Comparability Ratio

About every 10-20 years the International Classification of Diseases (ICD) is revised to stay abreast of advances in medical science and changes in medical terminology. Each of these revisions produces breaks in the continuity of cause-of-death statistics. Discontinuities across revisions are due to changes in classification and rules for selecting underlying cause of death. Classification and rule changes impact cause-of-death trend data by shifting deaths away from some cause-of-death categories and into others. Comparability ratios measure the effect of changes in classification and coding rules. For causes shown in Table VI, comparability ratios range between 0.9754 and 1.0588, except for Influenza and pneumonia, for which the comparability ratio is 0.6982, indicating that Influenza and pneumonia is 30 percent less likely to be selected as the underlying cause-of-death in ICD-10 than in ICD-9; and HIV disease with a comparability ratio of 1.1448, indicates that HIV disease is more than 14 percent more likely to be selected as the underlying cause using ICD-10 coding.

Another factor also contributes to discontinuities in death rates across revisions. For selected causes of death, the ICD-9 codes used to calculate death rates for 1980-98 in this report differ from the ICD-9 codes most nearly comparable with the corresponding ICD-10 cause-of-death category.

Examples of these causes are ischemic heart disease, cerebrovascular diseases, trachea, bronchus and lung cancer, unintentional injuries, and homicide. To address this source of discontinuity, mortality trends for 1980-98 were recalculated using ICD-9 codes that are more comparable with codes for corresponding ICD-10 categories. Table V shows the ICD-9 codes used for these causes. While this modification may lessen the discontinuity between the Ninth and Tenth Revisions, the effect on the discontinuity between the Eighth and Ninth Revisions is not measured.

Preliminary comparability ratios shown in Table VI are based on a comparability study in which the same deaths were coded by both the Ninth and Tenth Revisions. The comparability ratio was calculated by dividing the number of deaths classified by ICD-10 by the number of deaths classified by ICD-9. The resulting ratios represent the net effect of the Tenth Revision on cause-of-death statistics and can be used to adjust mortality statistics for causes of death classified by the Ninth Revision to be comparable with cause-specific mortality statistics classified by the Tenth Revision.

The application of comparability ratios to mortality statistics helps to make the analysis of change between 1998 and 1999 more accurate and complete. The  1998 comparability-modified death rate is calculated by multiplying the comparability ratio by the 1998 death rate. Comparability-modified rates should be used to estimate mortality change between 1998 and 1999. 

Caution should be taken when applying the comparability ratios presented in Table VI to age-, race-, or sex-specific mortality data. Demographic subgroups will sometimes differ with regard to their cause-of-death distribution. This will result in some demographic variation in cause-specific comparability ratios.

SOURCE: Health, United States 

Related Links
Comparability of Cause of Death Between ICD-9 and ICD-10
National Vital Statistics Reports (Vol. 49 No. 2 and Vol. 49 No. 3)
Cause of Death
International Classification of Diseases

This page last reviewed June 18, 2004