Objective: To analyze the reliability of records held on the National Immunization Program Information System (SI-PNI) in a subsample of children included in the national vaccination coverage survey in Brazilian state capitals and Federal District in 2020.
Methods: This was a study of agreement between data recorded on vaccination cards (doses and dates) and on the SI-PNI for 4050 children with full coverage at 24 months.
Results: Data on 3587 children were held on the SI-PNI, with losses of 11% (95%CI: 10;12). Total agreement between doses and dates in the two sources was 86% (95%CI: 86;87), however taking each dose and vaccine individually, variation was greater, with 32% of data in only one source.
Conclusion: Part of the information was not recorded, but the discrepancy can be considered small. Nonetheless, underrecording of doses and children can compromise vaccination coverage estimates, altering the numerator and denominator data.
Main results: Subsample of 4,050 children, among those completing the full schedule at 24 months studied in the national survey, 11% had not been recorded on the SI-PNI, 32% had unrecorded doses (doses or dates) and there was 8% disagreement between vaccination cards and SI-PNI records.
Implications for services: Recognizing the difficulties faced by the SI-PNI and the discrepancies between sources is essential for adopting initiatives to improve data quality, so as to avoid inaccurate estimates of childhood vaccination coverage.
Perspectives: This study is expected to contribute to improving the quality of records and the usability of data for monitoring vaccination coverage of the immunization program from the local to the national level.
Objetivo: Evaluar la confiabilidad de los registros del Sistema de Información del Programa Nacional de Vacunación (SI-PNI), en una submuestra de niños incluidos en la encuesta nacional de cobertura vacunal en capitales de Brasil y Distrito Federal, en 2020 .
Método: Estudio de concordancia entre datos en carné vacunal (dosis y fechas) y en el SI-PNI para 4.050 niños con pauta completa a los 24 meses.
Resultados: Se ubicaron 3.587 niños en el SI-PNI con un 11% (IC95% 10,0;12,0) de pérdidas. La concordancia total entre dosis y fechas en las dos fuentes fue del 86% (IC95% 86,0;87,0), pero por dosis y vacuna la variación fue mayor, con 32% de datos en una única fuente.
Conclusión: Parte de la información no se ha registrado, pero para los datos existentes en ambas fuentes, la discrepancia puede considerarse pequeña. El subregistro de dosis y de niños puede comprometer las estimaciones de cobertura vacunal.