Comparative analysis of completeness of death registration, adult mortality and life expectancy at birth in Brazil at the subnational level.

BACKGROUND: Estimates of completeness of death registration are crucial to produce estimates of life tables and population projections and to estimate the burden of disease. They are an important step in assessing the quality of data. In the case of subnational data analysis in Brazil, it is important to consider spatial and temporal variation in the quality of mortality data. There are two main sources of data quality evaluation in Brazil, but there are few comparative studies and how they evolve over time. The aim of the paper is to compare and discuss alternative estimates of completeness of death registration, adult mortality (45q15) and life expectancy estimates produced by the National Statistics Office (IBGE), Institute for Health Metrics and Evaluation (IHME), and estimates presented in Queiroz et al. (2017) and Schmertmann and Gonzaga (2018), for 1980 and 2010. METHODS: We provide a descriptive and comparative analysis of aforementioned estimates from four (4) sources of estimates at subnational level (26 states and one Federal District) in Brazil from two different points in time. RESULTS: We found significant differences in estimates that affect both levels and trends of completeness of adult mortality in Brazil and states. IHME and Queiroz et al. (2017) estimates converge by 2010, but there are large differences when compared to estimates from the National Statistics Office (IBGE). Larger differences are observed for less developed states. We have showed that the quality of mortality data in Brazil has improved steadily overtime, but with large regional variations. However, we have observed that IBGE estimates show the lowest levels of completeness for the Northern of the country compared to other estimates. Choice of methods and approaches might lead to very unexpected results. CONCLUSION: We produced a detailed comparative analysis of estimates of completeness of death registration from different sources and discuss the main results and possible explanations for these differences. We have also showed that new improved methods are still needed to study adult mortality in less developed countries and at a subnational level. More comparative studies are important in order to improve quality of estimates in Brazil.

Cause-specific mortality for 249 causes in Brazil and states during 1990-2015: a systematic analysis for the global burden of disease study 2015.

BACKGROUND: Reliable data on cause of death (COD) are fundamental for planning and resource allocation priorities. We used GBD 2015 estimates to examine levels and trends for the leading causes of death in Brazil from 1990 to 2015. METHODS: We describe the main analytical approaches focused on both overall and specific causes of death for Brazil and Brazilian states. RESULTS: There was an overall improvement in life expectancy at birth from 1990 to 2015, but with important heterogeneity among states. Reduced mortality due to diarrhea, lower respiratory infections, and other infectious diseases contributed the most for increasing life expectancy in most states from the North and Northeast regions. Reduced mortality due to cardiovascular diseases was the highest contributor in the South, Southeast, and Center West regions. However, among men, intentional injuries reduced life expectancy in 17 out of 27 states. Although age-standardized rates due to ischemic heart disease (IHD) and cerebrovascular disease declined over time, these remained the leading CODs in the country and states. In contrast, leading causes of premature mortality changed substantially - e.g., diarrheal diseases moved from 1st to 13th and then the 36th position in 1990, 2005, and 2015, respectively, while violence moved from 7th to 1st and to 2nd. Overall, the total age-standardized years of life lost (YLL) rate was reduced from 1990 to 2015, bringing the burden of premature deaths closer to expected rates given the country's Socio-demographic Index (SDI). In 1990, IHD, stroke, diarrhea, neonatal preterm birth complications, road injury, and violence had ratios higher than the expected, while in 2015 only violence was higher, overall and in all states, according to the SDI. CONCLUSIONS: A widespread reduction of mortality levels occurred in Brazil from 1990 to 2015, particularly among children under 5 years old. Major shifts in mortality rates took place among communicable, maternal, neonatal, and nutritional disorders. The mortality profile has shifted to older ages with increases in non-communicable diseases as well as premature deaths due to violence. Policymakers should address health interventions accordingly.

Curricular transformations in medical school: an exploration of the promotion of change in Brazil.

