Geospatial Settlement Data and Immunization Reach in Polio SIAs: Evidence from Axis of Intractable Transmission States in Northern Nigeria

Global polio eradication situation in Nigeria

Nigeria, the seventh most populous country in the world, was declared free of WPV in 2020 but remains among the largest sources of circulating vaccine-derived poliovirus type 2 (cVDPV2) globally.⁶ Much of the risk for cVDPV outbreaks can be linked to a combination of inaccessibility, insecurity, a high concentration of zero-dose and under-immunized children, and population displacement.⁷

Vaccine delivery in Nigeria is managed by the National Program on Immunization (NPI), which uses vaccines approved by the WHO.⁸ Six vaccines have been used to stop polio transmission and eliminate wild poliovirus (WPV) in Nigeria.⁹ Until 2016, the trivalent oral polio vaccine (tOPV), containing Sabin strains of all three poliovirus serotypes, was in use.¹⁰ In 2015, an inactivated polio vaccine (IPV) dose was added at 14 weeks to the routine schedule. In April 2016, tOPV was replaced by the bivalent oral polio vaccine (bOPV), which included only Sabin serotypes 1 and 3. Monovalent oral polio vaccines for types 1, 2, and 3 were also introduced to target each serotype specifically.¹¹ Nigeria continued to face cVDPV2 outbreaks even after administering mOPV2 from 2018 to 2022.⁹ In 2021, a second IPV dose was added at six weeks of age, and a new oral poliovirus vaccine, type 2 (nOPV2), was introduced.¹²

In 2023, the National Primary Health Care Development Agency (NPHCDA), in partnership with the Global Polio Eradication Initiative (GPEI) and other stakeholders, prioritized eliminating all forms of poliovirus transmission worldwide.¹³ The northern part of Nigeria was identified as a critical region, focusing on specific local government areas (LGAs) with the highest number of unimmunized children.¹⁴ To support the polio response and stop the spread of cVDPV2, eHealth Africa received grants from the Gates Foundation in 2024 to implement the Geospatial Tracking System project.

This retrospective analysis assessed whether geospatial investments contributed to the improvement of immunization coverage and operational quality during Supplementary Immunization Activities (SIAs) from April 2024 to June 2025 across four axes of Intractable Transmission States (AITs) in northern Nigeria: Kebbi, Sokoto, Katsina, and Zamfara. Overall, this analysis highlights the importance of moving from hand-drawn maps to georeferenced ones, and it also incorporates geospatial, tracking, and immunization indicators. An important limitation is that the analysis relies on aggregated state-level data and cannot assess intra-state variations at the LGA or ward level.

Nigeria was certified WPV-free in 2020, but remains one of the largest contributors to cVDPV cases globally.¹⁵ Between 2018 and 2024, 53 countries reported cVDPV outbreaks, including over 20 African nations in 2024 alone.¹⁶ This resurgence highlights that eradication hinges not only on vaccine availability but also on equitable, sustained, and effective immunization, especially in hard-to-reach and underserved communities.¹⁷ Achieving and sustaining global polio eradication is crucial for global health security, preventing cross-border transmission, and safeguarding all populations from this debilitating disease. The post-certification period presents unique challenges as countries experience programmatic complacency and reduced political attention to polio immunization.^16,18 Nigeria has made significant efforts but remains endemic for wild poliovirus (WPV) as a result of several factors, mainly because of poor quality and coverage of the SIAs.¹⁹ Another hindrance was the inability to identify areas that were missed by timely intervention with vaccination to provide the required herd immunity to break the polio transmission.²⁰

Project context

To support the polio response, eHealth Africa received grants from the Gates Foundation in 2024 to implement the Geospatial Tracking System project, intended to strengthen operational efforts to interrupt and ultimately stop the spread of cVDPV2. The project aimed to improve geographical coverage of settlements and reduce the number of missed areas; increase access to and use of GTS data through an up-to-date geodatabase to support government and partner decision-making; reduce missed children; strengthen vaccination team performance and accountability; and improve reach to chronically missed settlements while continuously updating the Polio Master List of Settlements for northern states. Core activities included tracking vaccination teams with the GTS across 120 LGAs, supporting 19 states and the FCT in updating the Master List of Settlements (MLOS), supporting the development of under-five line-lists across seven states, and conducting capacity-building sessions in 19 states and the FCT. Powered by Novel-T, the GTS provided organizations with visibility into field activities by enabling geospatial tracking of field teams, remote retrieval of movement data, and robust analysis of tracking outputs to inform operational decisions. This assessment aimed to determine the geospatial completeness of settlement mapping and the extent to which the completeness of the geospatial Master List of Settlements datasets improves coverage rates of polio immunization campaigns in the axis of intractable transmission in Northern Nigeria.

