[Estimates of mortality from external causes in Brazil, 2010-2019: methodology for redistribution of garbage causes]

Insufficient quality of the underlying cause of death in the Brazilian Mortality Information System (SIM), acronym in Portuguese underlists violence and it is necessary to redistribute garbage causes (GC) into valid causes in public health prevention. This study estimated mortality from external causes using the GC redistribution method (GBD-Brazil) and compared it with SIM and estimated data from the GBD-IHME study from 2010 to 2019 in Brazil and its states. The GBD-Brazil GC redistribution algorithm applies previous steps of the GBD-IHME with modifications, using two criteria: proportion of target causes (valid) or reclassification of investigated causes. The SIM data were not adjusted. Standardized rates by direct method, local regression in the time series, and the ratio of the GBD-Brazil and SIM rates are used as correction factors for traffic injuries, falls, suicides, and homicides. Brazil recorded 1.34 million deaths with valid external causes and 171,700 CG in ten years. The redistribution of GC from GBD-Brazil increased valid causes by 12.2%, and the trend curve of the rates was similar to that found with SIM data, but diverged from each other and from the GBD-IHME in states in the North and Northeast regions. The GBD-Brazil estimates changed the pattern of external causes in the states, applying greater corrections to falls in the states of the North and Northeast and homicides in the other states. The GBD-Brazil method can be used in the analysis of violent deaths because it seeks greater methodological simplicity, which guarantees both replication by public managers and consistency of the estimated data, considering the local composition of the data in the redistribution process.