**4. Discussion**

We found that for every 100 females, 103 males were enrolled and vaccinated in the SEIR over the last 4 years. However, the sub-national analysis at the UC level shows the difference increased to 300 males being vaccinated for every 100 females in specific UCs. Merely observing the aggregate levels for evidence of gender differentials masked these nuanced yet more pronounced inequities. Moreover, recent reports by Gavi [22] and WHO [23] asserted that subnational variations in immunization coverage were 'one of the tractable but unfinished challenges of immunization inequity globally.' Differences at the micro-geographic level reflected subtle and persistent forms of gender bias and discrimination that continue to affect health outcomes for females over the long term. When comparing the male-to-female ratios and gender inequality ratios, we observed a larger number of males than females made contact with the immunization system (even after adjusting for the male-to-female baseline population). However, once they had been enrolled (in the SEIR), the vaccine coverage rates were similar for both females and males, although females still fell behind males in receiving timely vaccinations.

Our findings have important implications for the zero-dose children that have yet to make contact with the health system. Since more males than females have been enrolled in the immunization system, this reflects substantial inequities, indicating more females than males are left behind and being added to the higher proportion of zero-dose children. There is a need for rethinking and emphasizing the narrative of 'zero-dose females', and ensuring the use of gender-disaggregated data and gender-sensitive strategies in order to reach the missing children. We also observed that gender inequities continue to persist over time. The analysis of individual UCs suggested there were certain pockets and regions spread throughout the province where females continuously fell behind males on their vaccinations, year-on-year. Targeted, intensified efforts directed to hotspots showing high inequities could be a potential measure to break the pattern of persistent inequities.

Although parity in coverage rates among females and males enrolled in the SEIR was a positive finding, we observed equality was not uniformly reflected across all age groups. Females were more likely to be delayed in their vaccinations than males. While reflecting well on the overall view of equality, it was imperative to note that as females were delayed on their vaccination, they remained susceptible to vaccine-preventable diseases (VPDs) for longer periods, leading to a higher risk of morbidity and mortality over time. Delayed vaccination for females could have a considerable impact on child survival rates overall,

a pertinent implication for Pakistan, where infant and child mortality rates are some of the highest globally. A study from Bangladesh showed children receiving BCG within the first 6 months of life had a lower risk of diseases than those vaccinated later [24]. Similar results were also reported for the delayed administration of the diphtheria–tetanus and pertussis vaccines [25].

Our findings of higher inequities in the number of vaccinated females and males in rural areas, as compared to urban areas, and slums, as compared to non-slums [26,27], have been repeatedly emphasized in existing literature [28,29]. We went a step further to demonstrate that within the rural areas, the category of remote-rural and hard-to-reach areas fared even worse, with M:F ratios as high as 1.14. Several underlying factors have been cited to explain the inequities, the most prominent being the deep-rooted socio-cultural practice of "son preference", which is inherently common in Pakistan [30] and other South Asian countries [29,31]. Persistent patriarchal practices favor sons over daughters due to factors such as carrying forward the family lineage, providing old-age support, financial support, and practices pertaining to dowries. The phenomenon of son preference has been closely associated with several adverse practices, including gender-selective abortions, female infanticide, and neglect of the health and education of females. In rural and remoterural regions, not only are these practices more deeply entrenched, but when coupled with multiple other deprivations, including poverty, lack of affordable transportation, and long distances to healthcare services, they lead to discriminatory attitudes by caregivers in favor of males. This was underscored in our findings with higher inequities in the number of females and males vaccinated at fixed immunization centers. Immunizations administered during both routine and enhanced outreach tended to be more equitable for females, reflecting that caregivers were not inherently opposed to vaccination, but when faced with the logistical and financial challenges of taking children to vaccination centers, they were more likely to favor males over females.

