Substance-related poisonings and homelessness in Canada: a descriptive analysis of hospitalization data

Canada is currently experiencing an overdose crisis affecting people from all walks of life. The Public Health Agency of Canada (PHAC), in collaboration with the provincial and territorial (PT) offices of Chief Coroners and Chief Medical Examiners, PT public health and health partners, and Emergency Medical Service data providers, releases quarterly surveillance data on apparent opioid and stimulant toxicity (overdose) deaths and Emergency Medical Services responses for suspected opioid-related poisonings (overdoses) ^{Footnote 1}. PHAC also collaborates with Health Canada to report on hospitalizations for opioid- and stimulant-related poisonings (overdoses) using hospital administrative data from the Canadian Institute for Health Information (CIHI) ^{Footnote 1}. In addition to ongoing surveillance, PHAC releases targeted analytical reports to help fill gaps identified by stakeholders on specific topics related to substance-related harms.

There are many factors that can play a role in substance use and related harms, including a person's living situation. It is estimated that in Canada, there are on average at least 235,000 people experiencing homelessness in a given year and at minimum 35,000 on a given night ^{Footnote 2}. Across the country, an additional 50,000 people a night could be experiencing hidden homelessness ^{Footnote 2}. Hidden homelessness refers to people who are temporarily staying with friends, relatives or others because they have nowhere else to live and no immediate prospect of permanent housing ^{Footnote 2}. The number of people experiencing homelessness in Canada remains very difficult to estimate, yet this number is thought to be increasing ^{Footnote 2Footnote 3}. The rates of substance use are disproportionally high among people experiencing homelessness compared to people with secure housing ^{Footnote 4Footnote 5Footnote 6Footnote 7Footnote 8} and people experiencing homelessness are at a greater risk of substance-related harms ^{Footnote 9Footnote 10Footnote 11}. Further, during the COVID-19 pandemic, some disparities in health have widened, particularly among some hard to reach populations, including an increase in the number of people experiencing homelessness, as well as an increase in the number of substance-related poisonings across the country ^{Footnote 12Footnote 13}.

It recently became mandatory to capture homelessness status upon admission to hospitals in the CIHI's national hospitalization data effective 2018 ^{Footnote 14}. In response to this newly available data, the objective of this analysis is to describe patterns of substance-related poisoning hospitalizations in Canada (excluding Quebec) among people with housing and people experiencing homelessness using the CIHI's Discharge Abstract Database. Data are presented for the time period from April 1, 2019 to March 31, 2020, and therefore most of the data included in this analysis captures the pre-pandemic period. This report examines patterns by patient demographics (sex and age), context of the poisoning (substances involved and intention of poisoning), hospitalization characteristics and outcomes (length of stay, intensive care unit admission and discharge disposition), and co-diagnosed mental health conditions. This analysis includes poisonings due to opioids, stimulants, cannabis, hallucinogens, alcohol, other depressants, and psychotropic drugs. To our knowledge, substance-related poisoning hospitalizations among people experiencing homelessness have not yet been reported publicly at a national level using this data source. The results of this analysis can be used to help better understand the intersection of homelessness, mental health, and substance-related harms and how hospital care is experienced differently by people experiencing homelessness.

Key Findings

Substance-related poisoning hospitalizations among people experiencing homelessness

Between April 2019 and March 2020, in Canada (excluding Québec), there were 10,659 substance-related poisoning hospitalizations. Of those, approximately 6% (n=623) were among people experiencing homelessness on admission.

Sex and age

The majority of substance-related poisoning hospitalizations for people experiencing homelessness were among males (71%). This may reflect the higher proportion of men who experience homelessness ^{Footnote 2}, males experiencing higher rates of substance use disorders compared to females ^{Footnote 16}, in addition to potentially more women experiencing hidden homelessness ^{Footnote 17}. In contrast, a more even distribution of substance-related poisonings was observed between males (47%) and females (53%) among people with housing (Figure 1).

Figure 1 - Text Equivalent

Substance-related poisoning hospitalization	Male	Female
	% (N)	% (N)
People with housing	47% (4,715)	53% (5,316)
People experiencing homelessness	71% (441)	29% (182)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) 'Other sex' is excluded due to low numbers. Refer to the limitations section for more information regarding sex and gender data in the DAD.