Methodology

This retrospective analysis utilized programmatic data collected during Supplementary Immunization Activities (SIAs) conducted between April 2024 and June 2025 across four Axis of Intractable Transmission States (AITs) in Northern Nigeria which are Kebbi, Sokoto, Katsina, and Zamfara states. The analysis focused on settlement-level geolocation quality, operational indicators, and vaccination outcomes.

The campaign datasets generated during this timeframe were analyzed to evaluate the geolocation quality based on the completeness, accuracy, consistency of georeferenced settlement data and determine the effectiveness of settlement-level tracking and explore the extent to which the geospatial completeness of the Master List of Settlements (MLoS) influences immunization coverage performance. Vaccination outcomes were derived from administrative SIA coverage data. Operational indicators included measures of settlement reach, team performance, and microplanning completeness. Between April 2024 and June 2025, a total of nine SIAs were conducted across the AIT states of Kebbi, Sokoto, Katsina, and Zamfara. These campaigns followed an intensive outbreak-response schedule, with rounds implemented approximately every 4-8 weeks. The majority were sub-national immunization days (SNIDs), complemented by periodic national immunization days (NIDs) and interspersed in-between round activities (IBRAs) to reach missed and underserved populations.

Analytical approaches included stratified comparisons across levels of geolocation quality. While stratification was used to reduce heterogeneity across settings, the analysis did not fully account for all potential confounders.

Data sources

This study sourced data from the project monitoring and evaluation performance tracking database, including:

Measures

Data analysis

Descriptive analyses

In this study, we applied the state-level means and distributions to summarize geospatial completeness, settlement tracking, and vaccination performance across campaign rounds. Temporal trends were inspected to identify improvements in operational performance.

Association between MLoS completeness and tracking performance

To assess whether settlement geospatial completeness was associated with tracking coverage, mean geospatial coverage and mean tracking coverage were compared across states. Given the limited number of states (n = 4), emphasis was placed on directionality and consistency rather than statistical inference. The analysis evaluated whether more complete geospatial readiness translated into more stable tracking and operational performance.

Relationship between settlement tracking and vaccination outputs

Pearson correlation coefficients were computed between the number of settlements tracked and the number of children vaccinated, both across the pooled dataset and within individual states. Subnational stratification was required because terrain, population mobility, service access, and risk patterns differ across states, potentially influencing the relationship between geographic reach and population yield.

Temporal analysis of operational learning

To assess performance change over time, campaign rounds were indexed chronologically within each state. Tracking coverage was assessed across early rounds (1–3) and late rounds (7–9). Improvements were interpreted as evidence of operational adaptation, including settlement list refinement and improved supervisory routines.

Data treatment and quality assurance

The missing data were handled through within-state forward fill for categorical descriptors to maintain longitudinal continuity. No imputation was performed for population or immunization variables to avoid inflating performance metrics. The state-level differences were preserved. There was no pooling or normalization that was performed beyond descriptive averages, because heterogeneity was intrinsic to the research question.

Statistical interpretation and analytical considerations

The analytical approach used in this study quantifies the association between settlement-level geolocation quality and vaccination outcomes during Supplementary Immunization Activities (SIAs). The model is designed to estimate the strength and direction of this relationship while accounting for selected observable factors through stratification and/or adjustment. However, the model does not establish causality. The findings should therefore be interpreted as evidence of correlation rather than proof that improvements in geolocation quality directly led to improved vaccination performance. The observed association may reflect underlying contextual conditions that simultaneously influence both variables.

