Vaccinations are a reliable and cost-effective way to decrease the burden of disease and improve the overall health of a community (1). Every year, immunizations save 2 to 3 million lives, and are one of the most effective ways to reduce global health inequities (2, 3). Despite global improvements in childhood immunization over the past decade, disparities in vaccination status persist within and across countries (4). Particularly in low- and middle-income countries (LMICs), differences in vaccine coverage exist based on one's socioeconomic status, geographical location, education level, and sex (5).

LMICs often exhibit low vaccine coverage rates relative to other countries (6). A majority of the global population of children with an incomplete vaccination series reside in LMICs (7). Several factors contribute to low rates of vaccine uptake in LMICs, including a lack of political support for vaccination campaigns, greater resource allocation to other health issues, and poor education and awareness about vaccines among healthcare workers and parents, specifically mothers (8–10). The lack of education about the benefits of immunization and risks of disease in LMICs further limits vaccine uptake, leading individuals to believe that immunization has a high cost-benefit ratio (8). Furthermore, insufficient human resources, lack of computerized vaccination registries and efficient communication systems with patients, and frequent issues with cold supply chains negatively impact vaccination rates in LMICs (9, 11).

Madagascar, an island country in Sub-Saharan Africa home to roughly 25 million people, has been identified as having one of the largest disparities in immunization rates in the world (5). Since 1976, Madagascar has followed the World Health Organization's (WHO) Expanded Programme on Immunization (EPI) to create a vaccine schedule (12). Recent evidence shows that immunization rates remain low, with WHO reporting MCV and DTP3 coverage at 69% and 79% in 2019 (13).

One 2010 WHO report identified a significant difference in childhood vaccination status based on sex in Madagascar, with girls more likely than boys to be fully vaccinated (14). Additionally, in Moramanga, Madagascar, there is a vast disparity in risk of death by age 15, at 58.4% for girls and 77.6% for boys, further indicating potential differential healthcare access between sexes in Madagascar (12). Sex differences in vaccination status have been identified in several other countries in Sub-Saharan Africa such as Malawi, Tanzania, and Namibia (14), demonstrating the existence of sex differences in immunization in other SSA countries.

Although imperative to understanding potential differences in vaccine uptake between boys and girls, research specifically identifying and quantifying the association between a child's sex and immunization uptake in Madagascar is limited (15). The purpose of this review was to assess the association between a child's sex and vaccination rates among children eligible for routine immunization programs, determine the availability of sex-disaggregated vaccination data, and identify other obstacles to vaccination in Madagascar.

The systematic search of four libraries retrieved 585 results, 29 (49.6%) from the Cochrane library, four (6.8%) from GIM, 402 (68.7%) from Google Scholar, and 150 (25.6%) from PubMed. Following rounds of title and abstract screening, full-text screening was conducted for 100 remaining records, of which 96 were excluded. Of these 96 records, 23 (24.0%) were excluded for not reporting any vaccination data, 26 (27.0%) were excluded for not including data specific to Madagascar, 19 (19.8%) were reviews, two (3.4%) contained duplicated data sources, and one (1.0%) did not report data specific to children.

A total of 44 records were identified through the gray literature search. Of these, 42 (95.4%) were excluded because they did not contain sex-disaggregated vaccination data and one (2.3%) was excluded because it had an undefined study population. The remaining publication by GAVI (19) had an identical data source to an article previously identified and included by the systematic search by Mutua et al. (20). The GAVI source reported vaccination data in terms of percentages of children vaccinated as opposed to the slope index of inequality used in Mutua et al.'s study, so the GAVI source was included instead (Supplementary Table 4). In total, three articles from the systematic search and one database from GAVI were finally included for data extraction (Figure 1). All four included articles presented secondary data originally collected through DHS and MICS and all extracted data are publicly available.

Risk of bias assessments were conducted on the four included publications using the AXIS tool (21). Three studies had no risk of bias and were considered of high quality (19, 22, 23) and one study had a moderate risk of bias (24) (Table 1). The cross-sectional study conducted by Restrepo-Méndez et al. and the 2019 GAVI report did not demonstrate any risk of bias and thus are of high quality (19, 22). Bosch-Capblanch et al. did not include a consolidated methods section but was otherwise of high quality (23). Hill et al. had a small sample size which, although recognized and acknowledged by the researchers, detracts from the quality of the study. Additionally, Hill et al. did not explicitly declare a conflicts of interest statement. Although this could be attributed to a change in the reporting of conflicts of interest over time, as this is the oldest study included in the review, this publication received a moderate quality grade (24) (Table 1).

