259 



Medicare. For example. Medicare discharges that had a primary diagnosis of lung cancer fell 

 into any one of 8 DRGs (depending on the procedure required or other complications) and, 

 therefore, were paid a different amount depending on the DRG. By placing the discharges 

 calculated above for each ICD-9 code into their appropriate DRG. we were able to adjust for 

 the case mix index (CMI), or the relative payment level for that discharge. In this way. we 

 were able to obtain a case-mix adjusted total for substance abuse-attributable discharges. We 

 then multiplied these weighted discharges by the standardized national average DRG payment 

 for 1991 ($3,974) to determine total Medicare substance abuse costs. 



. While adjusting for CMI allows us to capture the higher cost per discharge for 

 certain diagnoses, it does not measure the differential impact on length of stay when substance 

 abuse is recorded as a secondary- diagnosis. To capture the incremental costs of substance 

 abuse as a complicating factor in treating conditions unrelated to substance abuse, we also 

 analyzed the marginal impact of substance abuse as a secondary diagnosis on hospital length 

 of stay. For this analysis, we defined substance abuse as only those diagnoses that explicitly 

 mention drug or alcohol use (e.g. alcohol poisoning) or that are the immediate reaction to 

 substance use (e.g. delirium tremens). 



We calculated the difference in length of stay for patients with and without these 

 substance abuse secondary diagnoses that had the same primary diagnoses (by gender and for 

 the under 65 and over 65 age groups) to determine the marginal days of care that were 

 substance-abuse related. Estimating an average cost of $604 per day for these extra days, we 

 then added these incremental costs to our total. 



16 



