George O'Neil, the founder of Australia's first Naltrexone clinic, has become embroiled in yet another controversy following a report that addicts coming off the treatment have a much higher risk of death than untreated addicts, or addicts on methadone.

Dr O'Neil, who provides clinical support to St Kilda's The First Step clinic, admits that 82 of his Perth patients have died in three-and-a-half years but defends the Naltrexone treatment, claiming the deaths occurred because the addicts stopped the therapy.

Perth psychiatrist James Fellows-Smith and GP John Edwards claim their two-year WA study shows heroin addicts on Naltrexone programs had a one-in-61 chance of dying and a significantly higher risk of dying than addicts on no treatment at all, who had a mortality rate of one-in-74. Addicts on the methadone program had demonstrated a mortality ratio of one-in-458.

The medical director of the WA Government-run clinic Next Step, Alan Quigley, said there were clearly concerns about the report's findings on increased mortality rates.

"What is disappointing for us is that overseas research has already indicated Naltrexone programs are associated with higher mortality rates, and it's been very difficult to get that message understood,"Dr Quigley said.

"Now this paper really does confirm what is already known. It's good that it's been reported because ... perhaps some of the media enthusiasm for Naltrexone will be tempered by WA's experience of the drug."

Dr Quigley said clinical staff at Next Step had felt under enormous pressure from consumers and the public to provide treatment they did not consider professionally appropriate.

However WA academic Gary Hulse, who is chief investigator of the WA Health Department's Naltrexone studies trial, was highly critical of the report.

He said the study was poorly designed and had been retrospectively "cobbled together".

Dr Hulse also questioned the release of the report to peers and the media before it had been accepted for publication by a journal of review.

But one of the authors of the study, Dr John Edwards, said he and Dr Fellows-Smith had released the report to cut into some of the "media hype" and to try to achieve a more balanced view of Naltrexone.

"(Naltrexone) has found a place for about 5 per cent of people who are addicted to heroin, who happen to be the most stable, usually professional people.

"The ones who are most at risk are the chaotic people who are really not interested in taking on treatment ... they're at very high risk of dying because of lowered tolerance, and also because of other drugs used with Naltrexone to help with withdrawal."

But Dr O'Neil says people who have died following Naltrexone treatment do not die on Naltrexone or from Naltrexone therapy, they die because they stop the therapy.

While the dangers of dying from overdose are real, there are ways of controlling the risk, and his clinic had set up systems to do that.

"If you're going to use Naltrexone you have to do it properly or not at all. The success rate is 100 per cent if the Naltrexone is delivered."