Geospatial completeness of settlement mapping

Across the study period, the completeness of geospatial coordinates in the Master List of Settlements (MLoS) was consistently high in all four high-risk northern states ( Table 1). Kebbi state demonstrated full geocoordinate completeness, achieving almost 100% mapping coverage across all campaign rounds. Katsina and Sokoto states followed closely, with mean completeness of 97.3% and 96.7% respectively, while Zamfara had the lowest level of mapping maturity at 92.1%.

Results of settlement tracking

The analysis of the tracking data shows variation across campaign rounds. In Sokoto state, settlements tracked increased from 7,789 in March 2024, when the intervention began, to 10,572 in October 2025, with a mean of 7,982 settlements tracked per campaign. Katsina consistently had the highest settlement tracked, with an average of 12,870 settlements per round. In Kebbi, the settlements tracked were above 13,000 in most rounds, with an average of 9,987 settlements tracked per campaign. Zamfara recorded the lowest tracking data, with an average of 7,072 settlements tracked per campaign ( Table 2, Figure 1).

Figure 1. Map of AITS, including the settlement tracked.

Legend: The figure shows the countries where GTS was implemented during campaign rounds.

The bivariate exploratory analysis shows that the relationship between geospatial completeness and vaccination tracking performance was evident: states with more fully geolocated settlements had higher tracking coverage during campaign delivery. Kebbi, with full geospatial mapping, maintained a mean tracking coverage of 93%, whereas Zamfara, with the lowest mapping completeness, achieved only 84.8% ( Figure 2). This pattern shows that the expansion of geospatial enumeration capabilities is directly associated with the reliability of settlement-level campaign operations. In practical terms, the presence of validated coordinates appears to reduce settlement invisibility, improve microplanning accuracy, and support stronger supervisory reach during implementation.

Figure 2. The relationship between geospatial completeness and vaccination tracking performance, using the average values across the four states.

Legend: Geospatial completeness and vaccination tracking performance, using the average values across the four states.

Relationship between settlement tracking and immunization results

The relationship between the number of settlements tracked during campaigns and the number of children vaccinated varied across states and reflected the campaign operations and their delivery. When all campaign rounds were pooled, settlement tracking was moderately positively associated with immunization results. Campaigns covering more settlements tended to vaccinate more children, indicating that geographic expansion of field activities was generally linked to wider children population reach. The correlation between settlements tracked and children immunized was 0.5560589 (approximately 0.56) ( Figure 3). This relationship was strongest in Kebbi and Sokoto, where higher tracking results had a relationship with increased immunization numbers.

Figure 3. A positive correlation of 0.56 shows that tracking more settlements leads to improvement in immunization coverage.

Legend: Positive correlation of tracked settlements and their immunization coverage.

In contrast, Katsina displayed a near-zero association between tracking and vaccinations ( Table 3). The number of children vaccinated remained high regardless of how many settlements were tracked, suggesting that demand, campaign saturation, or historical experience may enable the system to reach children, in addition to vaccination tracking. Zamfara demonstrated a negative association between these indicators, as campaigns that expanded settlement coverage tended to vaccinate fewer children, reflecting a dilution effect, where widening the campaign footprint resulted in lower yield per settlement, or where smaller or more remote settlements were added late in the process.

Operational learning across campaign rounds

Tracking coverage showed evidence of operational improvement across repeated campaign cycles. When campaigns were sequenced chronologically from the first to the ninth round in each state, coverage tracking rose consistently over time in Sokoto, Katsina, and Zamfara. Sokoto improved from an average of 86.3% in its first three rounds to 94.7% in its most recent three, reflecting an 8.4 percentage point increase. Katsina and Zamfara demonstrated similar upward trajectories, each recording improvements of approximately 10 percentage points between early and late rounds. This progressive enhancement indicates campaign-level learning, likely driven by improved settlement lists, supervisory discipline, and frontline team familiarity with geospatial tools. Kebbi displayed minimal temporal change, though for positive reasons. The state began with performance at saturation levels, averaging 93.3% tracking coverage in its early rounds and maintaining 94% in later rounds ( Figure 4).

Figure 4. Trends in vaccination coverage across campaign rounds 2024-2025.

Legend: Shows trends in vaccination coverage across campaign rounds 2024-2025 in the four states.

