A retrospective observational study of reportable deaths involving MA was conducted using data obtained from the Victorian Institute of Forensic Medicine (VIFM) and the National Coronial Information System (NCIS). Regarding the NCIS, the Victorian Department of Justice and Community Safety was the organizational source of the data and the NCIS was the database source.

The VIFM is a centralized independent statutory authority in Victoria, Australia, that serves a population of ~6.7 million people. Reportable deaths that must be investigated by the Victorian State Coroner include violent, unnatural, and unexpected deaths. The VIFM assists in > 7000 coronial death investigations per year through the provision of forensic medical and scientific services.

As part of the normal process on admission to the mortuary, PM femoral blood was collected by upper thigh arterial puncture as soon as practicable. PM blood specimens were preserved with 1% w/v sodium fluoride and potassium oxalate and stored at 4°C until analysis, usually within 24 h. Ante-mortem (AM) toxicology specimens were obtained where possible if the deceased person died in hospital and stored at −20 °C. The VIFM database was searched for deaths in which MA was detected between 1 January 2010 and 31 December 2019.

The NCIS is a national database on deaths reported to the Coroner in Australia and New Zealand [11]. “Methylamphetamine” was entered into the Query Design search screen and the pharmaceutical substances for human use code set to complement the search of the VIFM database.

Validated methods of liquid chromatography-tandem mass spectrometry (LC–MS/MS) were used to detect the most common and toxicologically relevant basic and neutral drugs [12]. Toxicologically relevant acidic drugs were detected by an additional LC–MS/MS method [13]. MA blood concentrations were determined by a quantitative LC–MS/MS method. In brief, 100 µL of blood was pipetted into a 2 mL polypropylene tube with added internal standard. Then, 200 µL of trizma base (2 M, pH 9.2) was used to buffer the sample and 1000 µL of butyl chloride was added for a liquid–liquid extraction. The extract was evaporated to dryness and reconstituted in 100 µL of methanol. Samples were injected into the LC–MS/MS system and separated on an Eclipse XBD C18 column (150 × 4.6 mm, 5 µm) using a gradient elution of 50 mM ammonium formate in water (pH 3.5) and 0.1% formic acid in acetonitrile. MA was detected using an Applied Biosystems 5500 Q-TRAP MS/MS (positive electrospray, multiple reaction monitoring mode). The relative standard deviation estimates of accuracy, within- and between-day precision of the method were 3.1, 11, and 16% at 0.03 mg/L; 8.5, 6.6, and 6.6% at 0.45 mg/L; and 2.9, 7.2, and 9.1% at 0.9 mg/L, respectively.

Specialist forensic pathologists conduct the medico-legal death investigation in Victoria. A decision is made by the Coroner, with pathologist advice, about the type of autopsy required for each case. Autopsy may be limited to external examination in suitable cases where the cause of death is apparent. Furthermore, external examination alone may also be completed due to objections to internal autopsy by senior next-of-kin based on religious, cultural, or other reasons are considered (these are usually granted). Internal examination of the deceased typically involves investigation of the major organ systems by macroscopy and histology, with additional ancillary tests as required (e.g., vitreous humor, biochemistry, and microbiology). All deceased have a PM computerized tomography scan on admission to VIFM.

The cohort comprised Victorian deaths reported to the Coroner (≥ 18 years old) due to primarily natural causes and accidental drug toxicity in which MA was detected in blood (≥ 0.02 mg/L). Deaths due to other unintentional causes (e.g., motor vehicle crashes or other accidental causes where MA was not included in the cause of death), intentional self-harm, assault, and undetermined causes were excluded. Deceased persons with altered body condition (i.e., traumatized, burnt, or decomposed) identified at mortuary admission were also excluded. Furthermore, cases without internal examination were not included since underlying natural disease may have been an unknown causative or contributory factor to death. Only cases that had PM femoral blood were included and toxicological results of AM blood were preferred if available.

Cases were classified into five groups: deaths due to MA toxicity and absence of other factors (Group A1); deaths due to MA toxicity in the setting of other potentially contributing factors (Group A2); deaths due to MA toxicity in the setting of significant natural disease (Group B); deaths primarily due to multiple-drug toxicity (Group C); and deaths primarily due to natural causes (Group D).

Anatomical findings at autopsy were reviewed by a forensic pathologist for significant natural disease defined as any natural disease of the type and severity potentially causal or contributory to death. For example, only severe coronary artery atherosclerosis (≥ 70% stenosis) was defined as significant. Cardiac hypertrophy was determined using an online heart weight calculator (https://calc.chuv.ch/heartweight) and reference cardiac dimensions [14, 15]. Mental illnesses were not considered organic pathologies and were outside the scope of the study.

The contribution of other toxicologically significant drugs was reviewed by a forensic toxicologist. Using similar methods to Pilgrim et al., if one or more drugs were detected at potentially toxic concentrations, the other drugs detected at 'therapeutic' concentrations were considered incidental [6, 16]. For example, in a case with MA and diazepam concentrations of 1 mg/L and 0.2 mg/L, respectively, diazepam was not considered significant. Additionally, MIMS Drug Interactions and CredibleMeds® QTdrugs List were used to assess drug interactions and the risk of QT prolongation, respectively [17, 18]. Drugs administered by medical staff were excluded where medical depositions and ambulance reports permitted. In the absence of 6-acetylmorphine, the likelihood of heroin use relied on the morphine:codeine ratio and circumstantial evidence of intravenous drug use [19].

In some cases, the circumstances suggested the possibility that other factors may have contributed to the death and were classified as Group A2. The final classification of cases was reviewed by a panel of forensic pharmacologists/toxicologists and a forensic pathologist.

Continuous variables were summarized using means or medians, where appropriate, and categorical variables as percentages. Observations with low frequencies have been reported as < 5 unless explicit endorsement was obtained from the Victorian State Coroner. The Kruskal–Wallis H test or Mann–Whitney U test were used to compare MA concentrations. Assumptions were satisfied and post-hoc analysis was performed by Dunn’s procedure with Bonferroni adjustment for multiple comparisons. p-values < 0.05 were considered statistically significant. All statistical analyses were performed using IBM® SPSS® Statistics v28.0.1.1.

This study was approved by the VIFM Ethics Committee (1188–1180/1), Coroners Court of Victoria Research Committee (RC 408), and Justice Human Research Ethics Committee (CF/21/M0481).