Oxycodone involvement in drug abuse deaths: a DAWN-based classification scheme applied to an oxycodone postmortem database containing over 1000 cases

An oxycodone postmortem database was created from 1243 solicited cases from Medical Examiner and Coroner (ME/C) offices in 23 states in the United States over the period from August 27, 1999, through January 17, 2002. The request for cases was specific to only those cases in which the ME/C opined that the death involved oxycodone. Each case was evaluated to determine the role of oxycodone and the specific drug product OxyContin tablets in the death. Oxycodone identification was based on toxicology testing, and OxyContin identification was based on evidence found at the scene, credible witness reports, or identification of tablets in gastrointestinal contents. A system of case categorization was developed for this study based on the Drug Abuse Warning Network (DAWN) system for reporting drug abuse mortality data in the United States, using the same standardized, well-understood terminology. Of the 1243 cases, 79 cases were incomplete and could not be evaluated. There were an additional 150 cases submitted in which oxycodone was not identified by the originating ME/C. Of the remaining 1014 cases, 919 (90.6%) were related to drug abuse, whereas 95 (9.4%) cases were categorized as not involving drug abuse. Only 30 (3.3%) of the drug abuse cases involved oxycodone as the single reported chemical entity; of these, 12 cases had OxyContin identified as a source of oxycodone. Of the 919 drug abuse cases, the vast majority (N = 889, 96.7%) were multiple drug abuse deaths in which there was at least one other plausible contributory drug in addition to oxycodone. The most prevalent drug combinations were oxycodone in combination with benzodiazepines, alcohol, cocaine, other narcotics, marijuana, or antidepressants. Using the DAWN definitions, drug abuse cases were further categorized as drug-induced or drug-related. A total of 851 (92.6%) cases met the criteria for classification as being drug-induced, and the remaining 68 (7.4%) cases were categorized as drug-related. Cause of death (COD) statements from the originating ME/C indicated a general recognition of the role of abuse of multiple drugs in causing fatalities. Approximately 70% of the 889 cases in the multiple-drug-induced categories were listed in the COD or contributing COD statements as multiple-drug deaths. A variety of terms were employed in the COD statements to indicate multiple drug involvement such as "polydrug toxicity", "polypharmacy", "multiple drug poisoning", and "polypharmaceutical overdose". The system for death classification employed in this study recognizes the problems inherent in COD attribution when multiple drugs are involved. Use of this new system for reporting mortality data in future studies involving opioids is recommended.