Among those who were experiencing homelessness, most substance-related poisoning hospitalizations occurred in the 20-29 (23%), 30-39 (33%) and 40-49 (20%) age groups (Figure 2). In contrast, among people with housing, there was a more even distribution across age groups from 20-29 to 70+ years. This may reflect the relatively younger population among those experiencing homelessness ^{Footnote 18}.

Figure 2 - Text Equivalent

Substance-related poisoning hospitalization	0-19 Years	20-29 Years	30-39 Years	40-49 Years	50-59 Years	60-69 Years	70+ Years
	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)
People with housing	11% (1,146)	19% (1,915)	18% (1,819)	14% (1,435)	16% (1,619)	11% (1,149)	9% (953)
People experiencing homelessness	2% (12)	23% (143)	33% (207)	20% (126)	14% (88)	6% (36)	2% (11)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).

Generally, among hospitalizations for substance-related poisonings, people experiencing homelessness were younger compared to people with housing. The median age for people experiencing homelessness was 37 years (mean = 39), while the median age for people with housing was 40 years (mean = 43).

Hospitalization characteristics and outcomes

Length of stay

When examining the length of stay in hospital among people with housing and people experiencing homelessness, a substantial difference was observed (Figure 3). The total length of stay in hospital adds the number of days a patient was in acute inpatient care (patients receiving necessary treatment for a disease or severe episode of illness for a short period) and alternate level of care (patients occupying a bed but not requiring the intensity of services provided in that care setting) ^{Footnote 19Footnote 20}. Overall, the average total length of stay in hospital for a substance-related poisoning was longer among people experiencing homelessness. People experiencing homelessness hospitalized for substance-related poisonings spent an average of 7.3 days in acute inpatient care and 3.7 days in alternate level of care, for a total of 11.0 days spent in hospital. In comparison, people with housing hospitalized for substance-related poisonings spent an average of 5.8 days in acute inpatient care and 0.8 days in alternate level of care, for a total of 6.6 days spent in hospital. Please see Appendix A for information regarding median length of stay.

Figure 3 - Text Equivalent

Substance-related poisoning hospitalization	Mean length of stay in acute inpatient care	Mean length of stay in alternate level of care	Mean total length of stay
People with housing	5.8	0.8	6.6
People experiencing homelessness	7.3	3.7	11.0
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).

Intensive care unit admission

The Intensive Care Unit (ICU) is an organized system to provide specialized medical care to patients who are critically ill, and allow for enhanced monitoring and life support ^{Footnote 21}. This care could involve mechanical ventilation to help a person breathe if they are unable to do so on their own. A similar proportion of hospitalizations were admitted to the ICU when comparing people with housing and people experiencing homelessness with a substance-related poisoning. Among people experiencing homelessness, 38% were admitted to the ICU compared to 36% of people with housing.

Discharge disposition

Discharge disposition describes the status of the patient on discharge or where the patient was discharged to, and varies between people with housing and people experiencing homelessness (Figure 4). A notable difference was observed in the proportion who left the hospital against medical advice. For substance-related poisonings among people with housing, 8% left the hospital against medical advice. In comparison, for substance-related poisonings among people experiencing homelessness, almost a quarter (23%) left the hospital against medical advice. Patients who leave against medical advice may not have received all of the medical care or supports available.

The majority (68%) of substance-related poisoning hospitalizations among people with housing indicated that the patient was discharged home. In comparison, 49% among people experiencing homelessness were "discharged home". However, it is important to note this category also includes patients who were experiencing homelessness on admission and refused shelter upon discharge, as there was not a value to capture this concept in the data. As a result, "home" in these instances could refer to discharging a patient to living on a street or place that is not intended for human habitation (e.g., park, car, sidewalk).

A similar percentage of substance-related poisoning hospitalizations among people with housing (21%) and people experiencing homelessness (24%) resulted in the patient being transferred, for instance, to a specialty hospital for inpatient rehabilitation, mental health or addiction treatment centre, long-term care, shelter or assisted living with supportive housing. Upon examining those who fall within this category further, 12% of people experiencing homelessness were transferred to a shelter, compared to only 3% of people with housing on admission.

No differences were observed between those with housing and those experiencing homelessness for deaths occurring in hospital.

