Lies,
Damn
Lies
and
the
Drug
Abuse
Warning
Network
Statistics
6/23/00
The University of Maryland's Center for Substance Abuse Research (CESAR, http://www.cesar.umd.edu) devoted its latest weekly CESAR FAX (6/19), which goes out to some 4500 subscribers, to some dramatic but misleading numbers from the Drug Abuse Warning Network (DAWN). DAWN reports on emergency room visits for "drug-related" episodes, which it defines as emergency room visits "induced by or related to the use of an illegal drug(s) or the non-medical use of a legal drug for patients 6 years or older." The fax's headline tells the story: "Marijuana-Related Emergency Department Visits Now as Common as Heroin-Related." The brief accompanying text describes a steady increase in marijuana-related ER visits, from 16,251 in 1991 to a projected 79,088 this year. Heroin-related visits are estimated to hit 77,009 this year. Cocaine, meanwhile, will account for a projected 154,956 ER visits, the highest number for any illicit drug. The inference to be drawn from the headline is that marijuana and heroin are somehow having a roughly equal impact on users, at least in hospital emergency rooms. The headline, however, ignores the huge difference in the universe of heroin users and that of marijuana users. The ratio of marijuana users to heroin users is roughly 20:1. A more apt headline could have read, "Heroin 20 Times More Likely than Marijuana to Be Mentioned in ER Visits." In the fax's text, CESAR descends even further into distortion. The text describes ER incidents where marijuana has been mentioned as occurring "because of marijuana use." "CESAR should know better," said Doctor of Public Health David Duncan of Westat, a research organization doing work under contract with the federal government. One of Westat's ongoing contracts involves evaluating the utility and efficacy of the DAWN program. "Merely because marijuana was mentioned," Duncan told DRCNet, "doesn't mean the visit should be attributed to the drug." One of the problems with the DAWN numbers, Duncan said, is the confusion between mention and causation. "If you are struck by lightning and the autopsy finds cocaine in your system, for example, "that would be listed as a cocaine-related ER incident." Similarly, DAWN methodology allows for the mention of up to four drugs for each incident, so someone who comes in with a heroin overdose but mentions having smoked marijuana will go into the statistics as a marijuana mention as well as a heroin mention. According to Chuck Thomas of the Marijuana Policy Project, another problem with DAWN is that it does not distinguish the qualitative differences among different ER incidents. Unlike life-threatening hard drug overdoses, Thomas told DRCNet, most marijuana-related ER incidents are presumably "anxiety attacks." "Someone gets too high and his friends aren't smart enough to put on some mellow music and tell him to calm down," said Thomas, "so they go to the hospital and the staff there tells him to calm down." DAWN's limitations are familiar to professionals in the field. Problems arise, however, when the data are misinterpreted or misunderstood, either willfully or because of lack of understanding of methodological issues. Likewise, the selective cherry-picking of isolated statistics, as CESAR appears to have done, distorts the overall findings of the DAWN reports. Among the findings CESAR did not see fit to mention:
According to Dr. Duncan, the paramount point for people to remember when discussing the DAWN numbers is "that they represent only people who mentioned the drug in some way, not that it had any role in causing the visit." And, he cautions, "These numbers track only ER visits for the substances listed; they do not track drug use levels. Neither do they track alcohol, the most widely abused drug in America, because the DEA wasn't interested in that." The DAWN report is available online at http://www.samhsa.gov/OAS/DAWN/DawnMidYr/99mid_year.pdf.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
