0.14–0.38)	1.30 (0.62–2.52)	0.20 (0.04–0.61)
Distance to the nearest health center
≤ 1km	ref	ref	ref	ref
>1km to ≤ 2km	0.93 (0.55–1.68)	0.62 (0.45–0.88)	0.74 (0.42–1.25)	0.38 (0.17–0.74)
> 2km	0.41 (0.21–0.85)	0.66 (0.45–1.09)	0.84 (0.32–1.74)	0.12 (0.04–0.39)
Wealth index
Low	ref	ref	ref	ref
Middle	2.09 (1.36–3.65)	1.36 (1.01–1.92)	1.43 (0.74–2.97)	1.00 (0.25–1.97)
High	2.79 (1.66–4.58)	1.81 (1.33–2.62)	1.95 (0.89–4.74)	2.91 (1.32–7.41)

The spatial distributions are shown in Fig. 3 and Supplementary Fig. 1, illustrating clear geographical aggregations for uptake of the LLINs and vaccines. Moran's I was calculated as 0.29 for net distribution, 0.08 for net usage, 0.19 for vaccine uptake, and 0.30 for vaccine completion, indicating strong evidence (P < 0.05) of a spatial correlation in all variables of interest. However, these geographical aggregations varied depending on each outcome. For instance, in the southernmost part of the study area, a higher proportion of households reported that none of their children had ever been vaccinated, despite the area's high coverage of net distribution.

Factors associated with the uptake of the new LLINs and malaria vaccines

We showed OR of each of four outcomes regarding 1) the mean age of mother/caregiver/head, 2) the number of children per household, 3) distance to the nearest health center, and 4) wealth index by the spatial regression in Table 1. The household with an older mean age of mother/caregiver/head had increased odds of net distribution, vaccine uptake, and vaccine completion. Compared to the household with the mean age categorized as 15–24, the ORs of net distribution, vaccine uptake, and vaccine completion in the household with the mean age over 40 years old were 1.85 (95% CrI 0.93–4.61), 1.83 (0.69–4.66), and 2.36 (1.09–5.46), respectively. For net usage, the OR was 0.62 (0.40–0.96). If the household had more than two children, the odds of net usage and vaccine completion for all children were decreased, but more tend to get new LLIN. The ORs of net distribution, net usage, vaccine completion in the household with five or more children were 5.51 (1.69–53.8), 0.24 (0.14–0.38), and 0.20 (0.04–0.61), respectively. There was no clear trend about the association between the number of children and vaccine uptake. Distance to the nearest health center was negatively associated with all four outcomes. Especially, if the household was located more than 2 km from the nearest health center, the OR of vaccine completion was the lowest at 0.12 (0.04–0.39). Finally, the wealth index of households was positively associated with all outcomes. Compared to the household with the lowest wealth index, the ORs of net distribution, net usage, vaccine uptake, and vaccine completion in the household with the highest wealth index were 2.79 (1.66–4.58), 1.81 (1.33–2.62), 1.95 (0.89–4.74), and 2.91 (1.32–7.41), respectively. Supplementary Tables 2 and 3 showed the number and proportion of each stratified population regarding net and vaccine outcomes, respectively. With lower DIC and WAIC, CAR models for all four outcomes showed better fits to data than GLMs that did not incorporate a spatial weight matrix (Supplementary Table 4).

This study described the uptakes of LLINs and the RTS,S/AS01 malaria vaccine in the Lake Victoria region of Kenya in 2024 and examined the factors influencing the uptakes. Although several reports from various areas have addressed LLIN^15,27 and malaria vaccine^16,17,28 uptake independently, this is the first study to investigate simultaneously both interventions in the same area. We found some common household-level characteristics with four outcomes: net distribution, net usage, vaccine uptake, and vaccine full take. Older average age of the mother, caregiver, or head of the household, proximity to a health center, and higher household wealth were linked to a higher likelihood of receiving or complying with the interventions. However, geographical differences in adoption and compliance could be clearly discerned, and the spatial patterns differed among the measured outcomes. These findings suggest that different sets of actions should be considered to improve the coverage and compliance of these interventions in different areas.

