Although much is known about the correlates of heroin overdose, less is known about pharmaceutical opioid (PO) overdose. This study aimed to examine correlates of opioid overdose deaths by opioid and compare correlates between opioids.

Analysis of opioid overdose deaths in Australia between 2000-2015, extracted from the National Coronial Information System (NCIS). The NCIS is an online database of deaths reportable to the coroner, and contains coroner's findings, autopsy and toxicology reports. Deaths were categorized into mutually exclusive groups: 1) Heroin deaths; and 2) PO deaths (excluding heroin). PO deaths were examined by individual opioid.

There were 10,795 opioid overdose deaths over the study period. Relative to deaths occurring in major cities, deaths in regional/remote areas had 15.2 (95 % CI: 11.5-20.2) times the risk of being attributed to pharmaceutical fentanyl than heroin. Relative to deaths among people without a recorded history of chronic pain, deaths among people with a recorded history of chronic pain had a 1.9-10.7-fold increased risk of the death being attributed to POs than heroin. Deaths among people with a recorded history of substance use problems where the opioid was injected prior to death had 7.2 and 1.7 times the risk of being attributed to methadone and pharmaceutical fentanyl (respectively) than heroin.

Findings suggest the need to: educate PO consumers about the risks of overdose at the time of prescribing; increase coverage and engagement in opioid dependence treatment (particularly in regional/remote areas); and increase uptake of take-home naloxone to reduce opioid overdose mortality.