Tilting at windmills and the evidence base on injecting drug use.

The art of medicine Tilting at windmills and the evidence base on injecting drug use After reading through many of the 500 or so papers on harm reduction in drug users listed in the PubMed database, it's hard not to conclude that researchers are really bad at understanding the evidence base. Paper after paper begins its introduction with a homily about the need for evidence-based policies. Paper after paper concludes with a stirring call for policies that deliver what the research promises: lower rates of HIV, blood-borne infections, mental illness, and social disruption for places that are bold enough to give drug users easy access to sterile injecting equipment, opioid substitution therapy, cognitive therapy, and the other lesser jewels in the harm reduction crown. So says the science. So say we, the high priests of evidence-based policy. But there's another body of evidence, just as overwhelming. It shows, time after time, that elected politicians don't give a damn about scientific evidence when it comes to policies that benefit a minority that holds little sway over other voters. Injecting drug users form just such a minority. People who choose to sell sex for a living do, too. Policies around addiction and prostitution bulldoze happily through the scientific evidence in a quest to do what works best at the ballot box. What's the response of the research community? Most often, we call for more research. The notion that 501 studies will prove more persuasive than the 493 current is at best quixotic. At worst, a cynic might say, it is self-serving, although to be fair to the cavaliers of harm reduction research, most are driven more by the quest for decent services for addicts than by the wealth and glory that might cascade from a more commercially and thus politically popular goal—a cure for cancer, for example. But the windmills of public opinion at which they tilt are unlikely to be skewered on the lance of another well-honed p value. This is not to suggest that well-designed research does not shape policy. It can, and it sometimes does even in areas such as addiction where a scientific evidence base is built on a marshy foundation of morality and the slurry of legal and cultural history. The evidence base didn't change between US President Barack Obama's election in November, 2008, and the end of the ban on the use of federal funding for …