We found that 89.9% of households received new PBO-pyrethroid LLINs around five months after the mass net distribution. This figure is slightly lower compared to the previous report in Uganda^15,27, which showed 93.4% of households owned at least one new LLIN after 1–5 months of mass net distribution in 2020–2021. Moreover, Bhatt et al. estimated a median LLIN retention time of 23 months in 40 African countries⁵. Efforts should be made to get as close to 100% coverage of LLIN distribution as possible, given the possibility of nets being damaged or lost over time. From the regression analysis, we found the households that did not receive the new LLINs were more likely to be headed by younger adults, have fewer children, live farther from a health center, and be less wealthy. Given that nets can be an asset, the reasons for not receiving the new nets are likely to be different from the reasons for not using them or not having their children vaccinated. More likely, people wanted new nets but were omitted from the mass distribution. During the net distribution process, household heads were instructed to collect their new nets from designated locations such as health facilities and primary schools. Households that did not receive new nets either failed to visit the collection point or were not included in the distribution registry. As these people who did not get nets could be socially isolated, it is important to incorporate these people into public policy targets. In addition, the mass net distribution campaign in 2023 was the first time an electronic registration system was used to identify eligible households in the study area. Anecdotally some CHPs participating in both our study and the mass net distribution mentioned that there was some difficulty in utilizing the new technology, possibly resulting in the omission of some households.

Compared to net distribution, net usage showed less geographical clustering, suggesting that measures need to be directed towards a wider area population in order to increase net usage. A good approach to reaching a wider population is spreading the knowledge about malaria prevention. Kanyagarara et al. reported that knowledge of LLIN was associated with a 30–40% increased OR of net use in Zimbabwe and Zambia²⁹. Additionally, previous studies in SSA reported that household size was strongly associated with inequality in the use of LLIN^27,29,30, which is consistent with our results. Moreover, Tamari et al. reported sharing LLIN with two or more individuals may compromise its protective effect in areas close to our study site³¹. It is important to recognize that in large households, every member may not have equal access to LLINs, and a single net is often shared by several members. Another issue we observed in the field was the lack of awareness regarding the expiry date of the LLINs. Although this was not part of a structured survey, we identified several people who were unaware that the nets are treated with insecticide and will lose their effectiveness after a few years. As a result, people tend to keep using old nets until they are physically damaged, which may lead to lower acquisition and usage of the new nets³².

Regarding malaria vaccine coverage, we found that only the dose 1 uptake met the WHO coverage target of 80% across all health centers in the study area. After dose 2, coverage rates dropped dramatically, a trend commonly reported in other studies of malaria vaccines^14,17,28 and other multi-dose childhood vaccines^33,34. Since the regimen of the nelwy launched R21/Matrix-M vaccine is similar to that of the RTS,S vaccine, factors that lower the full uptake of the latter are likely to apply to the former. From the regression analysis, we found that households that did not complete the full RTS,S regimen were typically headed by younger adults, had more children, lived farther from a health center, and were less wealthy. One possible explanation for these findings is that households with these characteristics may not have the time and resources to ensure that all children are fully vaccinated. These populations should be prioritized to increase full-dose coverage of malaria vaccines. Although many efforts have been made to increase vaccination completion under the current immunization schedule, there have also been discussions about the possibility of changing the schedule itself. Alongside the development of a new vaccine with a shorter regimen³⁵, there is an ongoing study evaluating modification of the regimen and dosage of RTS,S vaccines without compromising their efficacy³⁶. This should be focused on which regimen has a chance to increase the completion rate in each area setting based on the local context. We believe this could be the ultimate form of tailor-made malaria control strategy.