Figure 4 - Text Equivalent

Substance-related poisoning hospitalization	Home (includes patients experiencing homelessness on admission who refuse shelter upon discharge)	Transferred (includes transfers to shelters or supportive/transitional housing)	Left against medical advice absent or did not return from pass	Died in hospital
	% (N)	% (N)	% (N)	% (N)
People with housing	68% (6,829)	21% (2,069)	8% (786)	4% (352)
People experiencing homelessness	49% (306)	24% (149)	23% (142)	4% (26)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) The "home" category also includes patients experiencing homelessness on admission and refused shelter upon discharge, as there was not a value to capture this concept in the data. As a result, "home" in these instances could refer to discharging a patient to living on a street or place that is not intended for human habitation (e.g., park, car, sidewalk).

Substances involved and intentionality

Type of substance(s) involved

Differences were observed when examining the substance(s) involved in the poisoning for people with housing and people experiencing homelessness (Figure 5). For substance-related poisoning hospitalizations among people experiencing homelessness, opioids were the most common type of substance involved. Opioids were present in 61% of poisoning hospitalizations among people experiencing homelessness and in 40% of poisoning hospitalizations among people with housing. Stimulants (e.g., methamphetamine) were also involved in a greater percentage of substance-related poisoning hospitalizations among people experiencing homelessness (29%) than among people with housing (19%). Overall, for substance-related poisoning hospitalizations among people with housing, cannabis, alcohol and other central nervous system depressants (e.g., benzodiazepines and other sedatives) were more commonly involved compared to people experiencing homelessness. Of note, multiple types of substances may be involved in any given poisoning and therefore percentages will not add up to 100%. Refer to the technical notes and Appendix B for the methodology used to identify types of substance-related poisoning hospitalizations.

Figure 5 - Text Equivalent

Substance-related poisoning hospitalization	Opioid	CNS† Stimulant	Other CNS† Depressant	Alcohol	Cannabis	Hallucinogen	Other/unspecified psychotropic drug
	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)
People with housing	40% (4,056)	19% (1,953)	39% (3,897)	23% (2,300)	6% (594)	<1% (35)	1% (67)
People experiencing homelessness	61% (377)	29% (182)	19% (118)	18% (114)	4% (24)	Suppressed	Suppressed
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) *Counts less than five have been suppressed as per CIHI privacy guidelines. †CNS = Central nervous system Multiple types of substances may be involved in any given poisoning. Refer to the technical notes and Appendix B for information regarding ICD-10-CA codes included and methodology.

Type of opioid(s) involved

Overall, the opioid(s) involved in the substance-related poisoning hospitalization differed between people with housing and people experiencing homelessness (Figure 6). Most notably, among people experiencing homelessness, 34% of opioid-related poisoning hospitalizations involved fentanyl and its analogues, whereas 20% of such hospitalizations were observed among people with housing. Heroin was also involved in a greater proportion of opioid-related poisoning hospitalizations among people experiencing homelessness than among people with housing, at 15% and 7% respectively. While fentanyl and heroin were most commonly observed among opioid-related hospitalizations for people experiencing homelessness, other opioid types were more prevalent among people with housing. In particular, tramadol, methadone, oxycodone, hydromorphone, morphine, opium and codeine were more commonly involved among people with housing. Of note, the category for other/unspecified opioids represented the highest proportion for both people with housing and people experiencing homelessness. This finding may reflect further confirmatory testing to distinguish the specific type of opioid being unavailable or not necessary for treatment, new and emerging substances not being captured in the toxicology screen, or an inability or unwillingness of the patient to disclose the substance used.

It was not possible to determine which poisonings were a result of pharmaceutical opioids, non-pharmaceutical opioids, or a combination of both. Multiple types of opioids may be involved in any given poisoning and therefore percentages will not add up to 100%. Refer to the technical notes and Appendix B for the methodology used to identify types of opioid-related poisonings.

Figure 6 - Text Equivalent

Substance-related poisoning hospitalization	Fentanyl†	Heroin	Methadone	Oxycodone	Codeine†	Hydromorphone	Tramadol	Morphine	Opium	Other/Unspecified§
	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)
People with housing	20% (792)	7% (272)	8% (341)	10% (393)	10% (424)	10% (393)	3% (115)	7% (284)	<1% (8)	35% (1,415)
People experiencing homelessness	34% (130)	15% (57)	6% (22)	3% (13)	3% (11)	2% (7)	Suppressed	Suppressed	0% (0)	41% (154)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) * Counts less than five have been suppressed as per CIHI privacy guidelines. † And its derivatives § Includes other opioids not elsewhere classified / other synthetic narcotics / unspecified narcotics. Refer to the technical notes and Appendix B for information regarding ICD-10-CA codes included and methodology. Multiple types of opioids may be involved in any given poisoning.

