Circumstances surrounding alcohol toxicity deaths and prior pharmacotherapy for alcohol use disorder in Ontario, Canada

Abstract

Introduction

Despite a high prevalence of alcohol use disorder (AUD) in Canada, access to medication-based treatment remains poor. Therefore, our aim was to explore patterns of alcohol toxicity deaths in Ontario, Canada, circumstances surrounding death, prior healthcare interactions, and pharmacotherapy for AUD.

Methods

We conducted a population-based repeated cross-sectional study of alcohol toxicity deaths occurring between 1 January 2018 and 30 June 2022 in Ontario, Canada. We reported trends in deaths over time and determined demographic characteristics of decedents, circumstances surrounding death, and prior healthcare interactions. Among a subset of the cohort with an AUD diagnosis eligible for public drug benefits, we reported receipt of medications used to treat AUD before death.

Results

We identified 1346 alcohol toxicity deaths over the study period, at a median age of 42 years, with 73.8% occurring among men. The majority of alcohol toxicity deaths involved other substances, including opioids (75.2%), benzodiazepines (10.8%), and/or stimulants (45.2%). Half had an AUD (50.4%) and 62.7% had an opioid, benzodiazepine or stimulant use disorder. Among decedents who were public drug beneficiaries with an AUD (N = 361), only 3.6% were actively prescribed first-line AUD pharmacotherapies (naltrexone and/or acamprosate) at time of death.

Conclusions

We found that the majority of alcohol toxicity deaths in Ontario involved other non-alcohol substances. We also detected a high prevalence of prior healthcare encounters for substance use disorders (SUDs) and low prevalence of evidence-based AUD pharmacotherapy. This suggests a need for integrated treatment across concurrent SUDs and improved access to pharmacotherapies for AUD across Ontario.

Short Summary: Most alcohol toxicity deaths in Ontario involved other substances and occurred among individuals with prior substance use disorders (SUDs). Uptake of first-line alcohol use disorder (AUD) pharmacotherapy was less than 5% among the deceased who were public drug beneficiaries. Addressing concurrent SUDs and improving AUD treatment access is therefore warranted.

Introduction

Alcohol use imposes a considerable public health burden, with 7% of the world’s adult population (400 million individuals) globally living with an alcohol use disorder (AUD) in 2019 (World Health Organization 2024). Alcohol is also one of the most consumed substance in Canada, and despite its regulation, it carries a high risk of adverse health outcomes, including liver disease, heart disease, cancer, depression, injury as well as alcohol tolerance, withdrawal, and addiction (World Health Organization 2018, Canadian Centre on Substance Use and Addiction 2025). About 18% of Canadians meet the criteria for an AUD, meaning that 7.4 million individuals are at high risk for alcohol related harms, including alcohol toxicity (Pearson et al. 2013). Considering that alcohol is a highly consumed substance in Canada, even those without an AUD are at risk of alcohol toxicity. Alcohol toxicity, also referred to as alcohol overdose or poisoning, results when the body’s ability to metabolize alcohol is overwhelmed, which can lead to death (Mirijello et al. 2023). In Ontario, alcohol directly contributed to 13.4% of substance toxicity deaths from 2018 to 2021 (The Ontario Drug Policy Research Network and Public Health Ontario 2023), with a similar percentage observed in the United States (16%, 2020 and 2021) (National Institute on Alcohol Abuse and Alcoholism 2024). Increases in alcohol toxicity deaths were observed in Canada (Statistics Canada 2022, The Ontario Drug Policy Research Network and Public Health Ontario 2023, Yaseen et al. 2025), the United States (National Institute on Alcohol Abuse and Alcoholism 2024), and Australia (Darke et al. 2024) during the COVID-19 pandemic. In addition, increased alcohol consumption among those with AUD was observed in Europe, coinciding with the pandemic (Kilian et al. 2022). Alcohol toxicity deaths increased in parallel with deaths due to other substances such as opioids and stimulants in North America during the pandemic (Government of Canada 2024, Zhao et al. 2024). Importantly, individuals using alcohol alongside opioids or other sedating substances can face heightened risk of harms—including toxicity death—compared to when using either substance alone (Konefal et al. 2022). Further, the permeation of fentanyl into the unregulated drug supply has resulted in increased toxicity deaths across North America, complicating responses for those who use alcohol and unregulated opioids concurrently (Government of Canada 2024, Holton et al. 2024). Understanding the role of opioid, stimulant and benzodiazepine involvement in alcohol use toxicities is crucial for informing appropriate responses.

