Alcohol and injury: a comparison of three emergency room samples in two regions

Alcohol's involvement in injury has been found to vary greatly depending on the type and location of emergency room (ER), and associations may also vary by regional variations in drinking patterns. Similar data on alcohol and injury are compared in patients from three distinctly different ER types (trauma center vs county/ community and HMO) in two regions of the country (Mississippi and California).

Representative samples of injured ER patients from the University of Mississippi Medical Center in Jackson (n = 356), San Francisco General Hospital (SFGH) (n = 555) and Contra Costa County, California, (n = 1,630) were breath analyzed and interviewed at the time of ER admission.

Men in the Jackson sample were less likely than those in SFGH, but more likely than those in Contra Costa, to be positive on the breath analyzer and to report drinking prior to the event, while women in Jackson were less likely to be positive than those in the other two samples. Similar rates of heavy problem drinking were found for those in Jackson and Contra Costa and both rates were lower than those in SFGH. Subjects in Jackson were more likely to be drinking in their own home, to be drinking an alcoholic beverage other than wine and to report a shorter time lapse between the last drink and the event, compared to the other two samples. Quantity-frequency of drinking was found to be predictive of injury in addition to breath analyzer reading and reporting drinking prior to the event, and significant interaction terms were found for drinking variables by site.

The data suggest that breath analyzer readings may be closely associated with type of ER, urbanicity of the area and socio-demographic characteristics of the clientele, while drinking prior to the injury may be closely associated with regional variations in drinking. Drinking variables appear to be more important predictors of injury in the Jackson sample than in Contra Costa or SFGH.