Cargando…

Comparing U.S. Injury Death Estimates from GBD 2015 and CDC WONDER

Objective: The purpose of the present study was to examine consistency in injury death statistics from the United States CDC Wide-ranging Online Data for Epidemiologic Research (CDC WONDER) with those from GBD 2015 estimates. Methods: Differences in deaths and the percent difference in deaths betwee...

Descripción completa

Detalles Bibliográficos
Autores principales: Wu, Yue, Cheng, Xunjie, Ning, Peishan, Cheng, Peixia, Schwebel, David C., Hu, Guoqing
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2018
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5800186/
https://www.ncbi.nlm.nih.gov/pubmed/29316676
http://dx.doi.org/10.3390/ijerph15010087
Descripción
Sumario:Objective: The purpose of the present study was to examine consistency in injury death statistics from the United States CDC Wide-ranging Online Data for Epidemiologic Research (CDC WONDER) with those from GBD 2015 estimates. Methods: Differences in deaths and the percent difference in deaths between GBD 2015 and CDC WONDER were assessed, as were changes in deaths between 2000 and 2015 for the two datasets. Results: From 2000 to 2015, GBD 2015 estimates for the U.S. injury deaths were somewhat higher than CDC WONDER estimates in most categories, with the exception of deaths from falls and from forces of nature, war, and legal intervention in 2015. Encouragingly, the difference in total injury deaths between the two data sources narrowed from 44,897 (percent difference in deaths = 41%) in 2000 to 34,877 (percent difference in deaths = 25%) in 2015. Differences in deaths and percent difference in deaths between the two data sources varied greatly across injury cause and over the assessment years. The two data sources present consistent changes in direction from 2000 to 2015 for all injury causes except for forces of nature, war, and legal intervention, and adverse effects of medical treatment. Conclusions: We conclude that further studies are warranted to interpret the inconsistencies in data and develop estimation approaches that increase the consistency of the two datasets.