Prompt AUD diagnosis and treatment is an effective strategy to prevent deaths due to alcohol toxicity. Canadian guidelines for AUD, first released in 2023, strongly recommend that adults with moderate to severe AUD be offered naltrexone or acamprosate as a first-line pharmacotherapy (Wood et al. 2023). As of 28 June 2018, the Ontario Public Drug Program expanded coverage of these first-line therapies, removing administrative barriers to their use (Konstantelos et al. 2023). Canadian guidelines also recommend offering off-label medications such as topiramate and gabapentin for those with moderate to severe AUD where first-line treatments are ineffective, contraindicated or are not preferred by individuals (Wood et al. 2023). The use of pharmacotherapies has been demonstrated to be effective and safe in treating AUD; however, uptake and adherence is a significant challenge (Socias et al. 2023). Research in some Canadian provinces suggest that <1% to 2% of eligible individuals receive pharmacotherapy for AUD (Spithoff et al. 2017, Wood et al. 2023). Similarly low uptake of AUD pharmacotherapy among individuals with AUD is also evident in other countries, such as the United States and Australia, with relatively higher uptake in the United Kingdom and Sweden (Wallhed Finn et al. 2021, Mackillop et al. 2022, Mansson et al. 2024). Although some individuals may prefer not to reduce alcohol consumption, making treatment with AUD pharmacotherapy less appropriate, others may benefit but face barriers to access. For example, many of those with AUD are reluctant to seek these treatments due to stigma or lack of awareness regarding treatment (Gregory et al. 2022, Clarkson 2023). Geographical barriers also make it difficult for people to access treatment in rural or remote Northern areas, while financial barriers may also impede access (Gregory et al. 2022). Therefore, while there is increasing clinical recognition of the importance of pharmacotherapy for the treatment of AUD (Wood et al. 2023), research and interventions for optimizing uptake and adherence are still evolving (Wood et al. 2023, Biswal et al. 2024).

Given the high prevalence of AUD across Canada, the worsening substance toxicity crisis, and the known harms associated with polysubstance use, research is needed to further understand patterns of alcohol-related harms and potential gaps in treatment of AUD among those who died of an alcohol toxicity. Therefore, we sought to explore trends in accidental alcohol toxicity deaths; characteristics and circumstances surrounding these deaths; prior healthcare interactions and prescribing patterns; and receipt of pharmacotherapy for AUD among decedents.

Materials and methods

Setting

We used administrative health data linked with coronial records to conduct a repeated cross-sectional study of all accidental, acute alcohol toxicity deaths occurring between 1 January 2018 and 30 June 2022 in Ontario, Canada. All deaths involved alcohol as a direct contributor to death. Our reporting aligns with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (Vandenbroucke et al. 2007). The use of these data is authorized under section 45 of Ontario’s Personal Health Information Protection Act (PHIPA), which does not require review by a Research Ethics Board.

