Examination of mortality rates in a retrospective
cohort of patients treated with oral or implant
naltrexone for problematic opiate use
Erin Kelty
1,2
& Gary Hulse
1
School of Psychiatry and Clinical Neuroscience, University of Western Australia, Sir Charles Gairdner Hospital, Nedlands, Western Australian, Australia
1
and Fresh
Start Recovery Programme, Subiaco, Western Australia, Australia
2
ABSTRACT
Aims To examine and compare mortality rates in patients treated with oral and implant naltrexone. Design A
retrospective cohort study. Setting A community not-for-profit drug treatment clinic. Participants Patients treated
with oral naltrexone (n = 2155, 17 207 patient-years) and implant naltrexone (n = 2389, 11 678 patient-years) for
problematic opiate use between August 1997 and December 2009. Measurements Crude gender, age, treatment
period and cause-specific mortality rates were calculated using data obtained from the National Death Index.
Findings Crude mortality rates for patients treated with oral naltrexone [8.78 deaths per 1000 patient-years (ptpy),
95% confidence interval (CI): 7.38–10.17] were significantly different to those treated with implant naltrexone
(6.59 ptpy, 95% CI: 5.13–8.06) (P = 0.0339). During the first 4 months following treatment, differences in the two
groups were particularly apparent, with a mortality rate of 26.28 ptpy in patients treated with oral naltrexone
compared to 7.34 ptpy in patients treated with implant naltrexone (P = 0.0003). Differences in initial mortality rates
following treatment were associated predominantly with high rates of opiate overdoses in oral naltrexone patients
during the first 4 months following treatment (17.22 ptpy compared with 0.67 ptpy in implant naltrexone patients)
(P < 0.0001). Conclusions The use of implant naltrexone can reduce all-cause mortality and opiate overdose during
the first 4 months following treatment compared with patients treated with oral naltrexone.
Keywords Naltrexone, mortality, opiates, overdose death.
Correspondence to: Erin Kelty, School of Psychiatry and Clinical Neuroscience, University of Western Australia, Sir Charles Gairdner Hospital, QE II
Medical Centre, Stirling Highway, Nedlands, Western Australian 6009, Australia. E-mail: erin.kelty@freshstart.org.au
Submitted 9 January 2012; initial review completed 23 February 2012; final version accepted 28 March 2012
INTRODUCTION
Since its initial US registration in 1984 for the treatment
of opiate dependence, a number of safety concerns have
been raised regarding the use of oral naltrexone [1].
These are: (i) an observed increase in the number of
opiate overdoses following cessation of oral naltrexone
[2,3]; (ii) the potential for increased use of non-opiate
drugs resulting in increases in fatal and non-fatal
non-opiate drug overdoses [4,5]; and (iii) increases in
depression and suicide due possibly to the inability of
endogenous opiates to bind to the opiate receptor [1,6].
Mortality rates for heroin-using populations have
been estimated at between 6.8 and 77.6 per 1000
patient-years (ptpy), with clear geographical differences
in mortality [7]. Mortality rates in Australia are among
the lowest in the world, with crude mortality rates of
between 8.9 and 18.0 [7]. Patients in methadone main-
tenance therapy (MMT) and buprenorphine mainte-
nance have mortality rates generally between 4.5 and
15.2 ptpy [8–10]. Mortality rates for oral naltrexone
remain comparable to other drug treatments (approxi-
mately 10 ptpy); however, mortality rates have been cal-
culated to be as high as 221 ptpy in the 2 weeks following
cessation of oral treatment [3,11].
The observed increase in opiate fatalities following
the cessation of oral naltrexone is postulated to be asso-
ciated with a reduction in a patient’s opiate tolerance as
patients transition from regular opiate use to naltrexone
[3,12,13]. While taking daily naltrexone, patients are
RESEARCH REPORT doi:10.1111/j.1360-0443.2012.03910.x
© 2012 The Authors. Addiction © 2012 Society for the Study of Addiction Addiction, 107, 1817–1824