Multiple substance poisoning

Substance-related poisoning hospitalizations which involved more than one substance were relatively similar among people with housing and people experiencing homelessness. The involvement of only one substance of interest accounted for 77% of all substance-related poisoning hospitalizations among people with housing, and 74% of all substance-related poisoning hospitalizations among people experiencing homelessness. For people with housing, 19% of substance-related poisoning hospitalizations involved two substances and 5% of substance-related poisoning hospitalizations involved three or more substances. Similarly, for people experiencing homelessness, 20% of substance-related poisoning hospitalizations involved two substances and 5% of substance-related poisoning hospitalizations involved three or more substances. Of note, the following substance types were explored for the multiple poisoning analysis: opioids, stimulants, cannabis, hallucinogens, alcohol, other depressants, and psychotropic drugs. Substances outside of this scope, for example poisonings that also involved anti-depressants, are not reported.

Intention

Intention refers to the documented reason for the poisoning, which may rely on self-report, and is classified as accidental, intentional self-harm and undetermined. Definitions and methodology are in the technical notes and Appendix C. People experiencing homelessness had a higher proportion of substance-related poisoning hospitalizations recorded to be accidental in nature compared to people with housing, 62% vs. 45% respectively (Figure 7). Conversely, people with housing had a higher proportion of substance-related poisoning hospitalizations recorded to be intentional self-harm compared to people experiencing homelessness, 46% vs. 26% respectively. Similar proportions were observed for substance-related poisoning hospitalizations of undetermined intention between people experiencing homelessness and people with housing.

Figure 7 - Text Equivalent

Intention for the poisoning	Accidental	Intentional self-harm	Undetermined
	% (N)	% (N)	% (N)
People with housing	45% (4,519)	46% (4,551)	10% (949)
People experiencing homelessness	62% (382)	26% (162)	12% (76)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) Multiple types of substances may be involved in a given poisoning. A small volume of abstracts had two different intention types and therefore summed percentages exceed 100%. Records where one or more poisonings are missing an associated intention code are excluded. Refer to the technical notes and Appendix C for information regarding ICD-10-CA codes included and methodology.

Intention by sex

Among males, accidental substance-related poisoning hospitalizations were most common for both people with housing (54%) and people experiencing homelessness (65%). Among females with housing, intentional substance-related poisoning hospitalizations were most common (56%), while for females experiencing homelessness, accidental substance-related poisoning hospitalizations were most common (54%) (Figure 8).

Figure 8 - Text Equivalent

Intention for poisoning	Accidental	Intentional self-harm	Undetermined
	% (N)	% (N)	% (N)
Females with housing	38% (1,986)	56% (2,928)	8% (398)
Females experiencing homelessness	54% (97)	37% (67)	10% (18)
Males with housing	54% (2,532)	35% (1,620)	12% (550)
Males experiencing homelessness	65% (285)	22% (95)	13% (58)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) 'Other sex' is excluded due to low numbers. Refer to the limitations section for more information regarding sex and gender data in the DAD. Multiple types of substances may be involved in any given poisoning. A small volume of abstracts had two different intention types and therefore summed percentages exceed 100%. Records where one or more poisonings are missing an associated intention code are excluded. Refer to the technical notes and Appendix C for information regarding ICD-10-CA codes included and methodology.

Co-Diagnosed Mental Health Conditions

Records were explored for co-diagnoses of mental health conditions (referred to as mental disorders in the ICD-10-CA) during the hospital stay for the poisoning; more detailed definitions and methodology are available in the technical notes and Appendix D. The presence of co-diagnosed mental health conditions during the hospital stay for a substance-related poisoning was high among both people with housing and people experiencing homelessness. However, people experiencing homelessness had a higher proportion of mental health conditions recorded during their hospital stay compared to people with housing, at 61% vs. 52% respectively (Figure 9).

