The idea of giving heroin addicts heroin to keep them from crime and other dangers has never been popular with American politicians. Yet several Western European countries routinely provide pharmaceutical-grade heroin to high-risk users in medically supervised facilities with minimal problems; the patients in these programs are much less likely to use street heroin when compared with patients in methadone programs. Given the tripling in heroin-related deaths in the last five years, it is time to give this innovation prompt consideration.

Though heroin has been a problem in American cities for nearly 50 years, it has been contained until recently. Baltimore, as memorialized in David Simon’s “The Wire,” was probably the city most affected. But nationally, from 1975 to about 2010, heroin use remained fairly stable, with only enough new users to replace those who died or, more rarely, quit.

Today’s opioid epidemic is different. The recent jump in overdose deaths in Maryland (62 percent in the first three quarters of 2016) involves more than heroin; many are caused by potent synthetic opioids, particularly fentanyl. Heroin, expensive as it is (about $10 per dose), became cheaper than street OxyContin. What really changed the situation was the introduction of fentanyl into illicit markets. Orders of magnitude more potent than heroin, it magnifies harm as users do not know their heroin has been spiked. Our current policy tools — prevention, substitution therapies and rehabilitation — are not specifically aimed at this risk of overdose. Even highly touted harm reduction interventions, such as naloxone, are blunted by fentanyl’s potency.

Taking opioid overdose deaths seriously requires consideration of new policies that get users out of street heroin markets. The principal advantage of heroin-assisted therapy (HAT) is that users are given heroin free of dangerous adulterants and of known quantity and purity. Programs in Germany, the Netherlands and Switzerland allow dependent, high-risk users to self-administer heroin several times a day under clinical supervision, as described in a 2009 Abell Foundation report, “Can Heroin Maintenance Help Baltimore?” Decades of studies have shown that prescribing heroin to those that have failed with traditional heroin-substitution therapies, like methadone or buprenorphine, substantially reduces overdoses and illegal activity while improving patient retention rates and individual health. Most importantly for the current American problem, HAT removes the incentive to purchase in street markets since the patient can choose his or her own dose. Naloxone is on hand should an overdose occur.

Heroin-assisted therapy addresses the immediate overdose threat posed by fentanyl — something naloxone attempts to do after the fact. Prescribed heroin use in a clinical and supervised setting ensures that users are not consuming fentanyl and that staff are on hand should something go wrong. Though this is HAT’s main attraction, it also serves important public health and social outreach functions. Interactions with clinicians and substance abuse counselors can and do help guide individuals into rehabilitation. Those who drop out of the program mostly transition to other kinds of treatment. It may well be that having discovered that ready access to heroin doesn’t make for a wonderful life, they are motivated at last to tackle their addiction.

After 25 years of evaluations we know that HAT can have positive and life-saving impacts. The challenge now is designing real-world policies that benefit American heroin users. Heroin epidemics have traditionally been clustered in cities, making HAT accessible to users who can walk in several times a day for treatment. Yet the heroin epidemic today ravaging small towns throughout rural America poses serious implementation challenges for HAT. The supply of qualified clinicians and the diffusion of users over larger areas make HAT adoption tricky, though mobile vans may help mitigate those problems. There are legal obstacles to overcome as well. It will not be easy to try out heroin assisted therapy.

Our opioid crisis is a genuine crisis, unlike any drug problem of recent decades in terms of the toll in lives lost. We cannot rely on well-worn policies aimed at reducing drug use to quickly stem the rise in overdoses. Innovative harm reduction policies such as HAT have been shown to reduce the risk of overdose and save lives. Even if it is not a silver bullet and even if it only works in some places, it is time to try heroin-assisted therapy in America.

Bryce Pardo is a doctoral student at the University of Maryland, where Peter Reuter (preuter@umd.edu) is a professor in the School of Public Policy and Department of Criminology.