Opioid-induced deaths in Australia

Opioids accounted for just over 3 deaths per day in 2018. The majority of these opioid-induced fatalities were unintentional overdoses in middle aged males involving the use of pharmaceutical opioids, often in the presence of other substances. Opioid related harm, including mortality, is a serious public health issue both in Australia and internationally.

• Of the 1,740 registered drug-induced deaths in 2018, opioids were present in close to two thirds (1,123 deaths, 64.5%). Opioid-induced mortality in 2018 was slightly lower than that recorded in 2017 with a per capita rate of 4.6 per 100,000 people compared to 4.8.
• Pharmaceutical opioids are present in over 70% of opioid-induced deaths. The rate of opioid-induced deaths with synthetic opioids present has increased significantly over the last decade. In 2018 there was a decrease in the number of deaths with naturally derived and semi-synthetic opioids present compared to 2017.
• There were 438 heroin-induced deaths in 2018. This is the highest number of heroin-induced deaths since the year 2000, with the increase being significant over the last 5 years.
• This paper focusses on opioid-induced deaths. An opioid-induced death is one where death is attributable directly to drug use and where an opioid was present. An opioid may have been present singularly, for example heroin was the only drug found on toxicology, or the opioid may have been found in combination with other substances. Excluded from analysis are opioid-related deaths, which are deaths where opioid use contributed to death, but the cause of death was something other than drug-induced (e.g. a traffic accident where the deceased was under the influence of heroin).

Opioid class drugs work by binding to opioid receptors in the brain which control pain and reward to inhibit messages of pain sent to the body (Le Merrer et al., 2009). Pharmaceutical opioids are prescribed for pain relief. For those seeking to understand and measure opioid use the following characteristics are important:

• Illicit nature, e.g. some opioids are prescription medications prescribed for pain (e.g. codeine), or to treat opioid dependence (methadone). Some opioids are illegal (e.g. heroin).
• The strength of the opioid and how much is needed to produce desired effects, e.g. heroin is a strong opioid and tramadol is a weak opioid.
• Individual drug type, e.g. fentanyl, heroin, tramadol.
• Derivation properties, for example those which are derived via the poppy plant (e.g. codeine, heroin) and those which are synthetically derived to act on opioid receptors (e.g. fentanyl).
   

Deaths in the National Mortality Dataset are coded to the International Classification of Diseases, 10th Revision (ICD-10). The ICD-10 is structured to assign drug related mortality to a number of different chapters depending on circumstance of the death. Some drug addiction and misuse related deaths may be coded to the mental and behavioural disorders chapter. The majority of drug-induced deaths are due to acute overdose and are assigned to the external cause chapter. Intent of death (e.g. accident, suicide), mechanism of death (e.g. exposure to narcotics and noxious substances) and a poisoning code that specifies the actual type of drug involved in an overdose (e.g. heroin) are all captured for acute overdoses.

In Australia, most opioid-induced deaths are referred to a coroner and subject to forensic pathology and toxicology. Autopsy and toxicology reports provide detailed drug information including the identification of specific drugs in the system, approximate levels of drugs in the system and the relatedness of drugs to the death. This information is accessed by the ABS for coding and statistical dissemination.

The ICD-10 coded data are a rich resource for understanding drug-induced mortality in Australia. However, it is important to consider how drugs are captured in ICD-10, as there are certain limitations with the current structure. Listed below are ICD-10 poisoning codes as they relate to opioids, alongside important notes on coding and dissemination considerations.

T400 Opium : Opium is a dried latex obtained from the poppy plant. It contains a collection of alkaloids (known as opiates) which have an effect on the brain and spinal cord. Opium is an illicit substance in Australia and is considered a "weak" opioid. Opium as a drug is rare in mortality statistics in Australia.

T401 Heroin: Heroin is a narcotic derived from morphine, a natural substance taken from the seed of the poppy plant. Heroin is an illicit substance in Australia and considered a strong opioid. Heroin is rapidly metabolised by the body, and is converted to monoacetylmorphine (MAM) and then to morphine. The presence of MAM indicates heroin use as opposed to morphine use. At times, toxicology is not able to determine MAM, and in these cases the death is coded to T402 "Other opioids", as only the morphine derivative can be identified.

T402 Other Opioids: Pharmaceutical opioids which are derived or synthesised from the poppy plant are coded to this category. The category includes codeine, oxycodone, hydromorphone and morphine. This category includes both weak (codeine) and strong (oxycodone) opioids. How the drug is used, the single drug within the category and the strength of the drug cannot be further disseminated in statistical output.

T403 Methadone: Methadone is a synthetic opioid which is used to treat heroin or other opioid dependence. It is a strong opioid and is available only by prescription in Australia.

T404 Other Synthetic Narcotics: Pharmaceutical opioids which are synthetic (man-made) are coded to this category. The category includes fentanyl, tramadol, pethidine and tapentadol. This category includes both weak (tramadol) and strong (fentanyl) opioids. How the drug is used, the single drug within the category and the strength of the drug cannot be further disseminated in statistical output.

