Speed misreading: meth survey on crack.

IN JANUARY USA Today announced that "methamphetamine accounts for more emergency room visits than any other drug." That story, like similar news accounts the same month, was based on a survey of hospital emergency rooms by the National Association of Counties (NACo), a group that has been pushing for more federal funding of its members' anti-drug efforts by warning that the U.S. faces a nationwide "meth epidemic." NACo used impressionistic responses from a skewed survey sample to claim "there are more meth-related emergency visits than [visits] for any other drug and the number of these visits has increased substantially over the last five years."

NACo started with a non-representative sample by focusing on county-run hospitals, facilities of last resort that disproportionately treat pooh uninsured patients. The results may have been further skewed by response bias, since it's plausible that hospitals seeing a lot of meth users would be more likely to participate. NACo said it "contacted county public hospital or regional hospital emergency rooms in 48 states" and got 200 responses from E.R. officials in 39 states, representing about percent of the nation's emergency rooms.

The responses came overwhelmingly from Midwestern states (72 percent) and from counties with populations below 50,000 (80 percent)--places where methamphetamine use is especially common. About three-quarters of respondents said they'd seen an increase in "hospital presentations where methamphetamine was involved" during the previous five years. Nearly half said meth was the illegal drug most commonly involved in E.R. visits. No data were required to back up either impression.

By contrast, the most recent data from the Drug Abuse Warning Network, which focuses on the urban areas that account for 58 percent of emergency rooms and 82 percent of E.R. visits in the U.S., indicate that meth-related cases are far less common than cases involving cocaine, heroin, or marijuana. "We never said this was a comprehensive study," a NACo spokesman told the Drug War Chronicle. No, they only implied it.