Table 2.

Data Sources and Values Used in the Current Study

Authors, Year	Description of source	Description of how data were used in the current study	Source’s definition of heavy drinker (where applicable)
Clark and Hilton, 1991	1984 National Alcohol Survey	% of population who are heavy drinkers (PctPop_HD=11% in sensitivity analysis) % of alcohol consumed by heavy drinkers (PctAlc_HD=67% in sensitivity analysis)	Consumption of 5 or more drinks on one occasion at least once a week
LaVallee and Yi, 2010	Estimated annual alcohol consumption nationally and by state	Average alcohol consumption per person in 1989-1994 (c=303 oz)	N.A.
Levy et al., 2002 Plus supplemental analyses	Analysis of Oklahoma data on hospitalized spinal cord, burn, drowning and traumatic brain injuries; victims ages 15 to 64; alcohol involvement noted in >90% of injuries.	See Table 3 for values. Alcohol involvement averaged over all four categories (p=28% in sensitivity analysis)	N.A.
Manning et al., 1991	RAND Health Insurance Experiment 1983 National Health Interview Survey	% of population who are heavy drinkers (PctPop_HD=10% in sensitivity analysis) % of alcohol consumed by heavy drinkers (PctAlc_HD=68% in sensitivity analysis)	Consumption of an average of at least 5 drinks per day
Miller et al., 2001 Plus supplemental analyses	A case-control study that compared the odds of injury among people aged 14-64 whose 1987-1989 health insurance claims include treatment for diagnoses that the Alcohol Epidemiologic Data System of the National Institute on Alcohol Abuse and Alcoholism (Cases et al., 1994) classifies as alcohol- related.	Risk of injury among heavy drinkers relative to other people (RR_HD); reruns at ages 18-64 to better match the Soderstrom et al. (1997) age distribution are used in the main analysis (RR_HD=1.82; sensitivity range 1.67 when drug-only cases are used in the estimation, 2.09 when excess risk for alcohol and drug users is attributed to alcohol, 2.56 for ages 15-64). For spinal cord and burn injuries (RR_HD=1.8 and 1.27, respectively, with sensitivity ranges of 1.63- 1.93 and 1.2-1.4)	Any treatment for alcohol-related diagnoses that include chronic alcohol abuse indicators such as cirrhosis, subacute indicators such as alcoholic psychoses, and acute ethanol poisoning.
Ray and Ksir, 1999	Book on the biological and social aspects of alcohol use and abuse.	People metabolize an average of 0.45 ounces of absolute ethanol per hour.	N.A.
Smith et al., 1999b; Selway et al., 2008; Miller et al., 2010	Smith reviews the literature on the role of alcohol in injury. The other studies fill gaps or provide more robust estimates for selected mechanisms	Alcohol involvement in injury by cause. See Table 3 for values.	N.A.
Soderstrom et al., 1992	Interim report of Soderstrom et al. (1997) study	Percent of hospitalized injuries to heavy drinkers ages 18 and over who were alcohol-positive (PInj_HD=62% in sensitivity analysis)	Alcohol abuse or dependence according to DSM-III-R
Soderstrom et al., 1997	Examined the percentage of serious trauma patients who were alcohol-positive, by alcohol dependency	Percent of hospitalized injuries to heavy drinkers ages 18 and over that were alcohol- positive (PInj_HD=56.7%)	Alcohol abuse or dependence according to DSM-III-R
Substance Abuse and Mental Health Services Administration, 1998	U.S. Census Bureau survey that probes alcohol use of a national probability sample of U.S. households	% of population ages 18 and over who did not consume alcohol, past month (44.6%) % of alcohol consumed by people ages 18 and over (96.5%) % of population who are heavy or dependent drinkers (PctPop_HD= 9%; 6.2% heavy drinkers in sensitivity analysis) % of alcohol consumed by heavy and dependent drinkers (PctAlc_HD=53%; 47.5% heavy drinkers in sensitivity analysis)	Drank 5 drinks or more on at least 5 days in the past month (6.2%) or were alcohol-dependent according to an abridged version of the DSM- IV (3.6%) or were treated for alcohol problems in the past year (0.3%)