As a surgeon, Jonathan Shepherd has seen at first hand the consequences of violent crime on countless occasions.
Treating injured victims in the operating theatre, he has done what he can with a medical training based on generations of research to repair the damage.
Now, as director of Cardiff University's Violence Research Group, he wants the research-led approach that made medicine the profession it is today to be replicated in the criminal justice system, on the undeniable basis that prevention is better than cure.
Speaking after he picked up a prestigious award for his work in the field, Professor Shepherd said research on crime and justice was languishing "in the Victorian age" and required the creation of a new academic discipline to overhaul it.
He said: "At the end of the 19th century, Sir William Osler, professor of medicine at Oxford and at the University of Pennsylvania, was faced with few effective remedies and great uncertainty about 'what works'.
"He said universities must invade the hospitals, and he got his way. The idea was to evaluate through scientific methods the medical treatments of the time, to ditch what didn't work and to invest in what did.
"Medical schools go back many centuries before that, but that invasion of the hospitals by the universities set the scene for an explosion of high-quality experiments, clinical trials and new remedies that came in the 20th century.
"It put applied science in its rightful place, buffered from political fad and fashion. It helped prevent enormous harm and saved countless lives."
Professor Shepherd argued that a similar "invasion" is necessary in the country's criminal justice system, a view that may find favour in the current climate of concern about levels of knife and gun crime among young people.
"There are professors, clinical academics and research teams across every medical specialty that you'd care to name, employed largely by Russell Group universities.
"Move across to criminal justice and we have criminology. To some extent, psychology and the social sciences play a part, but the applied side, the equivalent of surgery, medicine and psychiatry, just does not exist.
"We could have university schools of police science or offender management, for example, but we don't. In order to find out what works in crime and justice, we must invest in those applied crime sciences," he said.
At Cardiff University, the Universities' Police Science Institute is already attempting to replicate the medical school model for the police service, but Professor Shepherd's involvement in the field goes back to his days as a dentistry undergraduate, when he worked at a legal advice centre in a poor area of London.
Later, as a trainee surgeon in Leeds, he was struck by the surge of violence during the miners' strikes in the Thatcher years, which sparked an academic interest in the field.
This led to his PhD at the University of Bristol in the 1980s. For this, he analysed crime figures and hospital records and found a startling mismatch - only a quarter of the violence that put people in hospital appeared in police records.
"I did not know then that the weekly forays up the back stairs of the Bristol central police station would have lifelong implications," he joked.
One barrier to integrating university-led research and policing and probation work, Professor Shepherd said, has been a "credibility gap" that existed between the two.
"There are very few police officers with PhDs, and it's currently almost impossible for a police officer to get a training in police science at a Russell Group university.
"There are lots of graduates in the police service, but would you want your doctor to have a degree in biochemistry or physiology? No, you would want them to have a degree in medicine.
"In the same way, it seems to me, a police officer who is going to run policing in a city and is allocating resources should know and should be very close to the evidence of what works, just as a consultant or GP is in the health service.
"There's a credibility gap. Police officer training has traditionally been in military-style establishments, such as Hendon for the Met, that are not integrated with universities, where criminologists work.
"As a result of that, I have heard a very senior police officer say, 'I can't think of one practical step forward that has come out of criminology in this country' and at the same time I have heard a criminologist on a campus say, 'You can't teach police officers how to do research.'
"You would never hear that in medicine, that it is impossible to teach doctors how to do research, because they are embedded in Russell Group universities churning out high-quality evaluations," he said.
Professor Shepherd has had a unique insight into changing patterns in offending.
He said: "Since 2000, the trend in violence based on A&E data has been one of steady decline in terms of overall numbers. But at the same time we have seen increases in the number of people admitted to hospital, including for knife injuries, so the overall picture is one of decreasing rates of violence, but increase in severity."