Data sources

We conducted all analyses using data from ICES, an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. Specifically, we used the Drug and Drug/Alcohol Related Death (DDARD) dataset from the Office of the Chief Coroner in Ontario to report on alcohol toxicity deaths, including circumstances surrounding death as determined by the investigating coroner. In Ontario, all sudden and unexpected deaths are investigated by a medical coroner, with post-mortem toxicology conducted among all suspected substance-related deaths. We used the Registered Persons Database (RPDB) to capture demographic characteristics of the deceased and population denominators. We obtained data on previous healthcare interactions using the Canadian Institute for Health Information’s (CIHI) National Ambulatory Care Reporting System (CIHI-NACRS—for emergency department visits), Discharge Abstract Database (CIHI-DAD—for inpatient hospitalizations), Ontario Mental Health Reporting System (CIHI-OMHRS—for mental health hospitalizations), and the Ontario Health Insurance Plan (OHIP—for outpatient physician visits). The Narcotics Monitoring System (NMS) was used to identify prior prescriptions for monitored substances, capturing all opioids, stimulants, or benzodiazepines dispensed by community pharmacies. To identify prior prescriptions for AUD among individuals eligible for the public drug program who died of an alcohol toxicity, we used the Ontario Drug Benefit (ODB) dataset. These databases were linked using unique encoded identifiers and analyzed at ICES using SAS Enterprise Guide Version 8.3.

Study population

Our cohort included all Ontario residents who died from a confirmed accidental alcohol toxicity from 1 January 2018 and 30 June 2022, identified using data from the DDARD. Our analysis only included cases where alcohol was a direct contributor to death—either alone, or in combination with other substances. Individuals were excluded if their data could not be linked at ICES, if death was non-accidental, or if they had missing or invalid data on age or sex.

Characteristics

We determined a number of demographic characteristics and circumstances surrounding death on the date of death. Demographic characteristics included age group (0 to 24, 25 to 44, 45 to 64, and 65 years and older), sex (male, female, or other), neighborhood income quintile and location of residence (urban vs. rural, and Northern Ontario vs. Southern Ontario). We determined the average yearly alcohol toxicity death rate across the study period by location of residence (per 100 000 population) using population denominators at the mid-point of the study (1 July 2020). For circumstances surrounding deaths, we used coronial records from the DDARD to report location of death and the number and type of substances directly contributing to death (alcohol, opioids, stimulants and/or benzodiazepines).

We also defined healthcare interactions and prescribing patterns prior to death, including non-fatal toxicities, substance use disorder (SUD) diagnoses and prescriptions for monitored substances (opioids, stimulants, benzodiazepines). Any SUD diagnosis was defined as hospital-based encounters (i.e. ED visit, inpatient or mental-health hospitalization) with an alcohol, opioid, benzodiazepine or stimulant diagnosis in the 5 years prior to death; any opioid agonist therapy prescriptions or related outpatient visits in the 5 years prior to death; a general SUD outpatient diagnosis in the 1 year prior to death; or an AUD outpatient diagnosis in the 1 year prior to death (see Supplement Table S1 for classification codes). AUD was defined as any hospital-based encounter (prior 5 years) or outpatient visit (prior 1 year) for AUD. Any opioid, benzodiazepine, or stimulant use disorder was defined as hospital-based encounters (i.e. ED visit, inpatient or mental-health hospitalization) with an opioid, benzodiazepine or stimulant diagnosis in the 5 years prior to death; any opioid agonist therapy prescriptions or related outpatient visits in the 5 years prior to death; or a general SUD outpatient diagnosis in the 1 year prior to death (see Supplement Table S1 for classification codes). Non-fatal toxicities included those treated in an emergency department or hospital (see Supplement Table S2 for classification codes).

Prior prescriptions for alcohol use disorder

To assess treatment with medications for AUD, we restricted our cohort of alcohol toxicity decedents to those with a confirmed AUD diagnosis at time of death (as defined above) and eligible for Ontario’s provincial public drug program (the ODB). In Ontario, residents are eligible for public drug coverage if they are aged <24 without private drug coverage, are receiving social assistance or disability support, reside in a long-term care home, have high medication costs compared to income, or are aged 65 years or older. We determined ODB eligibility as all individuals who received a publicly funded medication claim in the ODB database in the 181 to 365 days prior to death (to ensure at least 6 months of continuous eligibility). We reported exposure to any AUD pharmacotherapy, and then reported separately as exposure to: (i) first-line pharmacotherapy (naltrexone and/or acamprosate) (ii) second-line pharmacotherapy (topiramate and/or gabapentin). We assessed the presence of an AUD prescription 1 year prior to death, thirty days prior to death, and with a day’s supply overlapping with the day of death.

