The Clinton administration on Monday endorsed the practice of giving clean needles to drug addicts in order to prevent transmission of the AIDS virus. "A meticulous scientific review has now proven that needle-exchange programs can reduce the transmission of HIV and save lives without losing ground on the battle against illegal drugs," Secretary of Health and Human Services Donna Shalala announced.
The administration is not unanimous, however; the drug czar, Gen. Barry McCaffrey, who opposes needle exchange, was out of the country Monday. Who's right? As recently as a month ago, HHS had resisted needle-exchange programs. "We have not yet concluded that needle exchange programs do not encourage drug use," spokeswoman Melissa Skolfield told the Washington Post March 17. By Monday the department had reached that conclusion, though the scientific evidence that needle exchanges don't encourage drug use is as weak today as it was a month ago.
In fact, the evidence is far from clear that needle-exchange programs protect against HIV infection. Most studies have had serious methodological limitations, and new studies in Montreal and Vancouver have revealed a troubling pattern: In general, the better the study design, the less convincing the evidence that clean-needle giveaways protect against HIV.
The Montreal study, the most sophisticated yet, found that those who attended needle-exchange programs had a substantially higher risk of HIV infection than intravenous drug addicts who did not. In a much-discussed New York Times op-ed article two weeks ago, Julie Bruneau and Martin T. Schechter, authors of the Montreal and Vancouver studies respectively, explained the higher risk this way: "Because these programs are in inner-city neighborhoods, they serve users who are at greatest risk of infection. Those who didn't accept free needles . . . were less likely to engage in the riskiest activities."
Dr. Bruneau is apparently rejecting her own research. For her study had statistical controls to correct for precisely this factor. In the American Journal of Epidemiology, Dr. Bruneau wrote: "These findings cannot be explained solely on the basis of the concentration around needle-exchange programs of a higher risk intravenous drug user population with a greater baseline HIV prevalence."
Even more troubling, Dr. Bruneau reported that addicts who were initially HIV-negative were more likely to become positive after participation in the needle exchange. Dr. Bruneau speculated that needle-exchange programs "may have facilitated formation of new sharing networks, with the programs becoming the gathering places for isolated [addicts]."
Janet Lapey of Drug Watch International says needle-exchange programs often become "buyer's clubs" for addicts, attracting not only scattered users but opportunistic dealers. Not everyone agrees. Dr. Schechter says that when he asked his study's heroin users, they reported meeting elsewhere. But a delegation from Gen. McCaffrey's office returned from Vancouver in early April with some startling news: Although more than 2.5 million clean needles were given out last year, the death rate from illegal drugs has skyrocketed. "Vancouver is literally swamped with drugs," the delegation concluded. "With an at-risk population, without access to drug treatment, needle exchange appears to be nothing more than a facilitator for drug use."
The problem for science is that no study has used the most effective method for settling such issues--a randomized control trial. Moreover, needle-exchange programs are usually embedded in complex programs of outreach, education and treatment, which themselves affect HIV risk. A 1996 study showed that through outreach and education alone, HIV incidence in Chicago-area intravenous drug users was reduced 71% in the absence of a needle exchange.
Peter Lurie of the University of Michigan argues that "to defer public health action on those grounds [awaiting better research] is to surrender the science of epidemiology to thoughtless empiricism and to endanger the lives of thousands of intravenous drug users." But Dr. Lurie's reasoning appears circular. Only someone already convinced that needle-exchange programs are effective at preventing HIV can claim that addicts are jeopardized by further testing.
And drug use carries risks besides HIV infection. A recent article in the Journal of the American Medical Association warned that the arrival of a new drug from Mexico called "black-tar heroin," cut with dirt and shoe polish, is spreading "wound botulism." This potent toxin leads to paralysis and agonizing death, even when injected by a clean needle.
Thus, dispensing needles to the addicted could produce a public health tragedy if this policy does indeed place them at greater risk for HIV or enhances the legitimacy of hard drug use. Simply put, the administration's case is not proven.
Mr. Murray is director of research for the Statistical Assessment Service, a nonprofit group in Washington.
