Estimation and projection of probabilistic age- and sex-specific mortality rates across Brazilian municipalities between 2010 and 2030

Background: Small area age- and sex-specific mortality rates are useful measures for population projections, health, economic, and social planning. Mortality rate estimation in small areas can be difficult due the low number of events/exposure. If a country’s mortality registration has problems, such as incomplete information, then estimating mortality rates can be even more difficult. Previous studies in Brazil have combined demographic and statistical methods to overcome these issues. These approaches depend on a gold standard for age-specific mortality rates and do not estimate uncertainties. We estimated age- and sex-specific mortality rates for all 5,565 Brazilian municipalities in 2010, and forecasted mortality rates between 2010 and 2030. Methods: We used the Tool for Projecting Age-Specific Rates Using Linear Splines (TOPALS) and a Bayesian model to estimate age- and sex-specific mortality rates in all Brazilian municipalities in 2010 while incorporating two types of uncertainties: low exposure and incomplete coverage of death counts. We adapted the Lee-Carter model to forecast age- and sex-specific mortality rates between 2010 and 2030 for all municipalities. Results: The proposed methodology was robust in adjusting for the mortality age profile and in estimating mortality rate uncertainties at the municipal level. The forecasted mortality rates indicated a convergence in life expectancy at birth, and variability of age at death across Brazil’s municipalities, with a persistent sex differential. Conclusion: We estimated and forecasted mortality rates in small areas with limited and incomplete death counts, and high mortality heterogeneity. The methodological approach applied could be useful for countries with death data quality problems similar to Brazil. Our results incorporated the main sources of uncertainty in estimating age- and sex-specific mortality rates and could be used as an important input for policy planning at the municipal level.