17 October 2000

Hulse et al. omit deaths in naltrexone report.

re: Hulse GK, Basso MR. Reassessing naltrexone maintenance as a treatment for illicit heroin users. Drug Alcohol Review 1999 18:263-269

Comment letter published D&A Review 2000 19;3:347

[Note: no reply published from original authors]
Dear Editor,

I support the authors' contention that naltrexone be judged using the same criteria as other treatments to enable fair comparisons of outcomes.

It is self-evident that stricter acceptance criteria will yield lower apparent 'success' rates, but that these will not alter actual patient outcomes. Clinical trials are rarely directly comparable so one of the tasks of clinicians is to determine the significance of each addition to the scientific literature. For dependency treatments this is not difficult since the aims of treatment are relatively straightforward and can be verified by simple means.

Hulse and Basso have reported detailed outcomes of a naltrexone treatment trial using two criteria for reporting six month follow-up details. Despite using tables and text to illuminate these outcomes, they place these results out of clinical context by omitting the two deaths which were previously reported from this cohort [Hulse G. Naltrexone - Defining a framework for use. Australian Professional Society on Alcohol and other Drugs (APSAD) Newsletter 'Summer' 1998, p4].

Mortality is the most fundamental end point in any treatment trial. To ignore, omit or mis-state deaths or other end points makes interpretation of data impossible and is inappropriate in balanced scientific writing.

A comparison of methadone and naltrexone in dependency management should be rigorous and impartial. After twenty years and over 500 peer-reviewed papers demonstrating the benefits of methadone, there has been only very limited research favouring naltrexone treatment in heroin addicts in the same time period. Few if any accredited addiction treatment services use naltrexone as a first line drug although it shows promise in certain selected groups and is effective in alcoholism.

The authors of all scientific research studies should be scrupulous in all reporting, lest they be seen as advocates for particular causes instead of maintaining the detachment which is essential in ethical research.

Yours faithfully,

Andrew Byrne ..