The Economic Costs of Alcohol and Drug Abuse in the United States - 1992
Appendix A: Health Disorder Codes

Analyses of health care utilization and of mortalities resulting from alcohol and drug abuse are reported in chapters 4 and 5. These analyses rely on essentially the same set of diagnostic codes and attribution factors for the analysis of databases. The ICD-9 codes used in this study are presented in table A.1, along with the corresponding attribution factors. The table differentiates between conditions that are definitionally 100 percent caused by alcohol or drug abuse (e.g., drug psychoses) and those conditions for which alcohol or drug abuse are contributing factors.

Note also that the mortality data reported in chapter 5 rely on "E" codes for injuries and poisonings, whereas hospital and ambulatory care data reported in chapter 4 rely instead on the series 800 and 900 codes. The reason for this is that many injuries and poisonings never result in hospitalization–the "E" code assigned in an emergency room, for example, may be the only code that exists. On the other hand, if a patient is admitted or cared for by a physician outside the emergency room, standard diagnostic codes are apt to replace the E code as the primary diagnosis.

Sources: Stinson et al. (1993); Bloss (1995), personal communication.
*Note: A point of departure from the mortality estimates attributed to alcohol by Stinson et al. (1993) is that this study attributes 10 percent of nonfatal accidents and injuries to alcohol. Stinson's research concludes that alcohol caused 16 to 46 percent of certain accidents and injuries in which the victims died before admission to a hospital. This study attributes 10 percent of all nonfatal accidents and injuries that result in hospitalization or outpatient care to alcohol.