All publications reported coverage of the basic vaccines recommended through the EPI program: BCG, DTP3, Polio3, and Measles. Three of the publications reported vaccination rates for children 12–23 months old (19, 22, 24) and one for children 12–59 months old (23). Two publications measure the outcome of interest in the form of percentages of boys and girls vaccinated (19, 22); one publication as the difference between percentage of girls and boys vaccinated (24), and another publication used an odds ratio to illustrate the sex difference (23). Data sources for the publications included were the Demographic and Health Surveys (DHS) in 1992 (24), 2004 (23), and 2008 (22); and a Multiple Indicator Cluster Survey (MICS) in 2018 (19).

Three included studies reported higher vaccination in girls (22–24) and one reported higher vaccination in boys (19) (Table 2).

Among the 100 full-text articles assessed for reported additional obstacles to immunization, four publications mentioned obstacles other than sex (4, 5, 25, 26). Socioeconomic status, mother's literacy and education, geographic location, supply chain issues, father's education, number of children in the household, and media access were all found to impact immunization status in Madagascar. Three publications mentioned socioeconomic status (4, 5, 25), three mother's literacy and education (4, 5, 26), two geographic location (4, 5), two equipment management and transportation (4, 5), one father's education (4), one number of children in household (26), and one media access (26).

The evidence collected in this systematic review shows a minimal difference in reported vaccination rates between boys and girls (22–24). However, the scarcity of evidence and heterogeneous nature of the data limit our ability to conclusively confirm or deny the existence of sex-based disparities in Madagascar's vaccination program. Of the four included reports, one reports a statistically significant sex difference in vaccine uptake (23), one reports no difference (22), and two report a two percent difference in vaccine uptake between boys and girls but no measure of statistical significance (19, 24). These conflicting reports and the lack of overall data mean that additional sex-disaggregated data about the EPI vaccine uptake of children are required to confirm or deny sex-based disparities in vaccination status in Madagascar.

Sex differences in vaccine uptake have been previously identified in other African countries including Mozambique and the Democratic Republic of the Congo (20). Conversely, sex has also been found to have no impact on vaccine uptake in Nigeria, Ethiopia, and Kenya (27–29). Interestingly, one report found a statistically significant sex imbalance in vaccination in East Africa (20) despite additional reports claiming that sex did not influence vaccination status in East Africa (30) or Sub-Saharan African as a whole (31). Such inconsistencies may reflect the fact that local or regional disparities can be masked by regional or national averages (4) and the existence of countries with and without sex-based differences in vaccine uptake across Africa. Analyses of immunization trends in Africa thus indicate that the existence or lack of sex-based disparities in EPI vaccine uptake in Madagascar are both plausible, again highlighting a need for greater reporting of sex-disaggregated vaccination data.

Further demonstrating the importance of collecting and analyzing vaccination data based on individual demographic factors, we identified several other factors causing differential vaccine uptake in Madagascar including a household's socioeconomic status (4, 5, 25), geographical location (4, 5), mother's education and literacy status (4, 5, 26), and the management and transportation of immunization resources (4, 5). Such barriers in Madagascar reflect reports citing socioeconomic status (26, 30, 32) community wealth (30), parent's education (30, 32), and media access (30) as causes of differential vaccination uptake in other African countries. It is important to note that the articles screened for additional obstacles to vaccination were identified via the original search strategy focusing on a child's sex and vaccination, likely narrowing our results. Reports of additional barriers are intended to indicate the presence of other factors associated with vaccination status to provide ideas for future direction and focus rather than providing a comprehensive report.

The influence of maternal education level on vaccination status is of note and was found to strongly impact a child's vaccination status in Madagascar. A 2019 GAVI report found that 57% of children whose mothers had at least a secondary level of education were vaccinated compared to only 24% of children with uneducated mothers (19). Another study reported that children with educated mothers were 1.7 times more likely to be vaccinated than children with uneducated mothers (33). The increases in household wealth and decreases in fertility associated with educated mothers positively influence childhood survival rates in Madagascar, and we posit that such effects may account for the influence of mother's education over vaccination status as well (34). Further research to understand why female education status impacts a child's vaccination status in Madagascar is needed to confirm or deny these hypotheses.