CONTEXT: Curricular transformation is complex and involves many interrelated and influential factors. The idiosyncratic nature of the time-frame in which curricular change takes place makes it difficult to study these factors systematically. In 2002, the Brazilian government established the Programme of Incentives for Curricular Changes in Medical Schools (PROMED). Nineteen schools participated in the programme and were given resources to finance curricular reforms in an effort to align their course structures with the tenets of the government proposal. This study analyses reforms in coursework among the schools in this cohort in an effort to better understand the impact of such incentive programmes and factors that might influence the degree of impact across institutions. METHODS: We compared data on the schools before and after their participation in PROMED. To facilitate comparison, we used a scheme of axes and vectors to classify the schools according to the profundity of the curricular changes, ranging from the most conservative to the most innovative. The data used for the classification were obtained through document analysis, interviews and focus group discussions. RESULTS: Different trends were observed for each axis. Important changes were noticed in the pedagogic approach axis, particularly in terms of pedagogic changes, which called for the adoption of active teaching and learning methods. The practice scenarios axis also underwent considerable changes, specifically in terms of primary health care. The vector related to production of knowledge pertaining to health system needs showed fewer changes; none of the schools reached stage 3 (effective fulfilment of an educational innovation). CONCLUSIONS: The PROMED initiative provided considerable support for implementing and consolidating curricular reforms that placed greater emphasis on the needs of society and the health care system. The different trends observed revealed the complexity behind curricular transformation and highlighted the need for the collective construction of curricula with the participation of all groups involved.

Consistent declining trends in stroke mortality in Brazil: mission accomplished?

Stroke mortality rates are declining in Brazil, but diferences among regions need to be better investigated. The age-adjusted stroke mortality trends among adults (30-69 years-old) from Brazilian regions were studied between 1996 and 2011. Method Data were analyzed after: 1) reallocation of deaths with non-registered sex or age; 2) redistribution of garbage codes and 3) underreporting correction. A linear regression model with autoregressive errors and a state space model were fitted to the data, aiming the estimation of annual trends at every point in time. Results Although there were high values, a steady decrease of rates was observed. The decreasing trends among all regions were statistically significant, with higher values of decline among the Northeast and Northern regions, where rates were the highest. Conclusion Standardized methodology use is mandatory for correct interpretation of mortality estimates. Although declining, rates are still extremely high and efforts must be made towards prevention of stroke incidence, reduction of case-fatality rates and prevention of sequelae.

Consistent declining trends in stroke mortality in Brazil: mission accomplished? / Tendências decrescentes constantes na mortalidade por AVE no Brasil: missão cumprida?

ABSTRACT Stroke mortality rates are declining in Brazil, but diferences among regions need to be better investigated. The age-adjusted stroke mortality trends among adults (30-69 years-old) from Brazilian regions were studied between 1996 and 2011. Method Data were analyzed after: 1) reallocation of deaths with non-registered sex or age; 2) redistribution of garbage codes and 3) underreporting correction. A linear regression model with autoregressive errors and a state space model were fitted to the data, aiming the estimation of annual trends at every point in time. Results Although there were high values, a steady decrease of rates was observed. The decreasing trends among all regions were statistically significant, with higher values of decline among the Northeast and Northern regions, where rates were the highest. Conclusion Standardized methodology use is mandatory for correct interpretation of mortality estimates. Although declining, rates are still extremely high and efforts must be made towards prevention of stroke incidence, reduction of case-fatality rates and prevention of sequelae.

RESUMO Objetivos As taxas de mortalidade por acidente vascular encefálico (AVE) estão em declínio no Brasil, mas diferenças entre as regiões precisam ser melhor investigadas. Foram analisadas as tendências de mortalidade por AVE ajustadas por idade em adultos (30-69 anos) de regiões do Brasil, entre 1996 e 2011. Método Análise realizada após realocação dos óbitos sem registro de sexo ou idade; redistribuição de códigos mal definidos e correção do subregistro. Modelos de regressão linear e de espaço de estados foram utilizados visando estimativas de tendências anuais em todos pontos no tempo. Resultados Observou-se redução de cerca de 50% nas taxas de mortalidade em todas as regiões, para homens e mulheres, com maiores declínios nas regiões Norte e Nordeste, onde as taxas eram as mais elevadas. Conclusão Usar metodologia padronizada é fundamental para interpretação correta das estimativas de mortalidade. Esforços devem ser feitos para prevenção da incidência de AVE, redução das taxas de letalidade e prevenção de sequelas.