[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
[Skip to Content Landing]
February 6, 2002

Changes in HIV-Related Deaths as a Function of Coding

Author Affiliations

Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor

JAMA. 2002;287(5):588. doi:10.1001/jama.287.5.583

To the Editor: In their Research Letter, Dr Grigg and colleagues1 found that changes in the rules for selecting the underlying cause of death in Florida, which resulted from implementation of the International Classification of Diseases, 10th Revision (ICD-10), were responsible for a 14% increase in the number of deaths attributed to human immunodeficiency virus (HIV) disease between 1998 and 1999. This figure is in contrast with a 6% increase due to these rule changes reported by the National Center for Health Statistics (NCHS), which was based on analysis of a large sample of national mortality data from 1996.2 Grigg et al concluded that application of the original NCHS estimate may result in an inaccurate adjustment to the number of deaths caused by HIV disease in 1998 for comparison with deaths in 1999. At the NCHS, we also came to this conclusion. As a result, with the release of preliminary 1999 mortality data for the United States, NCHS (June 2001) used an estimate of the impact of the rule changes on HIV-related deaths based on data from 1998 instead of 1996.3 We found that the application of ICD-10 rules to 1998 data for the United States would cause a 14% increase in deaths attributed to HIV disease, which is consistent with the findings later reported from Florida. This revised estimate resulted in a much more accurate assessment of the change in HIV mortality from 1998 to 1999.