T406 Other and unspecified narcotics: This code is used when an opioid is known to be involved in an overdose death yet the type is not specified and a toxicology or autopsy report is not yet available for coding. Often when this code is used the coroner is still investigating the death and a more specific opioid code can be applied when data are revised.

The opioid-induced death rate peaked in the late 1990s at 6.5 per 100,000 people, largely due to heroin (Degenhardt et al., 2006). Rates were lower in the early to mid-2000s but have recently increased. Opioid-induced deaths over the last five years are significantly higher than those recorded ten years ago with a death rate of 4.6 recorded in 2018 compared to 4.0 in 2009.

Males consistently have a higher rate of opioid-induced death than females, being on average 2.3 times higher over the last two decades. The National Drug Strategy Household Survey 2016 (AIHW, 2017) found that males were more likely to misuse illicit drugs (including heroin), however the rate of misuse of pharmaceutical opioids was similar across both sexes.

Although females experience lower rates of opioid-induced deaths, rates over the last five years have been higher (3.5 per 100,000 in 2016) or comparable (2.9 in 2018) to that recorded in 1999 (3.2).

The age distribution of opioid-induced deaths differs considerably from that for all causes. The highest proportion of deaths (30.4%) occurs in those aged between 35-44, while 87.5% of deaths occur between the ages of 25-64. In total 39,221 years of potential life were lost, and on average a person dying from an overdose with opioid involvement died 34.9 years prematurely.

Drug misuse is associated with a range of adverse social and health outcomes including lower socioeconomic status, unemployment and emotional distress (AIHW, 2019). Of the 1,123 opioid-induced deaths, there were 458 people who had a mental health condition including schizophrenia, mood disorders and anxiety. A further 64.5% of people recorded a mental and behavioural disorder due to alcohol and drug use, including misuse and addiction. Of note, there were 174 people (15.5%) who had chronic pain coded as a contributory factor to death.

The rate of opioid-induced deaths among males is highest in those aged 35-44 (15.1 per 100,000). Among females the rate is highest in the 45-54 year age group (6.2).

The differences in age specific death rates for males and females are linked to the type and pattern of drug use. Males are more likely to have the illicit drug heroin identified as contributing to death while for females pharmaceutical opioids are more likely to be present. The median age at death for those who die with illicit opioids (42.1 years) is lower than that of pharmaceuticals (46.6 years).

There is a clear shift in the age at death for opioid-induced deaths, with peak rates gradually shifting from the younger age groups (15-34) to middle aged groups (35-54) over the last two decades. For those aged over 55 years, the rate of opioid-induced overdose is higher than ten and twenty years ago.

This is supported by figures in the National Drug Strategy Household Survey which reported an increase in the average age of people who used illicit and pharmaceutical substances between 2001 and 2016 (AIHW, 2017). The average age of people who used illicit drugs was 34 years (up from 30 in 2001) and for pharmaceutical drug misuse was 43 years (up from 39 in 2013). Older people have been identified as a priority population in the National Drug Strategy (DoH, 2017) with unique health circumstances such as pain, co-morbidities, and social circumstances such as isolation being highlighted as important factors to consider in the context of drug use.

Opioids have been the most common class of drug identified on toxicology reports for drug-induced deaths for the past two decades. With the exception of the opioid peak in 1999, mortality rates for all drug classes have increased over the past 20 years.

Before the year 2000, the rate of opioids detected at death was over two times higher than that of other depressant drugs, a class of drug which slow down the activity of the central nervous system to ease anxiety, relax people and initiate sleep (includes benzodiazepines). This rate ratio has decreased to 1.1 in 2018 (4.6 opioid deaths per 100,000 compared with 4.0 benzodiazepine deaths).

The change in rate ratio between opioids and stimulants, drugs which increase activity and alertness in the brain and include the illicit substances of amphetamines and cocaine, is also marked. Twenty years ago the rate of opioid detection at death was over 12 times higher than that of stimulants. This has reduced to a rate ratio of 2.5 in recent years.

Of the 1,123 registered opioid-induced deaths, 88.9% of these occurred in the setting of other substances. Benzodiazepines were the most common drug to appear alongside opioids with 708 deaths (63.1%) having both drugs present. Approximately one-quarter of opioid-induced deaths also recorded an anti-depressant or anti-psychotic drug.

The majority of opioid-induced deaths are accidental (80%), with a further 16% due to suicide. While most overdoses are accidental for both sexes, females have a higher proportion of suicidal overdoses involving opioids at 26.7% compared to 10.8% in males.

On average, accidental opioid-induced deaths occurred at an earlier age than intentional overdoses for both males and females. Accidental drug overdoses had a median age of 43.0 years for men and 46.8 years for women. In contrast, suicidal overdoses recorded a median age at death of 49.6 for men and 56.3 for women.

Pharmaceutical opioids are the most common opioid present in suicidal overdose. Of the 179 opioid-induced suicide deaths, close to 80% (140 deaths) had a natural or semi-synthetic opioid present.