Analysis

We reported monthly population-adjusted rates of alcohol toxicity deaths in Ontario over the study period, per 100 000 population. We also reported the average yearly population-adjusted alcohol toxicity death rate by location of residence using monthly data across the study period. We used descriptive statistics (percentages and medians with interquartile ranges) to summarize all other characteristics.

Results

We identified 1346 accidental alcohol toxicity deaths over the study period. The monthly rate of alcohol toxicity deaths (per 100 000 population) was 0.13 in January 2018, doubling to 0.29 at the peak in June 2020 (Fig. 1). This was followed by a decline at the start of 2021 and stabilization until the end of the study period in June 2022 (0.13 per 100 000 population).

Figure 1.

Alcohol toxicity death rate per 100 000 population from January 2018 to June 2022

Average alcohol toxicity death rates per year in rural Ontario (2.1 per 100 000 population) were similar to urban Ontario (2.0 per 100 000 population; P = .55, Table 1). Average yearly death rates in Northern Ontario (5.2 per 100,000 population) were nearly triple that of Southern Ontario (1.8 per 100 000 population, P < .00001). The median age at time of alcohol toxicity death was 42 years (interquartile range 34 to 53), with the highest proportion of deaths occurring among young (25 to 44 years; 51.6%) and middle-aged adults (45 to 64 years; 39.7%; Table 2). Overall, 73.8% of deaths occurred among males, and deaths were disproportionately concentrated in neighborhoods with lower incomes (39.9% in lowest income quintile). The majority (73.2%) of alcohol toxicity deaths occurred in private residences, while 5.4% occurred in rooming houses (or other collective dwellings), 2.8% in shelters (or other residential settings), and 7.4% outdoors. Only 16.1% of deaths involved alcohol alone, with 39.5% of deaths involving alcohol and one other substance, and 44.4% involving alcohol with two or more other substances. Opioids, stimulants and benzodiazepines were involved in 75.2%, 45.2%, and 10.8% of all alcohol toxicity deaths, respectively.

Table 1.

Average yearly alcohol toxicity death rate by location of residence from January 2018 to June 2022 (N = 1346).

Location of residence	N, rate per 100 000	P-value
Urban/rural
Urban	1190 (2.0)	0.55
Rural	139 (2.1)
Northern/Southern
Northern	195 (5.2)	<0.00001*
Southern	1151 (1.8)

N: number. Unknown/missing data comprised of 1.3% of the cohort for urban/rural location. The midpoint of the study time period (1 July 2020) was used to determine population denominators. Asterisk with bolded font indicates P < .05.

Table 2.

Characteristics and circumstances surrounding death among individuals who died of an alcohol toxicity from January 2018 to June 2022 (N = 1346).

Characteristic
Age (median, IQR)	42 (34–53)
Age category (N, %)
0–24	67 (5.0%)
25–44	694 (51.6%)
45–64	534 (39.7%)
65+	51 (3.8%)
Sex (N, %)
Female	353 (26.2%)
Male	993 (73.8%)
Income quintile (N, %)
1 (lowest)	537 (39.9%)
2	296 (22.0%)
3	202 (15.0%)
4	167 (12.4%)
5	126 (9.4%)
Location of death (N, %)
Private residence	985 (73.2%)
Rooming house or other collective dwellings	73 (5.4%)
Shelters or other residential settings	38 (2.8%)
Other indoor space (non-hospital)	108 (8.0%)
Outdoors	100 (7.4%)
Othera	27 (2.0%)
Number of substances involved in death (N, %)
1	217 (16.1%)
2	532 (39.5%)
≥3	597 (44.4%)
Other substances contributing to death (N, %)
Opioids	1012 (75.2%)
Stimulants	608 (45.2%)
Benzodiazepines	145 (10.8%)

N: number; IQR: inter-quartile range. Unknown/missing data comprised of 1.3% and 1.1% of the cohort for income quintile and location of death, respectively.

^aOther includes hospitals, in transit and other locations not otherwise specified.

