Can vaccination data in electronic systems be trusted? A comparative analysis of the aggregated immunization coverage database and records from electronic health system with consideration of settlement type

reliable, high-quality vaccination data is vital to monitor immunization coverage and identify programmatic gaps. In Ukraine, two independent systems operate in parallel for capturing immunization data: the aggregated, with duplication by paper, database UkrVak system and the electronic health system (eHealth), which accumulates individual-level vaccination records. While eHealth offers real-time data and improved traceability, questions remain about its completeness and usability, especially in light of operational challenges, infrastructure gaps, and lack of dose-level tracking functionality. Objective: to compare routine childhood immunization data captured in UkrVak and eHealth across healthcare facilities serving rural, semi-urban, and urban populations in Lviv and Rivne oblasts, and assess the consistency and reliability of both systems as tools for immunization monitoring. A cross-sectional, descriptive and analytical study was conducted using data from six healthcare facilities, including primary healthcare centers (PHCs) and clinical hospital outpatient departments (CHs). Facilities were stratified by type of settlement served. Four vaccine indicators were analyzed: DTP-3, Polio-3, MMR-1, and MMR-2. UkrVak data (aggregated, as of 9–10 months of 2024) were obtained via official requests from regional disease prevention centers. eHealth data (individual-level, as of October 2024) were extracted from national dashboards. The number of children receiving full vaccine courses in eHealth was estimated by dividing raw entries by 3 (or 2 for MMR-2). For each vaccine and facility, the following were calculated: absolute difference, absolute error, relative error. Pearson correlation coefficient (r) and linear regression were used to assess the statistical relationship between the systems. Results. A strong positive correlation was observed between UkrVak and eHealth data correspondence. (r = 0.987, p < 0.0001). The linear regression model showed that 97.3% of the variation in UkrVak data could be explained by eHealth entries (R² = 0.973, F = 800.4, p < 0.0001). However, systematic discrepancies were identified across vaccine types and facility characteristics. MMR-2 and Polio-3 exhibited the highest mean relative errors (45.52% and 25.7%, respectively), likely due to dose tracking limitations and delays in documentation. A particularly large discrepancy (117 children) was found for MMR-2 in a rural facility, highlighting risks of underreporting in eHealth. When stratified by settlement type, rural facilities had the largest discrepancies (mean relative error: 20.13%) compared to urban (4.56%) and semi-urban (10.00%) settings. Conclusions: While eHealth and UkrVak systems are statistically aligned, they are not functionally interchangeable. Discrepancies vary by vaccine, region, and settlement type, reflecting structural and operational constraints. Transitioning toward a unified electronic immunization registry with standardized data fields, rural infrastructure support, and real-time validation is critical to improve the accuracy and reliability of immunization coverage monitoring in Ukraine.