189 



2) Treatment of Diseases Totally Attributable to Substance Abuse - In Category 1, the 

 hospital stay was specifically for the treatment of the substance abuse problem. For this 

 and the next category, the hospital stay was for treatment of a medical disease that may 

 have been caused by the use or abuse of a substance. In this category are discharges that 

 had a diagnosis that either specifically mentioned a substance in its name (e.g. alcoholic 

 cirrhosis), or that the National Institute on Alcoholism and Alcohol Abuse considers as 

 solely attributable to alcohol (e.g. pellagra), or that involve a secondary diagnosis of 

 substance abuse in 1(K)% of the NHDS cases reported (e.g. esophageal varices). Since 

 the hospital stay was for medical treatment of diseases caused solely by substance abuse, 

 100% of these hospital days were attributed to substance abuse. 



3) Treatment of Diseases Where Substance Abuse is a Major Risk Factor - From an 

 extensive review of epidemiologic research (see Bibliography), CAS A identified 72 

 conditions and diseases that have substance abuse as a major, but not the exclusive risk 

 factor. These include diseases such as lung cancer and low birth weight associated with 

 smoking; accidents and cardiovascular diseases associated with alcohol use; and 

 premature strokes and AIDS associated with drug use. The prospective, population-based 

 or case control studies used for this analysis often calculated (or provided sufficient data 

 for CASA to calculate) a Population Attributable Risk (PAR) for a specific substance and 

 disease. PAR is an epidemiologic term meaning the percentage of a given illness that 

 could be prevented if the use of the substance were eliminated.' In other words, the 



' These PARs are based on the best available epidemiologic research investigating the relationship 

 between substance abuse and morbidity. For some diseases and conditions, there was clear evidence that 

 a relationship exists between substance abuse and the occurrence of the condition, but prospective or case 

 control studies which calculate PARs had not been conducted. In these cases, we employed other 



-14- 