Non-fatal alcohol, opioid, stimulant and benzodiazepine toxicity 1 year prior to death was experienced by 3.3%, 9.7%, 2.6%, and 1.9% of the cohort, respectively (Table 3). Overall, 63.7% of individuals who died of an alcohol toxicity were diagnosed with a SUD, with 50.4% diagnosed with an AUD and 62.7% diagnosed with an opioid, benzodiazepine or stimulant use disorder in the 5 years prior to death. Receipt of prescription opioids and benzodiazepines in the year prior to death occurred among 34.8% and 30.3% of individuals, respectively, whereas only 4.3% of people were dispensed stimulants in the year prior. In the 30 days prior to death, 17.2% and 15.6% of people were dispensed opioids or benzodiazepines, respectively.

Table 3.

Healthcare interactions and prescribing patterns among individuals who died of an alcohol toxicity from January 2018 to June 2022 (N = 1346).

Prior toxicities, 1 year (N, %)
Alcohol	44 (3.3%)
Opioid	130 (9.7%)
Stimulant	35 (2.6%)
Benzodiazepine	26 (1.9%)
SUD diagnoses (N, %)
Any SUD, including AUDa	857 (63.7%)
AUD diagnosisb	679 (50.4%)
Alcohol use-related outpatient visit (1 year   prior)	216 (16.0%)
Alcohol use-related hospital diagnosis (5 years   prior)	654 (48.6%)
Any opioid, benzodiazepine or stimulant use disorderc	844 (62.7%)
Prescriptions for monitored substances dispensed prior to death (N, %)
Any opioid
1 year	468 (34.8%)
30 days	231 (17.2%)
Stimulants
1 year	58 (4.3%)
30 days	28 (2.1%)
Benzodiazepines
1 year	408 (30.3%)
30 days	210 (15.6%)

^aSUD diagnosis is defined as any hospital-based encounter for alcohol, opioid, stimulant, or benzodiazepine use disorder in the past 5 years; opioid agonist therapy prescription or related outpatient visit in the past 5 years; general SUD outpatient diagnoses in the past year; or AUD outpatient visit in the past year.

^bAny AUD diagnosis is defined as a hospital-based encounter with AUD diagnosis in the past 5 years or AUD outpatient diagnoses in the past year.

^cAny opioid, benzodiazepine or stimulant use disorder diagnosis is defined as any hospital-based encounter for opioid, stimulant, or benzodiazepine use disorder in the past 5 years; opioid agonist therapy prescription or related outpatient visit in the past five years; or general SUD outpatient diagnoses in the past year.

In our analysis of prior treatment for AUD, we identified 361 accidental alcohol toxicity deaths among people with an AUD who were eligible for public drug benefits (26.8% of the full cohort). Among these individuals, 16.1% (N = 58) received a first-line treatment for AUD (naltrexone and/or acamprosate) in the year before death, which reduced to only 4.7% (N = 17) in the 30 days before death and 3.6% (N = 13) actively treated with one of these medications at time of death (Fig. 2). In comparison, 26.0% (N = 94) received second-line treatment options (topiramate and/or gabapentin) in the year prior to death, with 13.3% (N = 48) and 10.0% (N = 36) receiving one of these medications in the 30 days prior or actively on the same day of death, respectively.

Figure 2.

Prior receipt of pharmacological AUD treatment among public drug beneficiaries with an AUD who died of alcohol toxicity (N = 361). N: number. Bar labels are ordered as follows for each drug grouping, from left to right: 1 years, 30 days, same day (i.e. prescription overlapping with the death date)

Discussion

In our population-based study in Ontario, Canada, we identified 1346 alcohol toxicity deaths over 4.5 years following January 2018. Rates of alcohol toxicity deaths have fluctuated over time, with a peak in deaths three months after the declaration of the COVID-19 state of emergency in Ontario (June 2020, .29 per 100 000 population). This aligns with findings of other studies around the world, which have shown increased alcohol consumption, alcohol-related harms (Roberts et al. 2021, Zipursky et al. 2021, Yaseen et al. 2025), alcohol purchasing (Zipursky et al. 2021; Yaseen et al. 2025), and use of other non-alcohol substances during the COVID pandemic (Roberts et al. 2021). Factors such as heightened stress and increased isolation have been described as catalysts for alcohol and other substance use, contributing to higher toxicity deaths during the pandemic (Grigoletto et al. 2020, Callinan et al. 2021, Roberts et al. 2021).

