230 



Other studies have estimated the costs of specific diseases (Harwood, 1985), of specific sub- 

 populations (Phibbs, 1991; Rivo, 1990), of distinct hospital departments (Hauswald, 1989), and 

 of state health expenditures (Rice, 1991; Spiegel, 1990) associated with one or more substance. 

 Most of these studies employed some version of the Rice or Harwood methodology. CASA's 

 study also starts with Rice and Harwood's previous work, incorporating both the concept of 

 disease-specific attributable risks to substance abuse and the marginal affects of substance abuse 

 as a secondary diagnosis. 



-59- 