Within remote-rural settings, several additional dynamics are at play that adversely impacted equitable immunization, such as the higher marginal cost of reaching remote children, retention and motivation of personnel, geographic remoteness, and limited sociopolitical power among communities [32]. The factors were further undercut by gender issues where, in the event of male vaccinators, mothers and female caregivers faced even greater societal restrictions when accompanying children for immunization. Our findings showed that not only did this have an adverse impact on vaccination rates overall, but the lack of female vaccinators disproportionately and adversely affected vaccination outcomes for females, as compared to males. The absence of gender-sensitive policies for immunization was highlighted in our study (none of the 87 remote-rural UCs in Sindh Province had a single female vaccinator (Supplementary Table S4) and mentioned elsewhere including no segregated waiting rooms at immunization facilities for female caregivers and a shortage of female vaccinators in urban impoverished areas, which was a "discouraging factor" for the attendance of females and children at health facilities [26]. Our results showed that the districts of Ghotki, Jacobabad, and Kashmore had high prevalence rates of inequities for females at enrollment and for subsequent antigens. These districts are located within the northern belt of the province, which remains deeply rooted in conservative tribal culture with a high prevalence of other discriminatory practices against females, including domestic violence and forced child marriages [33]. Gender equity in immunizations is not an isolated concept but deeply intertwined with females' empowerment, agency, and autonomy. Increasing females access to education is a proven mechanism to break the perpetual cycle of discrimination. Within the context of immunizations, our findings were in line with others that showed higher maternal education [34,35] led to reduced vaccination inequities for females. However, our study showed that, even with very high levels of maternal education (>11 years), there remained UCs that had extreme inequalities (M:F ratio: 11.0). Upon closer geographic examination, we observed 4 of such UCs were clustered fairly close together, suggesting there could be other prevalent socio-cultural or logistical challenges causing inequities that even higher maternal education levels were unable to overcome. Vaccine hesitancy is one particular challenge that merits further investigation within the context of gender inequities in immunization. One study has articulated the reasons for vaccine hesitancy in Pakistan as a triad of religious traditions, misconceptions, and political factors. [36]. Vaccine hesitancy may contribute to gender inequities in immunization by perpetuating cultural norms and beliefs that prioritize males over females and fuel misinformation and misconceptions about vaccines that disproportionately affect females, limiting access to health services and decision-making power for females.

Addressing gender inequities in immunization requires multilevel, complementary approaches. Feasible policy measures include the inclusion of more female vaccinators in the health workforce. Due to sociocultural and gender norms in underserved communities of LMICs such as Pakistan, only female frontline health workers have unrestricted access to households, are able to interact with mothers and provide health education, and deliver vaccines to children. More female vaccinators could, therefore, promote building trust concerning vaccines and encourage immunization uptake among vulnerable communities. To enhance the female position in the immunization decision-making process for their children, we must focus on overall education for females, specifically in health literacy. A previous study revealed that females who were health literate, regardless of their educational level, were more likely to vaccinate their children, in both rural and urban settings [37]. Additionally, female groups in local settings and communities can be initiated or leveraged as a platform for counseling focused on health literacy. These groups could be complemented with programs to involve fathers, including facilitating regular sessions with females and males to foster collaborative parenting and decision-making. A gender-centric approach to the overall health system should be strengthened by measures such as separate waiting areas for females in immunization clinics and the introduction of female-only transport to immunization centers, which could increase immunization rates among females.

Our study had a few limitations. The adjusted M:F ratios reported in the analyses represented a best-guess given the lack of reliable sex ratios in birth data at district and UC levels. Although 1.055 represented an aggregate number for the country, this masked the heterogeneity and inequities in M:F ratios across districts and union councils. Additionally, studies have shown that sex ratios at birth varied by levels of maternal education [38–40], ethnicity, the birth order of the child, as well as the economic and cultural heterogeneities [41]. Therefore, adjustments using aggregate M:F ratios at birth could mask the true extent of prevailing inequities at the sub-national level. Moreover, the M:F ratios calculated for maternal literacy, geographic location, and mode of vaccination delivery were not adjusted for the underlying proportion of males and females in the population due to the unavailability of baseline population proportions for these categories. Nonetheless, we speculated that even if these were to be adjusted, the high M:F ratios (up to 5.00 at the UC level) still reflected substantial inequities between females and males. To validate our estimates of M:F ratios further, we correlated them with the gender-wise proportions in the Multiple Indicators Cluster Survey (MICs). However, since the sample size in the MICs was small when compared against our analysis categories, no meaningful, statistically significant correlations were found between the gender proportions in our analysis and MICs. Additionally, only 58% of the remote-rural UCs in the province, as per our source, were matched with the UC database in SEIR due to different names and a variation in UC categorization used by the health and education departments. Lastly, we acknowledge that a long-term horizon of four years to observe inequity trends did not account for various factors that typically change over time (district-level government staff including supervisors and vaccinators, external shocks such as COVID-19, and unprecedented flooding), and these may have confounded the impact of the gender-based inequities over the last four years. However, by also focusing on regions that have persistently demonstrated worse immunization outcomes for females, we showed the deep-seated inequities that continue to persist despite external changes over time.

