Improvement of the unspecified external causes classification based on the investigation of death in Brazil in 2017.

Background: Unspecified causes of death are among the traditional indicators of quality of information.
Objective: To verify the performance of the 60 cities in the Data for Health Initiative project and to analyze the reclassification of unspecified external causes of death (UEC).
Methods: Using the 2017 records from the Mortality Information System, the proportion and percent change in UEC were compared after investigation between project cities and other cities, and the percent of reclassification to specific external causes was calculated.
Results: The project cities comprised 52% (n = 11,759) of the total UEC in Brazil, of which 64.5% were reclassified after investigation, whereas the other cities reclassified 31% of UEC. Results were similar for men, youth, blacks, metropolitan cities, the Southeast region, and deaths attested by forensic institutes. In the project cities, pedestrian traffic accidents were external causes with greater reclassification. In men, the UEC was reclassified to homicides (23.8%) and accident of terrestrial transportation (ATT) (11.1%), with motorcyclists (4.4%) and pedestrians (4.3%) being the most prominent. In women, these causes were changed to other accident causes (20.8%), ATT (10.6%) and homicides (7.9%). UEC changed to ATT (18.3%) in the age groups of 0-14 years old and to homicides (32.5%) in the age groups of 15-44 years.
Conclusion: The project cities obtained better results after investigation of UEC, enabling analysis of the reclassification to specific causes by sex and age groups.