Households with eligible children that did not take any dose of malaria vaccines differ from those with children who were not fully vaccinated based on the spatial distribution (Fig. 3) and the regression analysis (Table 1). Especially, proximity to the nearest health center was not associated with vaccine uptake. This suggests that vaccine hesitancy may be more prominent in some areas and/or populations. Simbeye et al. reported that in Malawi, some did not take any dose of malaria vaccine because of religious beliefs³⁴. In our study, some CHPs reported that certain religious leaders discouraged their followers from accessing publicly available health services, including malaria vaccination, as a reason for the abundance of malaria-unvaccinated households in their areas. In addition, it is known that trust in vaccines–not just malaria vaccines–has a profound effect on vaccine hesitancy in SSA countries^37,38. Unfried et al. showed that individuals’ trust in the government and society are key predictors of hesitancy towards polio, human papillomavirus, and COVID-19 vaccines in six countries (Ghana, Kenya, Nigeria, South Africa, Tanzania, and Uganda). Incorporating factors related to vaccine trust would provide a more holistic understanding of the barriers of malaria vaccine in future studies.

One of the strengths of our analysis is the incorporation of spatial aspects. Although there is growing awareness of spatial dependencies driven by proximity or shared social and environmental factors³⁹, there are still limited studies that explicitly account for spatial aspects in malaria research⁴⁰. While our regression analysis reaffirmed the known factors influencing LLIN and vaccine uptake, it is noteworthy that we found similar associations even after considering spatial correlation.

There are several limitations in this study. First, as we utilized baseline survey data from a cluster randomized trial that was not originally designed for this study, we did not conduct a sample size calculation for the analysis. As a result, we may not have achieved precise estimates, particularly in the regression analysis. Nevertheless, we categorized each explanatory variable into three strata, which may suggest dose-response trends. Second, although CHPs were instructed to review the mother-child handbook to obtain malaria vaccination history, verbal reports were sometimes used if the mother had lost her handbook or if time constraints limited the survey. This introduces the possibility of recall bias, particularly regarding vaccination history. Additionally, the analysis may include individuals who moved into the study area from regions where the pilot vaccination program was not implemented.

In conclusion, as the characteristics of households who did not receive or comply with net distribution, net use, vaccination, and full vaccination overlap but differ slightly within the same area, tailored activities should be implemented to enhance the overall uptake of different interventions.

Acknowledgements

We would like to express our sincere gratitude to this study’s participants, field, and laboratory staff, and Naoko Magafu as our project administrator. In addition, we acknowledge the collaboration and support of the Homa Bay County Ministry of Health in Kenya.

Authors’ contributions

YKK and WK developed the original concept. All authors discussed and contributed to the study protocol. YKK, JK, VO, JO, PO, and KBM participated and supervised the cross-sectional survey. YKK drafted the manuscript. DY participated as a senior statistician. The authors read and approved the final manuscript.

Funding

YKK was financially supported by the Japan Society for the Promotion of Science. AK and JG received support from JICA/AMED joint research project (SATREPS) (Grant no. 20JM0110020H0002), Hitachi Fund Support for Research Related to Infectious Diseases, and Sumitomo Chemical Corporation. The funding bodies play no role in the study design, data collection, analysis, interpretation, and publication.

Availability of data and materials

The study regimes, consent forms, and study-related materials are accessible from the corresponding author. The corresponding author will make the de-identified datasets and source codes for all analysis available upon reasonable request.

