Data quality in tuberculosis: the case study of two ambulatories in the state of São Paulo, Brazil

Abstract Tuberculosis is the second leading responsible cause of death from infectious diseases. Tuberculosis effective control and related activities depends on the use of different systems that aim to assist and monitor patients through managing their health data, and facilitating the activities of several health professionals. However, we noticed disparities in the information provided by some systems, which can negatively impact planning and decision-making. The study is defined as a quantitative and descriptive study of patients’ data during the treatment of tuberculosis in two different ambulatories. The data was collected from four different sources, including three information systems and the local patient archive in the ambulatories. Collected data was cleansed and standardized for semantic and structure, which allowed data comparison and analysis for its reliability and completeness. Low reliability scores are due to the absence of a semantic standard and the careless validation on recording of data by the professionals. Therefore, this study was able to effectively detect inconsistencies between the different data sources, stressing the need of health standards for data consistency, interoperability, and promoting data quality.