The most commonly documented co-diagnosed mental health conditions for both people experiencing homelessness and people with housing were substance-related and addictive disorders and mood disorders, however patterns within these categories varied. People experiencing homelessness had more than double the proportion of co-diagnosed substance-related and addictive disorders compared to people with housing (51% vs. 25% respectively). People with housing had a substantially higher proportion of co-diagnosed mood disorders compared to people experiencing homelessness (21% vs. 11% respectively). People with housing also had a higher proportion of co-diagnosed anxiety disorders (9% as compared to 3%). While less frequent, 2% among people with housing and 4% among people experiencing homelessness had a co-diagnosis of schizophrenia or other psychotic disorders during their hospital stay. Patterns for other mental health condition categories were more similar by housing status.

Figure 9 - Text Equivalent

Co-diagnosed mental health conditions	Any mental health condition	Substance-related and addictive disorders	Mood disorders	Anxiety disorders	Schizophrenia and other psychotic disorders	Selected disorders of personality and behavior	Other mental health conditions
	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)	% (N)
People with housing	52% (5,196)	25% (2,509)	21% (2,122)	9% (865)	2% (236)	7% (694)	12% (1,232)
People experiencing homelessness	61% (380)	51% (319)	11% (68)	3% (18)	4% (28)	8% (51)	9% (53)
Data source Discharge Abstract Database (DAD), fiscal year 2019/2020, Canadian Institute for Health Information (CIHI).
Note(s) Co-diagnoses of mental health conditions in this report do not reflect the overall prevalence of mental health conditions among those hospitalized for substance-related poisonings, but rather mental health conditions that were recorded/deemed relevant to the patient's stay in hospital. A patient can be diagnosed with more than one mental health condition during a hospital stay. Refer to the technical notes and Appendix D for information regarding ICD-10-CA codes included and methodology.

In Summary

Canada continues to experience an overdose crisis and data has shown that substance-related harms and deaths continue to be of great concern ^{Footnote 5}. Furthering our understanding of substance-related harms, including the demographics, context, and implications among specific populations, is essential to inform the implementation of evidence-based clinical and public health actions, harm reduction strategies, and social services.

This report found that approximately 6% of hospitalizations for substance-related poisonings (that were primarily from opioids, stimulants, cannabis, alcohol, and other depressants) across the country were among people experiencing homelessness. When compared to people with secure housing, hospitalizations for substance-related poisonings among people experiencing homelessness had a higher proportion of:

• Males
• Younger individuals
• Poisonings which were accidental in nature
• Poisonings due to opioids and stimulants, notably involving fentanyl and heroin
• Recorded mental health conditions, mainly for substance-related and addictive disorders
• Hospital stays ending by leaving against medical advice before being formally discharged by a health care professional

Additionally, compared to people with secure housing, people experiencing homelessness:

• Had lengthier stays in hospital and stayed four days longer on average

These findings are highlighted for healthcare professionals, researchers, and policy makers to better understand the intersection of homelessness, mental health, and substance-related harms and can be used to inform sectors that interact with people experiencing homelessness. In particular, these findings demonstrate how substance-related harms may differ among people with housing and people experiencing homelessness. Care in hospital settings may also be experienced differently among people experiencing homelessness as they had a higher proportion of substance-related poisoning hospitalizations which ended by leaving against medical advice. This may be due to a variety of factors, such as care not meeting the needs of this population or due to a lack of trust or stigma, and warrants further investigation to reduce barriers to care for people experiencing homelessness.

It is important to note that at the time of this analysis, homelessness status was only mandatory to record upon admission to hospital, and therefore the data presented in this report is likely an underestimate, as it may not reflect people who are experiencing homelessness upon discharge. Additional research is needed to better understand the causes and consequences of substance-related harms among people experiencing homelessness, but also among other populations that are disproportionally affected by substance use, such as LGBTQ2+ populations, racialized populations, and Indigenous Peoples ^{Footnote 22Footnote 23}. The Government of Canada will continue to improve data and analysis to inform strategies and interventions to reduce substance-related harms across Canada.

Acknowledgements

We would like to acknowledge the Canadian Institute for Health Information (CIHI) for collecting and providing the data used in this report.

Disclaimer

Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information (CIHI). However, the analyses, conclusions, opinions and statements expressed herein are those of the authors, and not necessarily those of the Canadian Institute for Health Information (CIHI).