World Health Organization. World Malaria Report 2023. (2023).
Stokes, B. H., Ward, K. E. & Fidock, D. A. Evidence of Artemisinin-Resistant Malaria in Africa. The New England journal of medicine vol. 386 1385–1386 (2022).
Dhorda, M. et al. Artemisinin-resistant malaria in Africa demands urgent action. Science 385, 252–254 (2024).
World Health Organization. GLOBAL TECHNICAL STRATEGY FOR MALARIA 2016–2030. https://www.who.int/publications/i/item/9789240031357#:~:text=The%20Global%20technical%20strategy%20for%20malaria.
Bhatt, S. et al. The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015. Nature 526, 207–211 (2015).
Gleave, K., Lissenden, N., Chaplin, M., Choi, L. & Ranson, H. Piperonyl butoxide (PBO) combined with pyrethroids in insecticide-treated nets to prevent malaria in Africa. Cochrane Database Syst. Rev. 5, CD012776 (2021).
Protopopoff, N. et al. Effectiveness of a long-lasting piperonyl butoxide-treated insecticidal net and indoor residual spray interventions, separately and together, against malaria transmitted by pyrethroid-resistant mosquitoes: a cluster, randomised controlled, two-by-two factorial design trial. Lancet 391, 1577–1588 (2018).
Staedke, S. G. et al. Effect of long-lasting insecticidal nets with and without piperonyl butoxide on malaria indicators in Uganda (LLINEUP): a pragmatic, cluster-randomised trial embedded in a national LLIN distribution campaign. Lancet 395, 1292–1303 (2020).
RTS,S Clinical Trials Partnership. Efficacy and safety of RTS,S/AS01 malaria vaccine with or without a booster dose in infants and children in Africa: final results of a phase 3, individually randomised, controlled trial. Lancet 386, 31–45 (2015).
Wadman, M. First malaria vaccine slashes childhood deaths. Science 382, 357 (2023).
WHO. WHO Prequalifies a second malaria vaccine, a significant milestone in prevention of the disease. Geneva: World Health Organization. https://www.who.int/news/item/21-12-2023-who-prequalifies-a-second-malaria-vaccine-a-significant-milestone-in-prevention-of-the-disease (2023).
Datoo, M. S. et al. Safety and efficacy of malaria vaccine candidate R21/Matrix-M in African children: a multicentre, double-blind, randomised, phase 3 trial. Lancet 403, 533–544 (2024).
Bertozzi-Villa, A. et al. Maps and metrics of insecticide-treated net access, use, and nets-per-capita in Africa from 2000-2020. Nat. Commun. 12, 1–12 (2021).
Osoro, C. B. et al. Policy uptake and implementation of the RTS,S/AS01 malaria vaccine in sub-Saharan African countries: status 2 years following the WHO recommendation. BMJ Glob. Health 9, e014719 (2024).
Kwiringira, A. et al. Ownership and use of long-lasting insecticidal nets three months after a mass distribution campaign in Uganda, 2021. Malar. J. 21, 367 (2022).
Yeboah, D., Owusu-Marfo, J. & Agyeman, Y. N. Predictors of malaria vaccine uptake among children 6-24 months in the Kassena Nankana Municipality in the Upper East Region of Ghana. Malar. J. 21, 339 (2022).
Simbeye, A. J. et al. Factors associated with malaria vaccine uptake in Nsanje district, Malawi. Malar. J. 23, 105 (2024).
Nyalundja, A. D. et al. Socio-demographic factors influencing malaria vaccine acceptance for under-five children in a malaria-endemic region: A community-based study in the Democratic Republic of Congo. Vaccines (Basel) 12, (2024).
Jarvis, C., Di Tanna, G. L., Lewis, D., Alexander, N. & Edmunds, W. J. Spatial analysis of cluster randomised trials: a systematic review of analysis methods. Emerg. Themes Epidemiol. 14, 12 (2017).
Ministry of Health, Kenya. Mass net distribution campaign 2023/2024. https://www.health.go.ke/node/831.
Moturi, A. K. et al. Malaria vaccine coverage estimation using age-eligible populations and service user denominators in Kenya. Malar. J. 22, 287 (2023).
Ko, Y. K., Kagaya, W., Omondi, P., Musyoka, K. B. & Okai, T. Evaluation of the protective efficacy of OlysetPlus ceiling nets for reduction of malaria incidence in children in Homa Bay County, Kenya: a cluster-randomized …. medRxiv (2024).
Homabay County | Ndhiwa Municipality. https://www.homabay.go.ke/department/ndhiwa-municipality.
