The accuracy and consistency of public health data in Brazilian information systems: identification of gaps and challenges to be faced in a municipality in the Amazon region.

Background: In Brazil, health conditions of public importance are notified by municipal health departments with a standardized flow using the Notifiable Health Conditions Information System (Sinan). This information goes through a process of consolidation and transfer to DataSUS, the national system that centralizes health information in Brazil. This study assessed whether there are quantitative differences between notified health conditions through the municipal system (Sinan) and those consolidated in the national system (DataSUS).

Methods: This study was based on the municipality of Bragança, located in the eastern Amazon, which plays a strategic role due to its high annual number of notifications. To identify differences between the systems, we used data provided by Sinan and retrieved from DataSUS from 2019 to 2023. We tested the absolute loss in the number of notifications across years and health conditions between the two systems. Ethical approval was not required due to the anonymous nature of the data.

Results: Of the 19 health conditions identified and analyzed, 15 showed decreases between the systems, with losses reaching up to 91%. The largest discrepancies were observed for AIDS, syphilis in pregnant women, and dengue. Over the years, the loss was consistent, averaging approximately 42%.

Conclusion: The differences observed between the two systems may have direct implications for the design, planning, and implementation of public health policies. Reducing these gaps urgently requires strategies such as the training of healthcare professionals, the revision of data flow processes, and investments in technologies that support system integration.