Trends In Opiate Overdose Related Hospital Admission In a Nationally Representative Sample

Background: Heroin and opiate addiction harms individuals, families, and communities. Local and anecdotal data suggest increased opiate overdoses since the early 1990s in the United States, but few studies have examined trends, correlates, or demographic changes in overdose-related hospital admissions in a nationally representative sample. This project seeks to document the extent of damage done by heroin and other opiates by characterizing trends in opiate overdoses between 1993 and 2007. Using a nationally representative sample of hospital admissions and Drug Enforcement Administration (DEA) data on Heroin price and purity we attempt to describe demographic differences in those trends and their relationship to heroin price.

Method: Using the nationally representative Nationwide Inpatient Survey (NIS), we estimated the frequency of heroin and opiate overdose hospital admissions in 1004 hospitals nested in 22 cities between 1993 and 2007. Using heroin or opiate overdose ICD-9 codes we calculated overdoses by race, gender, age and city. Frequencies and means adjusted for complex sampling design were used to estimate changes in demographics. Data from the DEA heroin price and purity database (STRIDE) was used to estimate the price of heroin adjusted for purchase purity, city, and date. Growth models including coefficients for heroin price and year were used to predict changes in the frequency of opiate overdose hospital admissions nested in cities and hospitals.

Results: In 1993 there were 8,889 (34.5 per million US population) opiate overdose admissions to US hospitals, by 2007 this number had increased to 40,655 (135 per million), a 291% increase. Whites saw the largest increase from 23.48 per million to 119.37 per million, a 408% increase in overdoses. Black overdose admissions increased 105% to 80.65 per million in 2007, and Hispanics overdose admissions increased 48% to 44.82 per million. Female overdoses increased 414%, while male overdoses increased 214%. All 5-year age categories between 16 and 64 saw increased overdose admission rates. Overall there was a statistically significant and increasing trend in hospital-based opiate overdose admissions, with a model implied 11% increase in admissions each year between 1993 and 2007. Over the study period the model implied price of heroin decreased 6% each year. Each logged dollar decrease in the price of heroin was associated with a 4% increase in the number of hospital opiate overdose admissions.

Conclusions: There was a substantial increase in the rate of heroin and opiate overdose hospital admissions between 1993 and 2007 in the United States. We also found evidence that the heroin and opiate overdose epidemic has shifted, becoming more white and female. Furthermore, we found that reductions in the price of heroin in local communities were associated with substantial increases in the number of opiate overdoses in those communities. We believe that these data provide evidence for the need to shift more resources to social work and treatment oriented community-based organizations working to reduce the harm associated with opiate addiction and away from drug interdiction efforts, which appear to be failing to reduce availability of heroin and other opiates.