Citation

Jansson B. Am. J. Public Health 2005; 95(8): 1305-1306.

Affiliation

Department of Public Health Sciences, Division of Social Medicine, Norrbacka Bldg, Karolinska Institutet, Stockholm, Sweden

Copyright

(Copyright © 2005, American Public Health Association)

DOI

10.2105/AJPH.2005.067553

PMID

16006408

PMCID

PMC1449353

Abstract

Research Letter Summary:

When assigning priorities, it is important to have correct information on the scope and magnitude of a health problem. In the March issue of the American Journal of Public Health, Kannus et al.(doi: 10.2105/AJPH.2004.047779) reported an increasing trend in fall-induced injury deaths among the elderly. This is a well-known problem in most industrialized countries.

The quality of cause-of-death statistics is regularly reviewed, but most studies compare death certificates with clinical records or autopsy reports. Only a few studies have focused on changes in registration practice, mainly the changeover from revisions of the International Classification of Diseases (ICD). Reporting errors due to coding errors at national statistics agencies are seldom considered, although a recent study by Graber et al. focused on national adaptations of coding instructions.

While epidemiologists are often aware of artifacts due to the implementation of a new version of the ICD, it seems to be less well known that important changes in coding practice may take place at other times. The selection of underlying cause of death is not always the result of strict application of ICD selection rules, but is often more a reflection of what coders feel is the most important condition to be reported on a certificate. Because of differences in the application of ICD selection rules, a condition that is selected as the underlying cause of death in one country might have been regarded as a contributory cause in another country. Therefore, both underlying and contributing causes of death should be analyzed.

When analyzing fall injury mortality, at least three diagnoses are important: epilepsy, pneumonia and fracture of the neck or other parts of the femur. Differences in proportions between underlying and contributing causes have been shown to be as much as 67% for pneumonia and 38% for femur fractures. The problem occurs when a patient dies of multiple and late complications of an injury. This phenomenon,�  will result in a serious underestimation of fall injury mortality.

In conclusion, both underlying and contributing cause-of-death statistics should be considered in any analysis of mortality trends. However, it is not always easy to assess the importance of registration practice in relation to other factors, such as changes in the natural history of diseases, altered diagnostic criteria and treatment methods, or even differences in reporting habits among physicians. These factors must be considered when comparing cause-of-death statistics from different countries.

ePub