228 



created a methodology for addressing issues of comorbidity. Using tlie National Hospital 

 Discharge Survey (NHDS), Rice first estimated the cost of alcohol, drug, or mental illness- 

 related as a primary diagnoses following Harwood's model. Then, recognizing that secondary 

 diagnoses of substance abuse complicates the treatment of other diseases and thus adds to 

 hospital costs, Rice also calculated the additional days of care rqwrted for all primary diagnoses 

 that had a secondary ADM diagnosis. Rice acknowledges at the outset that her estimates are 

 low, restricted by the information reported on the medical records. In fact, many studies have 

 documented that underreporting of secondary diagnoses is common, especially for conditions 

 such as substance abuse that do not require direct treatment but contribute to longer stays and 

 are considered embarrassing by the patient. 



Costs of Smoking 



Quantifying the costs of smoking has been a major public health issue since the 1960's. 

 Annually, the Surgeon General issues a report on smoking and health which summarizes all 

 current epidemiologic evidence on the relationship between smoking and disease and death. 

 The most noteworthy of these was Reducing Health Consequences of Smoking: 25 Years of 

 Progress, issued in 1989, which reported smoking attributable fractions (SAFs) for ten selected 

 causes of death using data collected in a four year, fifty state study conducted by the National 

 Cancer Society. These SAFs represent the proportion of deaths for a given disease that could 

 have been avoided if cigarette smoking were eliminated. 



Many economic cost studies have relied on these estimates to calculate the number of smoking- 

 attributable deaths for specific regions and the number of years of potential life lost as a result 

 of smoking. Some have also employed these mortality statistics to estimate hospital utilization 



-57- 