Suggested Citation

Substance-related poisonings and homelessness in Canada: a descriptive analysis of hospitalization data. Ottawa: Public Health Agency of Canada; June 2021. https://www.canada.ca/en/health-canada/services/opioids/hospitalizations-substance-related-poisonings-homelessness.html

Technical notes

Methodology

Data from the Canadian Institute for Health Information's (CIHI) Discharge Abstract Database (DAD) for the fiscal year April 1, 2019 to March 31, 2020 were analyzed for this descriptive report. This analysis was limited to acute inpatient hospital discharges, which are nationally representative across Canada, except for Quebec. This analysis presents the number of acute inpatient hospital discharges for substance-related poisonings; it does not reflect the number of patients who were hospitalized in the analysis year. It is possible some patients may have been hospitalized more than once for a substance-related poisoning.

The International Classification of Diseases and Related Health Problems, Tenth Revision, Canada (ICD-10-CA) is used in the DAD to capture diagnoses from the patient's hospitalization. It is the national standard for reporting morbidity statistics. More information on the DAD and ICD-10-CA coding can be found on CIHI's website.

The methodology for identifying substance-related poisonings in this report was adapted from CIHI's Opioid-related Harms in Canada report and Hospital Stays for Harm Caused by Substance Use indicator. The methodology for identifying homelessness was followed as per the Canadian Coding Standards directives for version 2018 ICD-10-CA. The methodology for identifying co-diagnosed mental health conditions was adapted from CIHI's indicator for Repeat Hospital Stays for Mental Illness.

Identifying substance-related poisonings

A complete listing of all ICD-10-CA diagnosis codes used to identify substance-related poisonings presented in this report can be found in Appendix B. Of note, poisonings from the following substances of interest were included in the scope of this analysis: opioids, stimulants, cannabis, hallucinogens, alcohol, other depressants, and psychotropic drugs. Information on these substance-related poisonings are extracted from patient charts by trained coders, which may be based on toxicological analysis and/or patient self-report.

For substance-related poisonings, analyses were restricted to the following significant diagnosis types:

• Most responsible diagnosis ("M")
• Pre-admit comorbidity ("1")
• Post-admit comorbidity ("2")
• Service transfer diagnosis ("W", "X", and "Y")

These significant diagnosis types capture hospitalizations in which the substance-related harm poisoning was considered influential to the time spent and treatment received by the patient in hospital. Secondary diagnoses that do not meet the criteria for significance (diagnosis type 3) were excluded.

A diagnosis prefix of "Q", indicating unconfirmed diagnoses or query diagnoses recorded by the physician, were excluded from these analyses as the scope of the analysis was concerned with confirmed cases of poisonings only. Records indicating the patient was admitted to the facility as a cadaveric donor and stillbirths were also excluded.

Identifying intention of poisoning

When a patient experiences a poisoning, coders are directed to assign an external cause ICD-10-CA code to indicate the intention (i.e., reason) for the poisoning. Each substance-related poisoning of interest for this analysis was further examined by the documented intention for the poisoning, and classified in the following three groups in line with the Intentional Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Canada:

• Accidental: A substance-related poisoning hospitalization that is considered to be non-intentional in nature and is identified by an associated external cause ICD-10-CA code of X41, X42, or X45.


• Intentional: A substance-related poisoning hospitalization that occurred as a result of purposely self-inflicted harm and is identified by an associated external cause ICD-10-CA code of X61, X62, or X65.


• Undetermined: A substance-related poisoning hospitalization which has been documented to be of undetermined/unknown intention and is identified by an associated external cause ICD-10-CA code of Y11, Y12, or Y15.

In line with CIHI coding standards, a diagnosis type of "9" was used to identify intention using external cause codes. External cause codes are mandatory to assign with codes within the range S00-T98 Injury, poisoning, and certain other consequences of external causes. All poisonings captured in this report fall within this range, however a small number of records contained poisonings that were found to be missing associated external cause codes and therefore it was not possible to determine the intention for these. Hospital records where one or more poisonings were missing an associated external cause code to determine intention were excluded from the intention analysis, but were included in all other analyses. In total, 102 records were missing complete codes to determine the intention of the poisoning and excluded from the intention analyses.

External cause codes for intention with a diagnosis prefix of "Q", indicating unconfirmed diagnoses or query diagnoses, were included when identifying intention as the scope of this analysis was concerned with capturing suspected diagnoses for intention.

A list of all substance-related poisoning types of interest for this analysis and their associated intention codes can be found in Appendix C.

