Shootings and killings in deprived areas of Chicago and Baltimore have plummeted by between 41 and 73 per cent thanks to a programme that treats violence as if it is an infectious disease.
Pioneers of the programme, called CeaseFire, say it relies on simultaneously changing attitudes and behaviour and will work anywhere.
The key is to change social norms so that violence is seen as "uncool" both by potential perpetrators and their communities, instead of being the automatic way to settle a dispute.
On 30 June, pioneers of the programme publicised their high success rate so far to attract interest at a time in the year when violence peaks, triggered by the heat.
"Violence gets transmitted the same way as other communicable diseases, so we train 'violence interruptors' to prevent escalation," says Gary Slutkin, founder and executive director of CeaseFire.
"They change the norm from 'violence is what's expected of me' to 'violence will make me look stupid'," says Slutkin.
While violence interruptors work on the streets to intercept and defuse disputes before anyone gets hurt, outreach workers work in parallel to get the same message to the community, through schools and key members such as clergy.
The net effect is that the "default" norm of instant violence rapidly changes to one in which shooting is seen as unacceptable and unfashionable. "1800 of these types of events have been successfully mediated in the past 4 years," says Slutkin.
The first programme was launched in 1999, and now covers a quarter of Chicago and has more recently been introduced in Baltimore.
A three-year independent evaluation of CeaseFire published by the Department of Justice last year found that in Chicago, it reduced violence in every community where it was deployed.
Shootings and killing fell by between 41 and 73 per cent, with drops of 17 to 35 per cent the result of direct interventions by CeaseFire. Retaliation murders fell by 100 per cent in 5 of the 8 communities covered.
"This is a method that's now in demand," says Slutkin. "We've been visited by representatives from 30 cities, and training has started in at least six."
The most important but controversial element of the programme to tackle the epidemic of violence is sending reformed shooters out into the streets as mediators in disputes and mentors for youths.
"At age 17 I'd been arrested several times and convicted of four gun-related offences," says Jalon Arthur, a key violence interruptor speaking in a teleconference on 30 June.
"You must work with the shooters, and credible messengers like me need to understand the minds of the perpetrators and have the measure of street-credibility to overcome mistrust," said Arthur.
"For individuals who engage in gun activity, there's a great deal of paranoia which makes them very difficult to influence, but because violence interruptors have been through the transition themselves, they have the social networks and enough street cred to reach these people," said Arthur.
William Pollack, a clinical psychologist at Harvard Medical School who has been involved with CeaseFire, says that norm-changing is key to the programme because otherwise, the myth persists that boys can't help but be violent.
"People believe that boys are doomed to be aggressive, especially males of colour and poverty, and that's how boys must settle disputes," says Pollack. "But we now know it's the socialisation experience that creates this expectation, nurture rather than nature."
The CeaseFire model changes these social norms and behaviour," adds Pollack. "It's a scientifically based programme that works."
Slutkin says that once a behavioural norm reaches a critical level in a community, it's sustained, creating a long-term change.
Moreover, Slutkin is confident it will work anywhere. "We've been visited by 15 other countries, and the same training we use in the US is now being deployed in Brazil and the Caribbean, and we've had a few other visits from Mexico and even Basra in Iraq, where we've defused 100 events this year."
Arthur says he's convinced the programme would work in the UK, where knife crime has been escalating.
Slutkin said there are huge savings to be made in the money otherwise spent on hospital admissions, treatments and aftercare, as well as the costs to the criminal justice system of investigating shootings and killings. "We calculate that the cost of not doing it in the US amounts to $150 to $200 billion each year," he said.