Hill, J. et al. Integration of the RTS,S/AS01 malaria vaccine into the Essential Programme on Immunisation in western Kenya: a qualitative longitudinal study from the health system perspective. Lancet Glob Health 12, e672–e684 (2024).
Harris, P. A. et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J. Biomed. Inform. 42, 377–381 (2009).
Rustein, S. O. Steps to Construct New DHS Wealth Index. https://www.dhsprogram.com/programming/wealth%20index/Steps_to_constructing_the_new_DHS_Wealth_Index.pdf.
Okiring, J. et al. LLIN Evaluation in Uganda Project (LLINEUP2)-Factors associated with coverage and use of long‑lasting insecticidal nets following the 2020-21 national mass distribution campaign: a cross-sectional survey of 12 districts. Malar. J. 21, 293 (2022).
Hoyt, J. et al. RTS,S/AS01 malaria vaccine pilot implementation in western Kenya: a qualitative longitudinal study to understand immunisation barriers and optimise uptake. BMC Public Health 23, 2283 (2023).
Kanyangarara, M. et al. Malaria knowledge and bed net use in three transmission settings in southern Africa. Malar. J. 17, 41 (2018).
Hailu, A. et al. Equity in long-lasting insecticidal nets and indoor residual spraying for malaria prevention in a rural South Central Ethiopia. Malar. J. 15, 366 (2016).
Tamari, N. et al. Antimalarial bednet protection of children disappears when shared by three or more people in a high transmission setting of western Kenya. Parasitology 146, 363–371 (2019).
World Health Organization. Update on subnational tailoring of malaria interventions and strategies. https://cdn.who.int/media/docs/default-source/malaria/mpac-documentation/mpag-march2024-session4-subnational-tailoring-of-interventions-rev.pdf?sfvrsn=6eeebf97_3 (2024).
Acharya, P., Kismul, H., Mapatano, M. A. & Hatløy, A. Individual- and community-level determinants of child immunization in the Democratic Republic of Congo: A multilevel analysis. PLoS One 13, e0202742 (2018).
Biset, G., Woday, A., Mihret, S. & Tsihay, M. Full immunization coverage and associated factors among children age 12-23 months in Ethiopia: systematic review and meta-analysis of observational studies. Hum. Vaccin. Immunother. 17, 2326–2335 (2021).
Sissoko, M. S. et al. Safety and efficacy of a three-dose regimen of Plasmodium falciparum sporozoite vaccine in adults during an intense malaria transmission season in Mali: a randomised, controlled phase 1 trial. Lancet Infect. Dis. 22, 377–389 (2022).
Samuels, A. M. et al. Efficacy of RTS,S/AS01E malaria vaccine administered according to different full, fractional, and delayed third or early fourth dose regimens in children aged 5-17 months in Ghana and Kenya: an open-label, phase 2b, randomised controlled trial. Lancet Infect. Dis. 22, 1329–1342 (2022).
Unfried, K. & Priebe, J. Vaccine hesitancy and trust in sub-Saharan Africa. Sci. Rep. 14, 10860 (2024).
Wollburg, P., Markhof, Y., Kanyanda, S. & Zezza, A. Assessing COVID-19 vaccine hesitancy and barriers to uptake in Sub-Saharan Africa. Commun. Med. (Lond.) 3, 121 (2023).
Francetic, I. et al. Framework for identification and measurement of spillover effects in policy implementation: intended non-intended targeted non-targeted spillovers (INTENTS). Implement Sci Commun 3, 30 (2022).
Jarvis, C. I. et al. Spatial Effects of Permethrin-Impregnated Bed Nets on Child Mortality: 26 Years on, a Spatial Reanalysis of a Cluster Randomized Trial. Am. J. Trop. Med. Hyg.101, 1434–1441

No competing interests reported.

Supplementaryfile.docx

Download PDF

Editor assigned by journal
03 Nov, 2024
Submission checks completed at journal
25 Oct, 2024
First submitted to journal
25 Oct, 2024

You are reading this latest preprint version

Who are the hard-to-reach population? Spatial analysis from a cross-sectional study on the access to bed net and malaria vaccine in the Lake Victoria Region, Kenya

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study area

Study design and data collection

Descriptive analysis

Statistical analysis

Ethics declarations

Results

Uptake of the new LLINs and malaria vaccines

Factors associated with the uptake of the new LLINs and malaria vaccines

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1