Differences in vaccination rates between wealthy and poor or urban and rural households have also been identified (4, 5, 19, 25). Children in the affluent Itasy region of Madagascar are 3.4 times more likely to be vaccinated than those in the poorer and more remote Menabe region, revealing Madagascar's large regional and socioeconomic inequalities in vaccine uptake (5). Such socioeconomic and geographic differences in vaccination status reflect the obstacles that poorer, rural-dwelling Malagasy people face in accessing healthcare facilities. Unreliable transportation methods, poor roads, and rainy seasons that flood travel routes and render villages unreachable present considerable challenges to health care and vaccination access for rural residents (19, 35), negatively impacting vaccination status (4, 5). The management of resources needed for immunization such as refrigeration, gasoline, and the transportation of cold chain supplies is challenging in rural areas, making it hard to provide adequate resources for vaccinations (4). Furthermore, Madagascar only has 0.14 physicians for every 1,000 patients compared to a global average of 1.4 physicians, creating an additional barrier to healthcare access (36). We recommend taking an intersectional approach to assess whether such barriers identified here differentially affect boys and girls in Madagascar.

In addition to the notable scarcity in sex-disaggregated data and the lack of clear difference in vaccination rates by sex, this review also revealed a recent decrease in immunization rates in Madagascar. The latest comprehensive assessment published in 2019 by GAVI highlights a decrease in immunization in girls from 61.9% (22) to 40% (19) and in boys from 61.4% (22) to 42% (19) between 2008 and 2019. Effective vaccine coverage for the measles, a disease against which EPI vaccines provide protection, is estimated at 95% by WHO, indicating that Madagascar's vaccine coverage is at an alarming low level (37).

Notably, researchers identified stark differences in the reporting of vaccination coverage across sources, potentially explaining the seemingly low immunization rates reported in the 2019 GAVI assessment (19). One WHO publication reported rates of EPI vaccination ranging from 62% coverage for MCV1 to 81% coverage for DTP1 in 2018. This report notes the difference of up to 20% between coverage estimates provided by WHO-UNICEF, the Malagasy government, and administrative reports (38). Another WHO publication reported 2018 coverage for MCV1 at 85% and DTP1 at 97%, again demonstrating a vast discrepancy in immunization data collection and reporting (13). The inconsistencies in the reporting of immunization coverage are troubling and should be of concern to data collection organizations and the Malagasy government.

Regardless of these discrepancies in estimated vaccine coverage levels, vaccination coverage in Madagascar appears to be far below the levels needed to confer herd immunity and prevent outbreaks (37). In 2018, a measles outbreak struck Antananarivo, Madagascar's capital, and extended rapidly to all 22 regions in Madagascar due to low vaccination coverage estimated at 60% by WHO-UNICEF (39). Low immunization rates remain a critical threat to the health of citizens and the decrease in vaccination rates noted in this review is a cause for concern. The COVID-19 pandemic has led to further decreases in vaccination rates (40). As of October 2021, no global datasets had published sex-disaggregated data on COVID-19 vaccine uptake despite existing reports indicating that women in LMICs are less likely to receive COVID-19 vaccines than men (41). The impact of the COVID-19 pandemic on EPI vaccination programs must continue to be monitored closely, and the role of a child's sex must not be forgotten.

One limitation and simultaneously a finding of this research is the lack of sex-disaggregated vaccination data collected in demographic surveys. Many reports contain information about vaccination rates in Madagascar, but few are sex-disaggregated, limiting our ability to assess sex differences in vaccine uptake. Another limitation is the heterogeneous nature of the data collected, with many sources reporting vaccination data differently and preventing researchers from conducting a meta-analysis. A recent scoping review has highlighted the importance of high-quality vaccination data and the current paucity of such in LMICs (42). Our review supports this finding and underlines the urge for consistent methodology in data collection and methods of analyses to allow comparison on relevant levels and inform future adaptation of vaccine distribution strategies.

The lack of sex-disaggregated data identified in this review reveals a gap in current understandings of how a child's sex affects health status. The scarcity of available data not only precludes researchers from determining whether sex influences vaccine uptake in Madagascar but indicates an insufficient understanding of how sex may interact with the additional barriers identified here. These gaps in current understandings of vaccine uptake are concerning as they may allow inequities to remain unaddressed. A more comprehensive assessment of the vaccination status of Malagasy children based on defined variables including sex and household information is desirable to confirm potential demographic factors contributing to inequity in vaccine uptake. A thorough and consistent collection of sex-disaggregated vaccination data in Madagascar would allow for a monitoring of the impact of sex on vaccination status over time and intersectional analyses to understand the relationship between sex and other barriers to vaccination (43). Further, in the event of large-scale disease outbreaks such as the 2018 measles outbreak and the ongoing COVID-19 pandemic, any potential sex imbalances in vaccination could be readily identified and corrected.