Our results suggest that alcohol toxicity deaths most often involve other substances, with 84% of deaths involving either opioids, stimulants, and/or benzodiazepines as direct contributors to death. Three-quarters of alcohol toxicity deaths had opioids identified as direct contributors to death, and nearly half had stimulants directly contributing to deaths. Notably, despite the high proportion of alcohol toxicity deaths involving opioids, only 17.2% of individuals who died of alcohol toxicity had an opioid prescription in the 30 days prior to death, suggesting that unregulated opioids, rather than prescribed opioids, were more frequently involved in these deaths. This aligns with previous research in Ontario which suggests that 86% of opioid toxicity deaths involved non-pharmaceutical opioids between 2018 and 2022, as well as the impacts of increasing fentanyl into the unregulated drug supply during this period (Government of Canada 2024, Holton et al. 2024). Our results also reflect the rising trend of other substance involvement in alcohol toxicity deaths in Canada, which increased from 49% in 2014 to 74% in 2017 (Konefal et al. 2022), and broader patterns of rising polysubstance post-mortem detection reported throughout Canada and the United States (Konefal et al. 2022). Further, diagnosis of other non-alcohol SUDs was relatively common, with nearly two-thirds of deaths occurring among people with opioid, benzodiazepine or stimulant use disorders. These results highlight the importance of tailored strategies that consider polysubstance use and concurrent SUDs among those experiencing alcohol-related harms. Improved screening and integrated treatment for multiple SUDs, and expansion of harm reduction services tailored to those who use multiple substances, may be warranted. Given high rates of alcohol use in the population compared to other substances, widespread public health campaigns can be implemented to educate the public about the risks of using multiple sedating substances and how to implement harm reduction approaches to mitigate these risks.

Alcohol toxicity deaths are not distributed equally across demographic groups or geography in Ontario. In particular, we observed income and geographical inequities, where deaths due to alcohol toxicity were concentrated among those residing in lower income neighborhoods and with population-adjusted rates being nearly three times higher in Northern compared with Southern Ontario. Socioeconomic status is a predictor of AUD (Lasserre et al. 2022) and alcohol-related emergencies (Benny et al. 2024), which may result from lower access to mental health resources (e.g. counseling and rehabilitation programs), an increased need to cope due to financial concerns, lifestyle risk factors and the exacerbation of comorbidities which disproportionally impact individuals of lower socioeconomic status (Lasserre et al. 2022, Benny et al. 2024). Further, mental health and substance use delivery disparities are common in Northern Ontario as these communities are generally more remote compared to the more population-dense Southern Ontario (Canadian Mental Health Association 2009). Anonymity and confidentiality may also be more difficult for individuals in small remote communities, deterring individuals from seeking treatment for AUD and alcohol intoxication (Wolfe et al. 2023). Importantly, First Nations communities are concentrated in Northern Ontario (Statistics Canada 2018), and First Nations populations experience disproportionally higher rates of SUDs and related harms, driven by intergenerational trauma stemming from colonialization, forced relocation, and the legacy of residential schools (Haggarty, n.d., Lavalley et al. 2018, Russell et al. 2019, Chiefs of Ontario and Ontario Drug Policy Research Network 2023). Additionally, systemic racism in healthcare settings and a lack of culturally safe services further contribute to disparities in access to appropriate treatment for alcohol and SUDs. Given these challenges, there is a need for Indigenous-led research that further explores factors associated with alcohol-related harms and potential responses, which include community-led, culturally grounded programs among Northern communities, in addition to policies that mitigate the geographical and financial burden of accessing healthcare and substance use treatment.

