This chapter discusses the methodology of GBD’s analysis of levels and trends of cause-of-death aliases in the population and demonstrates the specific estimation process of GBD’s cause-of-death analysis. The greatest challenges in cause-of-death analysis include (1) inadequate data sources, poor quality of data for some causes of death, and the resulting methodological problems of rational adjustment, and (2) rational modeling of more than 300 causes of death. This chapter consists of six sections. Section 4.1 briefly reviewed the basic concepts, methods, and tools of cause-of-death analysis and described analyzing strategy of mortality causes in the GBD study. Section 4.2 focused on sources of cause-of-death data and quality assessment strategies in the GBD study. There are three main strategies for cause-of-death data adjustment: to redistribute “Garbage Codes,” to estimate the cause-of-death composition in a population using hospital cause-of-death data, and to harmonize data of cause of death with different versions of ICD codes through mapping. Section 4.3 discusses the five main principles for cause-of-death model development, validation, and reporting proposed by the GBD, and the tool that embodies these principles: the cause-of-death integration model, CODEm. Section 4.4 demonstrates the process of systematic analysis for global, regional, and national age–sex-specific and cause-specific mortality since GBD 2010. Section 4.5 shows the style of results expression of GBD series cause-of-death analysis. Finally, Section 4.6 summarized the characteristics of the GBD cause-of-death analysis, the significance of standardized cause-of-death analyses, and the reasons for differences between the results of the GBD study and the results of the usual national reports, most closely related to the redistribution of garbage codes. In addition, the issues of quality and representativeness of cause-of-death data, as well as strategies for adjusting these data, remain key elements affecting cause-of-death estimates. Further validation is therefore needed to determine whether the GBD findings are closest to the “true picture.” Overall, the GBD’s cause-of-death analysis, both in methodology and the results of the multiple rounds, has made a significant contribution and is a landmark advance in cause-of-death analysis. As a final note, the results of the GBD study are the result of a systematic analysis, not a direct estimate of empirical data. As with each GBD cycle, newly published iteration results have replaced the previous results of the GBD study. Therefore, it is not possible to simply take the data from different iteration cycles and use them as the results of a longitudinal analysis.

Systematic Analysis for Global, Regional and National Cause-Specific Mortality

章节
发表日期：

2025年

查看全文

Gonghuan Yang

摘要

This chapter discusses the methodology of GBD’s analysis of levels and trends of cause-of-death aliases in the population and demonstrates the specific estimation process of GBD’s cause-of-death analysis. The greatest challenges in cause-of-death analysis include (1) inadequate data sources, poor quality of data for some causes of death, and the resulting methodological problems of rational adjustment, and (2) rational modeling of more than 300 causes of death. This chapter consists of six sections. Section 4.1 briefly reviewed the basic concepts, methods, and tools of cause-of-death analysis and described analyzing strategy of mortality causes in the GBD study. Section 4.2 focused on sources of cause-of-death data and quality assessment strategies in the GBD study. There are three main strategies for cause-of-death data adjustment: to redistribute “Garbage Codes,” to estimate the cause-of-death composition in a population using hospital cause-of-death data, and to harmonize data of cause of death with different versions of ICD codes through mapping. Section 4.3 discusses the five main principles for cause-of-death model development, validation, and reporting proposed by the GBD, and the tool that embodies these principles: the cause-of-death integration model, CODEm. Section 4.4 demonstrates the process of systematic analysis for global, regional, and national age–sex-specific and cause-specific mortality since GBD 2010. Section 4.5 shows the style of results expression of GBD series cause-of-death analysis. Finally, Section 4.6 summarized the characteristics of the GBD cause-of-death analysis, the significance of standardized cause-of-death analyses, and the reasons for differences between the results of the GBD study and the results of the usual national reports, most closely related to the redistribution of garbage codes. In addition, the issues of quality and representativeness of cause-of-death data, as well as strategies for adjusting these data, remain key elements affecting cause-of-death estimates. Further validation is therefore needed to determine whether the GBD findings are closest to the “true picture.” Overall, the GBD’s cause-of-death analysis, both in methodology and the results of the multiple rounds, has made a significant contribution and is a landmark advance in cause-of-death analysis. As a final note, the results of the GBD study are the result of a systematic analysis, not a direct estimate of empirical data. As with each GBD cycle, newly published iteration results have replaced the previous results of the GBD study. Therefore, it is not possible to simply take the data from different iteration cycles and use them as the results of a longitudinal analysis.