Strengths and limitations

This study has several strengths. It leverages real-world operational data across multiple SIA rounds and states, allowing examination of temporal trends and cross-state heterogeneity. It also integrates geospatial, tracking, and immunization indicators, highlighting relational dynamics rather than focusing on a single metric such as administrative coverage. The findings are grounded in a robust body of evidence on GIS-supported microplanning and target population estimation challenges, and they speak directly to programme decision-makers tasked with sustaining high-quality campaigns in the post-certification era. However, this study has several limitations that should be considered when interpreting the findings. First, the analysis is based on observational data, which limits the ability to establish causal relationships between settlement-level geolocation quality and vaccination outcomes. While the study demonstrates an association between improved geolocation and higher vaccination performance, this relationship should be interpreted with caution, as it may be influenced by unmeasured confounding factors.

Geographic accessibility, security conditions, and social acceptance of immunization activities are likely to simultaneously influence both the quality of geolocation data and vaccination coverage. Settlements that are easier to access, both physically and socially, are more likely to have accurate geospatial data and also to achieve higher vaccination coverage. Although the analysis attempts to control for some of these factors through stratification, residual confounding is likely, and the observed associations may therefore overestimate the independent effect of geolocation quality.

Second, the study may be affected by measurement limitations related to both geolocation quality and vaccination outcomes. Geolocation accuracy may vary depending on the tools used, the skill level of field teams, and environmental factors such as terrain or network connectivity. Similarly, vaccination coverage estimates may be subject to reporting bias, including over-reporting or under-reporting during Supplementary Immunization Activities (SIAs), especially in settings where administrative data systems are relied upon.

Third, the analysis does not fully capture important contextual and social determinants that influence immunization performance. Factors such as community trust, political dynamics, caregiver perceptions, competing health priorities, and local leadership engagement play a critical role in vaccination uptake but are not directly measured in the dataset. In regions where communities may express frustration with the prioritization of polio over other high-burden diseases such as malaria, these dynamics can significantly affect campaign performance independent of geospatial data quality.

Fourth, security constraints and population mobility in the study areas, particularly in states such as Katsina, Sokoto, Zamfara, and Kebbi, may affect both data quality and program implementation. Insecurity can restrict access to certain settlements, resulting in incomplete geolocation data and missed vaccination opportunities. Additionally, migratory and nomadic populations may not be fully captured in static geospatial datasets, leading to potential underestimation of target populations and gaps in coverage.

Fifth, heterogeneity across states and local government areas is not fully explored in the analysis. Differences in infrastructure, terrain, governance, health system capacity, and implementation practices may influence both geolocation and vaccination outcomes in ways that are not adequately accounted for. As a result, findings from one context may not be directly generalizable to others.

Sixth, the study focuses primarily on technical and operational aspects of geospatial implementation and may underrepresent broader programmatic factors. Issues such as workforce capacity, training quality, supervision, and resource availability can significantly influence both data quality and vaccination performance but are not explicitly examined in the analysis.

Seventh, the temporal scope of the data may limit the ability to assess longer-term impacts of geospatial interventions. The study captures outcomes within a specific campaign period and may not reflect sustained improvements or changes over time. Longitudinal analyses would be required to better understand the durability of observed effects.

Finally, while the study contributes to the growing evidence base on geospatial tools in immunization programs, it should be viewed as part of a broader body of work rather than as definitive evidence of impact. Geospatial interventions operate within complex health systems, and their effectiveness depends on how they are integrated with other strategies addressing social, political, and operational challenges. Future research and programme priorities.

Future work should extend this analysis in three directions. First, linking settlement-level geospatial data, GPS tracks, and independent coverage surveys would allow a better assessment of zero-dose and under-immunized populations. Second, linking GPS data tracks with walk-through population estimates to forecast resources required at the ward level. Third, identify the prevailing factors that may have impacted the increase in immunization coverage even when the number of tracked settlements fluctuated, suggesting a level of saturation.

Future research and programme priorities

Future work should extend this analysis in three directions. First, linking settlement-level geospatial data, GPS tracks, and independent coverage surveys would allow a better assessment of zero-dose and under-immunized populations. Second, linking GPS data tracks with walk-through population estimates to forecast resources required at the ward level. Third, identify the prevailing factors that may have impacted the increase in immunization coverage even when the number of tracked settlements fluctuated, suggesting a level of saturation.