Mortality among regular or dependent users of heroin and other opioids: a systematic review and meta-analysis of cohort studies
Article first published online: 4 NOV 2010
© 2010 The Authors, Addiction © 2010 Society for the Study of Addiction
Volume 106, Issue 1, pages 32–51, January 2011
How to Cite
Degenhardt, L., Bucello, C., Mathers, B., Briegleb, C., Ali, H., Hickman, M. and McLaren, J. (2011), Mortality among regular or dependent users of heroin and other opioids: a systematic review and meta-analysis of cohort studies. Addiction, 106: 32–51. doi: 10.1111/j.1360-0443.2010.03140.x
- Issue published online: 10 DEC 2010
- Article first published online: 4 NOV 2010
- Submitted 22 February 2010; initial review completed 17 May 2010; final version accepted 21 July 2010
Aims To review the literature on mortality among dependent or regular users of opioids across regions, according to specific causes, and related to a number of demographic and clinical variables.
Methods Multiple search strategies included searches of Medline, EMBASE and PsycINFO, consistent with the methodology recommended by the Meta-analysis of Observational Studies in Epidemiology (MOOSE) group; grey literature searches; and contact of experts for any additional unpublished data from studies meeting inclusion criteria. Random-effects meta-analyses were conducted for crude mortality rates (CMRs) and standardized mortality ratios (SMRs), with stratified analyses where possible. Meta-regressions examined potentially important sources of heterogeneity across studies.
Results Fifty-eight prospective studies reported mortality rates from opioid-dependent samples. Very high heterogeneity across studies was observed; pooled all-cause CMR was 2.09 per 100 person-years (PY; 95% CI; 1.93, 2.26), and the pooled SMR was 14.66 (95% CI: 12.82, 16.50). Males had higher CMRs and lower SMRs than females. Out-of-treatment periods had higher mortality risk than in-treatment periods (pooled RR 2.38 (CI: 1.79, 3.17)). Causes of death varied across studies, but overdose was the most common cause. Multivariable regressions found the following predictors of mortality rates: country of origin; the proportion of sample injecting; the extent to which populations were recruited from an entire country (versus subnational); and year of publication.
Conclusions Mortality among opioid-dependent users varies across countries and populations. Treatment is clearly protective against mortality even in non-randomized observational studies. Study characteristics predict mortality levels; these should be taken into account in future studies.