Understanding the contribution of different types of opioids to the overall number of opioid-induced deaths is important as different interventions and prevention strategies are required depending on the drug. For example, the National Drug Strategy aims to minimise drug related harms through demand reduction (delaying or preventing the uptake of drugs), supply reduction (e.g. preventing supply of illegal drugs) and harm reduction (reducing adverse health consequences) (DoH, 2017).

Prescription opioids were identified in 70.7% of opioid-induced deaths. The natural and semi-synthetic opioids, including codeine, oxycodone and morphine were the most common prescription opioids present, followed by synthetic opioids. In 2017-18, under the Pharmaceutical Benefits Scheme (PBS) codeine (with paracetamol) was the most commonly dispensed prescription opioid, followed by oxycodone and tramadol (PBS, 2018).

The structural composition of opioid-induced deaths differs between the sexes. Close to half (46.3%) of all opioid-induced overdoses in males involve the illicit substance heroin, compared with just under a quarter (23.7%) for females. The higher use of illicit drugs by males has been demonstrated consistently in the National Drug Strategy Household Survey (AIHW, 2017). Over three-quarters of opioid deaths for females had a prescription opioid identified (82.6%).

Deaths with illicit substances identified have a lower median age at death compared with pharmaceutical opioids. This is consistent across both sexes. Among females the average age at death for heroin overdose was 41.5 years of age compared to 50.4 years of age for pharmaceutical opioids.

The graph below shows changes in death rates over time for selected opioids. Pharmaceutical opioids have been presented by specific type to show their contribution to opioid-induced deaths.

Natural and semi-synthetic opioids have had the highest mortality rate over time, but the rate has reduced in the past 2 years. This type of opioid includes codeine. From February 2018 codeine products were changed from Schedule 2 or 3 of the Standard for the Uniform Scheduling of Medicines and Poisons to Schedule 4, prescription only (TGA, 2018). Scheduling of drugs refers to how they are made available to the public. The doubling of deaths due to codeine between 2000 and 2009 and increasing harms associated with its use were cited as reasons for the change by the Therapeutic Goods Administration (Roxburgh, et al., 2015 in TGA, 2017). Codeine cannot be separately identified in the mortality dataset, but changes in mortality rates for this group of opioids will be monitored closely.

The death rate from synthetic opioids has increased over the past decade from 0.15 per 100,000 people in 2009 to 1.0 in 2018. Rates have stabilised over the past five years ranging from 0.9 to 1.0. Synthetic opioids include the drug fentanyl, a particularly strong opioid which has been linked to many deaths in the United States of America (CDC, 2019). As such changes in the mortality rate for this group of opioids would be of particular interest to government. The TGA established the Opioid Regulatory Advisory Group which have recommended stricter guidelines around prescription, including that fentanyl patches should only be prescribed for pain in cancer patients, palliative care or other exceptional circumstances (TGA, 2019).

After a period of stability in the mid-2000s the rate of opioid-induced deaths involving heroin have increased over the last five years. Though the rate of 1.8 per 100,000 people in 2018 is not as high as the peak in heroin deaths in the late 90s (2.3), it is the highest recorded since the year 2000 and has doubled since 2008. The National Wastewater Monitoring Report (ACIC, 2019) found that in the December quarter of 2018 there was an increase in the consumption of heroin (population weighted) in capital cities in New South Wales and Victoria.

People living outside of capital cities had a higher rate of opioid-induced deaths with 5.0 deaths per 100,000 people compared to 4.3 for those living in capital cities. The National Wastewater Monitoring program (ACIC, 2019) reported that apart from heroin, cocaine and alcohol, the regional average of drug consumption was higher than that in capital cities.

Those who died with heroin identified on toxicology were more likely to die in a capital city. Victoria had the highest count of heroin-induced deaths (165) across states and territories, while the highest death rate was recorded in Western Australia at 3.1 per 100,000 people.

Those who died with pharmaceutical opioids identified on toxicology were most likely to live outside of a capital city. People living outside of the Perth region in Western Australia had the highest rate of fatal pharmaceutical opioid overdose at 5.1 deaths per 100,000 people.

The structure of ICD-10 enables capture of sufficient detail to produce data on some specific drugs and some types of drugs. However, it cannot capture all relevant details relating to drug deaths and or separate all specific drugs that may be of particular interest or concern. For opioids the improvements in output will be notable. Opioids will be separated into mutually exclusive categories including fentanyl, tramadol, codeine and oxycodone enabling a finer level of analysis.

In May 2019, the World Health Assembly endorsed the revision of the ICD (ICD-11). The revised classification contains new structures which support the capture of additional details using extension codes. ICD-11 extension codes include a full drug classification - the International Nonproprietary Names (INN) - which will allow information to be captured on all individual drugs. These extension codes may also allow capture of other details relating to drug overdoses and deaths, adding considerably to the amount of information available to those seeking to reduce drug related harm

The ABS is developing plans for ICD-11 adoption in collaboration with the World Health Organization and other stakeholders. While ICD-11 adoption may still be some time away, trials of ICD-11 structures could begin to deliver additional details relating to drug deaths, while also serving to refine and enhance the classification in preparation for future adoption.