Among public drug beneficiaries with an AUD, we noted that 16.1% were dispensed naltrexone or acamprosate (the first-line treatments for AUD) in the previous year, dropping to only 3.6% actively dispensed these treatments at time of death. In comparison, previous research among public drug beneficiaries in Ontario from 2011 to 2012 demonstrated that <1% of individuals with an AUD diagnosis were dispensed first-line AUD medications in the previous year (Spithoff et al. 2017). The higher prevalence of AUD pharmacotherapy in our study likely reflects more severe AUD among alcohol toxicity decedents, which has been associated with higher rates of AUD pharmacotherapy use (Spithoff et al. 2017). When considering both first- or second-line AUD treatments dispensed in the past year, the percentage rose to 34.6%; however, because second-line treatments are used for AUD off-label, we cannot definitively conclude whether these second-line treatments were prescribed for AUD or for other ailments, such as epilepsy and pain-related conditions (National Health Service, n.d.), resulting in potential overestimation. Additionally, the higher prevalence of second- versus first-line pharmacotherapies may also reflect high AUD severity and other SUDs in our sample, as second-line treatments are often considered for individuals who do not benefit from, or have contraindications to, first-line options (e.g. naltrexone contraindicated with opioid use) (Wood et al. 2023).

Nonetheless, our study demonstrates that a minority of individuals received either a first- or second-line AUD treatment prior to alcohol toxicity death, highlighting low uptake and potential treatment gaps. Several barriers contribute to these treatment gaps, including knowledge deficits concerning AUD pharmacotherapy among clinicians (e.g. the first Canadian clinical guideline for AUD was not released until 2023) (Wood et al. 2023), limited clinical time, and addiction-related stigma from clinicians toward patients (Gregory et al. 2022, Johnson et al. 2022). At the individual-level, barriers to AUD pharmacotherapy include fear of stigmatization, feelings of shame, efficacy concerns, and a preference for managing alcohol problems independently (Gregory et al. 2022, Wolfe et al. 2023). Further, access to AUD pharmacotherapy has been hindered by drug shortages, financial constraints, and geographical barriers (Gregory et al. 2022). For example, an acamprosate shortage (February 2019 to July 2020) led to a 98.1% reduction in acamprosate use in Ontario (Konstantelos et al. 2023). These challenges highlight the need to improve access to, and education regarding, AUD pharmacotherapy. Lastly, given that nearly half of individuals who died from alcohol toxicity had been previously hospitalized with an AUD diagnosis identified during hospitalization, and considering evidence supporting AUD pharmacotherapy at hospital discharge (Bernstein et al. 2024), hospital discharge may represent a prime opportunity to initiate AUD pharmacotherapy and prevent future deaths.

A major strength of our study is its use of post-mortem toxicological findings from coronial data to accurately identify and characterize all deaths due to acute alcohol toxicity in Ontario, Canada. However, some limitations are worth noting. First, our dataset did not include deaths caused by the long-term health impacts of heavy alcohol use (e.g. liver cirrhosis, heart problems, stroke, cancer and mental health conditions) or alcohol-related accidents (e.g. impaired driving deaths, drownings as a result of alcohol impairment) as this data was unavailable through the coronial dataset used. Further, our study could not capture societal harms associated with alcohol, including domestic violence, crime and intergenerational trauma. Thus, our cohort is only generalizable to deaths directly caused by alcohol toxicity, representing a subset of a larger population that experienced alcohol-related harm.

In addition, we restricted our analysis of prior AUD pharmacotherapies to public drug beneficiaries due to data limitations, and therefore we are unable to characterize recent treatment patterns among decedents who may have accessed these medications through private insurance or out-of-pocket payments. In particular, this restriction impacts the generalizability of our findings related to AUD pharmacotherapy to individuals receiving public drug benefits. This would include individuals aged 65 years and older, aged <24 without private drug coverage, low income individuals, those with high medication costs, those receiving professional home and community care services, and those living in special care or long-term care homes (Bouck et al. 2024). Note that this limitation only applies to Fig. 2.