#### **5. Conclusions**

Our study demonstrated evidence of the gender-based inequities in Sindh Province, Pakistan, over the last four years, with a higher number of males than females being enrolled and immunized. Once enrolled, the coverage rates of females and males were similar, although females tended to be delayed in receiving their vaccinations, as compared to males. We also observed geographical pockets where females continued to fall behind males, year-on-year, reflecting the persistent nature of the inequalities. Our findings have important implications for the inequities among zero-dose children who are more likely to be females. We also demonstrated that certain factors such as maternal literacy, place of residence, and supply-side factors (mode of vaccination delivery, gender of vaccinators), were both a cause and consequence of gender-based inequities. Socio-cultural factors are inextricably linked to characteristics that lead to poor immunization outcomes for females. A deeper qualitative investigation at the sub-national level is needed to uncover the complex dynamics that impact equities in coverage, so that tailored and targeted strategies can be implemented to ensure females and males have the same opportunities to access and benefit from life-saving immunizations.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/vaccines11030685/s1, Figure S1: Location of Urban, Rural, and Remote-Rural UCs in Sindh Province, Pakistan (*n* = 1130); Figure S2: Annual male-to-female ratios among children (>6 weeks) who received Penta-1, Penta-3, and Measles-1 vaccinations in 2019–2022 birth cohorts enrolled in SEIR (1 January 2019–31 December 2022); Figure S3: Annual gender inequality ratios (GIR) among children (>6 weeks) who received Penta-1, Penta-3, and Measles-1 vaccinations in 2019–2022 birth cohorts enrolled in SEIR (1 January 2019–31 December 2022); Figure S4: Geographical distribution of UCs showing annual male-to-female ratios of >1.10 among children (>6 weeks) who received Penta-1, Penta-3, and Measles-1 vaccinations in 2019–2022 birth cohorts enrolled in SEIR (1 January 2019–31 December 2022); Figure S5: Location of UCs with differing sex ratios of vaccinators in Sindh Province, Pakistan (*n* = 3354); Table S1: Gender inequality ratios of 0–23-month-old children in 2019–2022 birth cohorts in SEIR at enrollment and vaccination coverage in Sindh Province, Pakistan, by district (*n* = 6,235,305) (1 January 2019–31 December 2022); Table S2: Gender inequality ratios of 0–23-month-old children in 2019–2022 birth cohorts in SEIR at enrollment and by antigens by maternal literacy levels, geographic location, vaccinators sex ratio, and modality of immunization delivery (*n* = 6,235,305) (1 January 2019–31 December 2022); Table S3: Number of UCs showing persistent gender inequities in vaccination (M:F > 1.05) from 2019–2022 (*n* = 1129);Table S4: Categorization of remote-rural, rural, and urban UCs by sex ratio of vaccinators (*n* = 1130).

**Author Contributions:** Conceptualization, D.A.S., H.S. and S.C.; Data curation, S.I. and M.S.; Formal analysis, S.I., M.S. and M.M.; Funding acquisition, D.A.S. and S.C.; Methodology, D.A.S., H.S. and S.C.; Project administration, V.K.D. and M.T.S.; Resources, V.K.D. and M.S.; Supervision, S.C.; Visualization, D.A.S., S.I., M.S., H.S. and S.C.; Writing—original draft, D.A.S. and M.M.; Writing—review & editing, H.S. and S.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by grants from GiveWell and Gavi, the Vaccine Alliance, to support the Government of Sindh for implementing SEIR. The funders had no role in study design, data collection, data interpretation, or report writing.

**Institutional Review Board Statement:** This analysis was deemed to be exempt by the Institutional Review Board of Interactive Research and Development under 45 CFR 46.101(b) (study number: IRD\_IRB\_2020\_04\_018). The IRB was registered with the U.S. Department of Health and Human Services Office for Human Research Protections with registration number IRB 404 00005148.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data may be obtained from a third party and are not publicly available. The data used for this analysis from Sindh Electronic Immunization Registry (SEIR; also known as Zindagi Mehfooz program) can be requested from the Government of Sindh's Expanded Programme on Immunization (EPI).

**Acknowledgments:** We thank Rozina Feroz Ali for supporting the literature review for the manuscript. We are grateful to the EPI-Sindh and the frontline health workers who vaccinate children and maintain the SEIR, and their supervisors and support staff.

**Conflicts of Interest:** The authors declare that they have no known competing financial interest or personal relationships that could have influenced the work reported in this paper.
