Death rates for community-acquired pneumonia based on relatively small-scale, published studies tend to exceed 15% to 20%. This study reexamined these estimates by using very large, population-based databases.
Death rates from 1993 associated with community-acquired pneumonia were reexamined with hospital discharge data from all of Washington, Illinois, and Florida.
These death rates were substantially lower (7.0%, 8.1%, and 9.7%, respectively) than what appears in the literature.
Significant risk factors for dying were being 65 years of age or older (odds ratio [OR]=2.9), being positive for human immunodeficiency virus (OR=2.9), and having a high severity of illness (OR=7.1).
Sampling bias associated with selection for hospital admissions explain the discrepancy between previous and this study's results.
Mots-clés Pascal : Pneumonie, Hospitalisation, Mortalité, Homme, Erreur estimation, Epidémiologie, Base donnée, Washington, Etats Unis, Amérique du Nord, Amérique, Floride, Illinois, Appareil respiratoire pathologie, Poumon pathologie, Infection communautaire
Mots-clés Pascal anglais : Pneumonia, Hospitalization, Mortality, Human, Estimation error, Epidemiology, Database, Washington, United States, North America, America, Florida, Illinois, Respiratory disease, Lung disease, Community infection
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 96-0396547
Code Inist : 002B05B02E. Création : 10/04/1997.