Also, we could not identify Indigenous People through our reporting, limiting our ability to specifically describe and interpret the rate of alcohol toxicities among Indigenous populations. Future efforts should prioritize meaningful engagement of Indigenous People, including engagement with First Nations communities which are concentrated in Northern Ontario. Engagement should adhere to OCAP® principles, which emphasize First Nations ownership, control, access, and possession of their data.

Conclusion

Alcohol use continues to impose a considerable public health burden across Canada, with over 1300 people dying from acute alcohol toxicities between 2018 and 2022 in Ontario. Importantly, these deaths often involve non-alcohol substances and occur frequently among people with AUD and other SUDs. Low uptake of first-line pharmacotherapy for AUD was also observed among public drug beneficiaries. These findings suggest the need for adaptive and integrated strategies that address the complexity of concurrent SUDs while minimizing barriers to AUD treatment. Tailored approaches should prioritize screening and integrated care models for concurrent SUDs and expand harm reduction services for these individuals. Importantly, efforts to improve knowledge of, and access to pharmacotherapies for AUD are warranted. To prevent further alcohol toxicity deaths, public health and policy efforts can expand harm reduction initiatives; promote evidence-based AUD pharmacotherapy, particularly at hospital discharge; and reduce barriers for individuals seeking substance use treatment.

Author contributions

William Wynne (Writing—original draft [lead], Writing—review & editing [lead]), Anita Iacono (Writing—original draft [lead], Writing—review & editing [lead]), Joanna Yang (Data curation [equal], Formal analysis [equal], Methodology [equal], Writing—review & editing [equal]), Bisola Hamzat (Writing—review & editing [equal]), Pamela Leece (Writing—review & editing [equal]), Gillian Kolla (Writing—review & editing [equal]), Rob Boyd (Writing—review & editing [equal]), Nikki Bozinoff (Writing—review & editing [equal]), Mike Franklyn (Writing—review & editing [equal]), Dana Shearer (Project administration [equal], Writing—review & editing [equal]), Ashley Smoke (Writing—review & editing [equal]), Fangyun Wu (Data curation [equal], Methodology [equal], Supervision [equal]), Paul Newcombe (Writing—review & editing [equal]), Tara Gomes (Conceptualization [equal], Funding acquisition [equal], Investigation [equal], Project administration [equal], Supervision [equal], Writing—review & editing [equal]).

Conflict of interests

Tara Gomes reports financial support from the Canadian Institutes of Health Research. Tara Gomes and Pamela Leece report financial support from the Public Health Agency of Canada. Mike Franklyn reports a relationship with the Northern Ontario School of Medicine that includes employment. Paul Newcombe and Ashley Smoke report a relationship with Public Health Agency of Canada which includes honorarium. Ashley Smoke reports board work for the Ontario Network for People Who Use Drugs and Canadian Association of People who Use Drugs; consultation or advisory work at St. Peter’s Centre, Canadian AIDS Treatment Information Exchange (CATIE), Public Health Agency of Canada (PHAC), Unity Health Toronto, Centre for Addiction and Mental Health (CAMH), University of Toronto, Mount Sinai, Chiefs of Ontario; employment at Native Women’s Association of Canada; speaker fees for CATIE, Anishinabek Nation, Chiefs of Ontario, Ontario Drug Policy Research Network (ODPRN), Dr Peter’s Centre, Haliburton, Kawartha, Pine Ridge District Health Unit (HKPR District Health Unit), Government of the Attorney General, Public Health Ontario, and Canadian Liver Foundation.

The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding

This work was supported in part by the Public Health Agency of Canada, Substance-Related Harms Division, and grants from the Canadian Institutes of Health Research (grant number 178163). T. Gomes is supported by a Tier 2 Canada Research Chair. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.

Data availability

The dataset from this study is held securely in coded form at ICES. While legal data sharing agreements prohibit ICES from making the dataset publicly available, access may be granted to those who meet pre-specified criteria for confidential access, available at http://www.ices.on.ca/DAS (email: das@ices.on.ca).