Trends in cannabis-related ambulance presentations from 2000 to 2013 in Melbourne, Australia

Highlights

• Cannabis-related drug presentations to ambulance services are on the rise across age groups.

• Polydrug use is common in particular alcohol and crystal methamphetamine.

• The majority of presentations result in transport to hospital.

• Further work is required to identify at risk populations.

Abstract

Aims

The current burden of cannabis-related presentations to emergency health services is largely unknown. This paper presents data collected over a 13-year period in metropolitan Melbourne, Australia as part of the Ambo Project, a unique surveillance system that analyses and codes paramedic records for drug-related trends and harms.

Methods

Cannabis-related ambulance attendances involving 15–59 year olds in metropolitan Melbourne were analysed retrospectively from 2000 to 2013 (n = 10,531). Trends and attendance characteristics were compared among cannabis only (CO)-, cannabis and alcohol (CA)- and cannabis with polydrug use (CP)-related attendances. Changes in alcohol and drug involvements in cannabis-related attendances were explored.

Results

Rates of cannabis-related ambulance attendances increased significantly over the study period. Increasing rate of attendances per 100,000 population per year changed from 0.6 (2000–2010) to 5.5 (2010–2013). This sharp change was driven by CO- and CP-related attendances (rate of CA-related attendance increased steadily). The highest increasing rate (15.6) was for CO-related attendances among 15–29 years old males (2010–2013). Crystal methamphetamine became the most common illicit co-intoxicant amongst cannabis presentations in 2013.

Conclusions

Relative to the total drug-related burden on ambulance services, cannabis-related presentations appear to be a small but significant and increasing problem. Significant changes in trends across other drug involvement and demographic subgroups suggest a possible shift in the cannabis using population and/or a change in using behaviours. Public health strategies should raise awareness of the increased risk posed by cannabis polydrug use and high attendance subpopulations should be determined.

Introduction

Cannabis sativa is the world's most commonly used illicit drug, with approximately 3.8% of the population (or about 178 million people globally) consuming the substance in the last year. The highest prevalence rates are found in North America (12% past-year) followed by Australia and New Zealand (9–11%), then Europe (4–6%) (UNODC, 2014). While the absolute harms attributable to cannabis use have been difficult to quantify due to myriad confounding factors and inherent limitations with study design, several large clinical and epidemiological studies have found that cannabis use is associated with a number of mental and physical health conditions (Aldington et al., 2008, Barber et al., 2013, Callaghan et al., 2013a, Degenhardt et al., 2013, Di Forti et al., 2015, Kuepper et al., 2011, McGrath et al., 2010, Mittleman et al., 2001, Moore et al., 2007, Pletcher et al., 2012, Rossler et al., 2012, Semple et al., 2005, Sherrill et al., 1991, Tetrault et al., 2007) and increased risk of injury through road traffic accidents (Asbridge et al., 2012, Callaghan et al., 2013b, Legrand et al., 2013). Polydrug use is common with national survey data suggesting that over half of young adults use cannabis with alcohol (Pape et al., 2009, Quek et al., 2013) and cannabis use is also strongly associated with the use of cocaine, stimulants, opiates and hallucinogens (Tzilos et al., 2014). Furthermore, polydrug use, in particular heavy alcohol and opiate use, is a significant risk factor for morbidity and mortality in drug-using populations (Degenhardt et al., 2005, Fridell and Hesse, 2006, Gossop et al., 2002, Herbeck et al., 2014), and therefore needs to be taken into consideration when assessing cannabis-related harms.

To date, there have been a limited number of studies assessing the impact of cannabis use on acute healthcare systems. By analysing nursing triage text in an Australian hospital emergency department (ED), Indig et al. (2010) demonstrated that cannabis was a significant factor in 12% of drug-related ED presentations, while Riddell et al. (2008) found that the cannabis-related burden on Australian inpatient services had increased in recent years following analysis of the principal diagnosis across hospital admission records. Roxburgh et al. (2010), who expanded on Riddel's initial work by including alcohol and other drug treatment data with hospital admission records, confirmed that cannabis-related hospital and treatment presentations were increasing across Australia. An observational study in Toulouse, France with a similar methodology, selected cases based on psychoactive substance-related ICD-10 codes contained within hospital discharge summaries (with retrospective review of toxicology reports and/or medical notes review), and found an incidence of 3.2 (CI 2.5–3.9) adverse medical events per 1000 regular cannabis users (Jouanjus et al., 2011).

Such studies highlight the importance of measuring cannabis-related harms and the cannabis-related burden on health services, however there are significant methodological limitations associated with the identification of recreational drug presentations using hospital administrative data and clinical codes (e.g., ICD-10; Shah et al., 2011, Wood et al., 2011). Often very limited information for drug use is available in hospital data. In the study by Indig et al. (2010), about half of the drug-related ED presentations were identified as unspecified drug/s-related. In addition, ICD-10 codes applied in hospital settings may not appropriately reflect the role of substance use due to inaccurate assessment or recording (e.g., coded generically as polydrug use-related events rather than as substance specific codes). Finally, individuals treated by other emergency health staff, such as paramedics, who are not transported to the ED (because of low clinical need or patient refusal) are not included in hospital data. These limitations present considerable challenges for monitoring the true burden of cannabis use on acute health services.

Utilising pre-hospital data for measuring drug-related harms and trends has a number of advantages over ED presentations and hospital admissions data (Clark and Bates, 2003, Dietze et al., 2000, Dietze et al., 2004, Dietze et al., 2008, Heilbronn et al., 2012, Lloyd and McElwee, 2011). Firstly, as the first health responder on the scene, the paramedic assessment is contemporaneous and arguably more likely to capture comprehensive detailed information of drug/s involvement from multiple sources (e.g., relatives, the police and care staff) and see physical evidence of drug use (e.g., drug paraphernalia). Secondly, ambulance records are able to capture patients experiencing severe drug-related harms who do not need to be transported or refuse to be transported. Thirdly, ambulance data differs from data collected in hospital settings in that it contains substantial additional information collected from the patient and others on the scene both as data fields and free text (e.g., description of the incidents, scene management and treatment activities). This allows drug use to be secondary evaluated specifically for the project purpose which reduces the risk of biases introduced by hospital disease category coding practice (e.g., ICD-10 codes). The additional information (e.g., location of presentation, transportation to hospital) also allows multiple outcomes and risk factors to be investigated.

Using data from a unique ongoing surveillance study utilising paramedic records to examine alcohol and other drugs (AOD)-related ambulance attendances in Victoria, Australia, we calculated rates of cannabis-related ambulance attendances from 2000 to 2013 in metropolitan Melbourne based on age and gender. A secondary aim was to determine if attendance rates, location of attendance and transportation to hospital differed among presentations where cannabis use alone was recorded by paramedics, compared to those where cannabis and alcohol use or cannabis and polydrug use was identified.