194 



costs. First, while we have attempted to pull together all available q)idemiologic research on 

 the health effects of substance abuse, more research is needed. Our results only reflect the 

 current state of the art in this area.' Second, studies reveal that identification and reporting of 

 substance abuse problems by medical practitioners is poor. For example, estimates of 

 underreporting of substance abuse secondary diagnoses run as high as 60%. For reasons of 

 confidentiality and concern over insurance reimbursement, physicians are reluctant to record 



* 



substance abuse unless it relates directly to the primary diagnosis or the treatment plan. 

 Assuming that only 40% of cases with substance abuse actually listed it on the medical record, 

 the complicating costs of substance abuse comorbidity may be two and a half times higher than 

 estimated here. Third, there is little identification of tobacco use or abuse of prescription 

 medications on the medical record: our estimates only include the complications of alcohol and 

 illicit drug abuse. Fourth, using an average hospital cost of $750 per day may be low if 

 substance abusers require a greater intensity of services. For example, if substance abuse bum 

 patients are more likely to stay longer in the Intensive Care Unit (ICU) at an average cost per 

 day of $3,000, these additional costs would not be captured in our analysis*. Finally, our 

 estimates do not include general hospitalization costs of caring for people who join the Medicaid 

 rolls, and benefit from its coverage, due to job loss, disability, or poverty, related to substance 

 abuse. 



' The association between illegal drug use and resulting illness has not been as thoroughly studied 

 as that of smoking and alcohol because drug use is less prevalent in the general population and more 

 difficult to identify since subjects are reluctant to admit openly to illegal conduct. Alcohol studies are 

 also somewhat limited, due in part to the greater difficulty in establishing level of use (self-reporting of 

 alcohol use is less reliable than that of tobacco because heavy use of alcohol has a negative social 

 stigma). Even for cigarette smoking, a great deal of research is available on illnesses highly prevalent 

 in the population such as lung cancer and heart disease, but less is available for less prevalent diseases, 

 such as Crohn 's disease. Thus, our study only includes those diseases and conditions that have been 

 clearly documented as related to substance abuse. We attempted to use the best research available, 

 recognizing that the field of epidemiology is constantly evolving and sharpening its findings. Further 

 inquiry into other related conditions would most likely significantly increase substance abuse-related 

 Medicaid hospitalization costs. 



'A study at Johns Hopkins Hospital revealed that 28 percent of 435 ICU admissions and 39 percent 

 of ICU costs were substance abuse-relaxed (Baldwin et al). 



