NDARC Technical Report No. 309 (2010)
Aims: At present no peer-reviewed publication exists regarding the extent of use and dependence of heroin/opioids at the global level. To address this need for information, we aim to present the first systematic review of prevalence estimates, across all 229 UN Member States, for heroin/opioid use and dependence. We include all countries' most recent estimates, with methodological details, for the prevalence of use and dependence on heroin (or other opioids, if heroin estimates were unavailable). We also suggest priorities for improving data quality and coverage.
Methods: Relevant evidence was identified through systematic searches of grey and peer reviewed literature (1990-2008), online databases, and the WWW; repeated consultation and feedback from experts worldwide; and a viral email to lists in the illicit drug and HIV fields. Data was extracted using manualised protocols, checked for internal consistency, classified using predefined variables and quality scored. This paper reports the most recent and highest graded prevalence estimate for the general population and school population and reports the proportion of coverage of the world's population for use and dependence estimates, general population and school surveys, age and sex specific estimates, and most recent year of estimates.
Results: Evidence of heroin/opioid use or dependence was found for 192 of 229 countries worldwide. For 101 countries, with 18.2% of the world population aged 15-64 years (WP15-64), no prevalence estimates were available. For 25 countries (33.5% WP15-64) dependence estimates were available. Fifty four countries (48.7% WP15-64) had estimates for use among the general population and 65 (40.4% WP15-19), for use among school-aged youth. Just 16 countries, constituting 5.6% WP15-64 had a national estimate of dependence using indirect (gold standard) methods as well as a national general population survey (direct) estimate of use. National past year prevalence estimates ranged from 0-3% (any use in the past year) and 0.1-0.8% (dependence). Age ranges, estimation methods and the types of 'opioid' assessed differed widely.
Conclusions: The available prevalence data is incomplete, inconsistent and therefore unable to meet the needs of public health policy makers attempting to plan scaled responses. There is a need for greater data coverage, more rigorous methods, and regular data collection to improve this situation.