Overview and Implications
ICD-10 for mortality was implemented in the United States beginning with data year 1999. Additional details on the implementation of ICD-10 can be found at the Mortality Data from the National Vital Statistics System Web page.
The ICD-10 injury mortality framework was developed to be as consistent as possible with the recommended framework developed based on the ICD-9 external cause of injury codes as published in Morbidity and Mortality Weekly Report, August 29, 1997, Vol. 46, No. RR-14 [PDF – 325 KB]. Colleagues in the ICE (International Collaborative Effort) on Injury Statistics as well as in the Injury Control and Emergency Health Services (ICEHS) section of APHA participated in its development. However, it must be recognized that ICD-10 external cause of injury mortality codes are very different, often more detailed and at times far less detailed than codes in ICD-9. Thus, NCHS has decided to finalize the matrix because of its logic and internal consistency rather than basing changes on final comparability ratios.
Several changes were made to the ICD-10 matrix that warrant attention:
Two rows have been added. The first is labeled “All transport” and it includes all transport related deaths that were classified as unintentional, suicide, homicide, intent undetermined and operations of war. In ICD-9, the codes for suicide and intent undetermined by crashing of a motor vehicle were included with motor vehicle traffic injuries. There is no indication in the actual codes that these are traffic deaths. The second row, “Other land transport” was added to accommodate new codes in ICD-10.
A change was made to the transportation and drowning categories. The ICD-10 codes for water transportation-related drowning, V90 and V92, are included with the “other transport” codes rather than with the drowning codes. In the ICD-9 version of the matrix, the comparable codes, E830 and E832, were included with drowning. This change was made to be consistent with the categorization of other mechanisms of injury (i.e., falls, fires, and machinery) involved with water transport-related injuries.
In the ICD-9 matrix, E846-E848 “vehicle accidents not elsewhere classifiable” were categorized with other codes into the category “Other specified and classifiable”. However, with the additional transportation categories in ICD-10 and to be consistent with the NCHS 113 Cause-of-Death list, ICD-10 codes V98-V99, “Other and unspecified transport accident” are included with other transportation codes (including water transport and air and space transport related accidents) in a group V90-V99, making them part of all transportation related accidents.
The November 2002 version of the matrix shows a correction in the motor vehicle traffic category. Codes V81.1 and V82.1 were moved from the occupant codes to the other codes. In addition, codes X82 for suicide, Y03 homicide and Y32 intent undetermined for the crashing of a motor vehicle are included in the group, “other land transport” and code Y36.1 is included with “other transport.” In the preliminary matrix, they were only included in the row for “All transportation.”
The November 2002 version also includes the newly developed US ICD-10 codes for terrorism.
Further, because substantial changes were made in the classification of transport-related injuries in ICD-10, a more detailed categorization of transport codes is being offered for those users who want more detail than this framework offers.
The SAS Program document has the SAS statements needed to run the ICD-10 matrix. The Matrix document contains the ICD-10 codes and the detailed transport code groupings. These detailed transport codes are compatible with categories of the transport codes in the NCHS ICD-10 Underlying Cause-of-Death List for 113 causes (the list that replaces the 72-Cause list based on ICD-9 codes).
Finally, this matrix should be considered to be the final ICD-10 matrix. The preliminary comparability ratios are reported in: Comparability of Cause of Death Between ICD-9 and ICD-10: Preliminary Estimates. Additional analyses of comparability data suggest that there will be no further changes to the matrix.
Please e-mail Lois A. Fingerhut with any questions.