This is the first systematic review to look at sex differences in EPI vaccination in Madagascar to our knowledge. The lack of sex-disaggregated vaccination data highlights an unmet need and gap in current understandings of vaccine uptake. Interestingly, socioeconomic status, geographic location, mother's education status, and immunization equipment management were all identified as causes of vaccine inequity in Madagascar, directing attention away from sex and toward overcoming other barriers. We recommend future analyses to weigh the relative impact of such factors and inform interventions designed to improve vaccine coverage.

Of note is our finding that vaccination rates in Madagascar have drastically decreased over the past 15 years, with an additional decrease due to the COVID-19 pandemic. Future monitoring of sex-disaggregated vaccination data would reveal whether similar drastic changes in vaccination rates differentially affect boys and girls. We suggest that increasing the vaccination rate in Madagascar is of the utmost importance. During this process, specific attention should be paid to closing the gaps that are identified between rural and urban, rich and poor, and educated and non-educated populations while maintaining an awareness of the potential for sex inequity in vaccine uptake in the future via the collection of sex-disaggregated vaccination data during routine EPI vaccinations.

Implementing interventions to expand vaccination coverage across LMICs is recommended. Suggested interventions include outreach campaigns (going to houses, schools, markets, and other places where people gather), healthcare education (particularly community-based), and home visits. Such interventions are intended to decrease the reliance on self-presentation by individuals (mothers, in most cases) to the healthcare facility in order to complete EPI in children. Similar strategies have previously been used to improve vaccination rates in LMICs (7). In particular, we recommend the continuation and expansion of National Immunization Days (NIDs), Subnational Immunization Days (SNIDs), Child Health Days, and Vaccination Weeks (VWs) in Madagascar and similar settings. These programs involve establishing additional outreach locations for vaccinations, knocking on doors to vaccinate people, hosting “mother and child health weeks” at primary care facilities, and catching children who missed or are late on a vaccine dose, and have previously helped improve vaccination rates in LMICs (44–47).

We additionally recommend a focus on implementing and expanding media coverage, mass informational campaigns, and direct physician to patient communication about vaccination and the dangers of vaccine-preventable diseases in LMICs. Increased public education and awareness of the benefits of immunization have been identified as useful interventions to increase political support, reduce sociocultural barriers, and improve community attitudes toward immunization (8, 9). Health and government organizations designing informational campaigns must take into account the literacy and education level of the target audience (8) as well as the specific barriers to immunization in a community to ensure the campaign is effective and well-received. We also urge the importance of increasing the quality of data collection and ensuring external consistency of immunization data to allow for accurate future analysis. To do so, potential weaknesses in the surveillance of vaccine uptake data should be assessed and strengthened accordingly.

Directing attention toward new vaccine delivery technologies such as vaccine-containing microarray patches (VMAPs), could have the potential to reduce barriers to immunization in LMICs. With an intradermal delivery method, higher thermo stability, and fewer training requirements for healthcare workers than ordinary vaccines, VMAPs are favorable for delivery to rural areas and regions in conflict (48, 49) Research and development of VMAPs for diseases including measles, rubella, HPV, and influenza are currently underway (49, 50) and should be monitored to determine whether they are suitable for implementation in Madagascar and other LMICs.

The recommendations outlined in this review and our suggestion to increase the collection of sex-disaggregated vaccination data have the potential to improve vaccination rates, attitudes about vaccines, and understandings of how demographic factors impact vaccination status in many LMICs. Greater understandings of the barriers to vaccination and existing inequities will allow for the implementation of specific interventions to target such obstacles. Doing so will ensure that LMICs increase their vaccination rates in a manner that is equitable across demographic groups.

The original contributions presented in the study are included in the article/Supplementary files, further inquiries can be directed to the corresponding author/s.

EH, AH, and LC-E contributed to the study design, methodology, article screening, and data interpretation. EH conducted the literature search and collected data. AH and LC-E provided supervision. GN and RR attempted to access immunization data from the Malagasy government. All authors contributed to manuscript revision, read, and approved the submitted version.

AH, BT, FM, GN, and RR are supported by a grant from the Bill and Melinda Gates Foundation (Grant Number: OPP1205877).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

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.