Identifying homelessness

The ICD-10-CA code of "Z59.0 - Homelessness" was used to identify homelessness status upon admission to hospital. Substance-related poisoning records with and without a corresponding ICD-10-CA code of "Z59.0 - Homelessness", were identified to create two groups for analyses:

1) Substance-related poisoning hospitalizations among people experiencing homelessness and;
2) Substance-related poisoning hospitalizations among people with housing.

When examining recorded homelessness during the patient's substance-related poisoning hospitalization, the following diagnosis types were included:

• Most responsible diagnosis ("M")
• Pre-admit comorbidity ("1")
• Post-admit comorbidity ("2")
• Secondary diagnosis ("3")
• Service transfer diagnosis ("W", "X", "Y")

Of note, diagnosis types of "3" were included as the scope of this analysis was concerned with capturing secondary diagnoses for homelessness as per CIHI coding directives. Unconfirmed diagnoses or query diagnoses, with a diagnosis prefix of "Q", were included when identifying homelessness as the scope of this analysis was also concerned with capturing suspected diagnoses for homelessness.

Identifying co-diagnosed mental health conditions

Specific examples of diagnoses included in each mental health condition category reported and associated ICD-10-CA codes can be found in Appendix D. When examining recorded co-diagnoses of mental health conditions (referred to as mental disorders in the ICD-10-CA) during the patient's substance-related poisoning hospital stay, the following diagnosis types were included:

• Most responsible diagnosis ("M")
• Pre-admit comorbidity ("1")
• Post-admit comorbidity ("2")
• Secondary diagnosis ("3")
• Service transfer diagnosis ("W", "X", "Y")

Of note, diagnosis types of "3" were included as the scope of this analysis was concerned with capturing secondary diagnoses for mental health conditions.

Unconfirmed diagnoses or query diagnoses (with a diagnosis prefix of "Q") were included when identifying mental health conditions as the scope of this analysis was also concerned with capturing suspected diagnoses for mental health conditions.

Recording diagnoses of mental health conditions is mandatory if the mental health condition(s):

• significantly affects the treatment received
• requires treatment beyond maintenance of the pre-existing condition, or
• increases the length of stay in hospital by at least 24 hours.

Therefore, the data presented in this report does not reflect the overall prevalence of mental health conditions among those hospitalized for substance-related poisonings, but rather mental health conditions that were deemed relevant to the patient's hospital stay.

Limitations

General notes

1. This analysis is descriptive in nature and while it will report differences in demographics, context of poisoning, and hospital characteristics, it does not evaluate why these differences occur.
2. It was not possible to determine which poisonings, if applicable, were a result of pharmaceutical substances, non-pharmaceutical substances, or a combination of both. It was also not possible to determine whether co-poisonings were a result of intended use of multiple substances at the same time or close in time, or if a substance unknowingly contained other substances.
3. The DAD captures patients discharged from hospital. Patients who were still hospitalized during the analysis year are not captured in this report.
4. The terminology used in this report aligns with the terms used in the International Classification of Diseases and Related Health Problems, Tenth Revision, Canada (ICD-10-CA).
5. This analysis does not include data from the province of Quebec.
6. Data released by provinces and territories may differ from the data provided in this report due to the availability of updated data, differences in the type of data reported, the use of alternate age groupings, differences in time periods presented and/or population estimates used for calculations, etc.

Data on sex

7. At the time of this analysis, no data was available in relation to gender identity. In the DAD, data on sex is typically recorded based on biological determination or legal documentation. Although an 'other' sex group was available, the counts were too low to include in this report. Therefore, the 'other' sex group was retained in the overall analyses, but excluded from all analyses by sex to avoid disclosure.

Data on homelessness

8. The ICD-10-CA code used to identify people experiencing homelessness (Z59.0) is mandatory to capture individuals who are experiencing homelessness upon admission to hospital. Patients who were newly experiencing homelessness upon discharge from hospital but not homeless on admission, may not have been captured. A new coding direction, effective on April 1, 2021, specifies it is mandatory to code Z59.0 for patients who are experiencing homelessness whenever documented (e.g., on admission, during admission) ^{Footnote 24}. As coverage improves, this analysis should be repeated.
9. Homelessness status can rely on self-reported information. It is possible some patients may have not disclosed they were experiencing homelessness, or were not able to do so due to disability or death. As a result, these instances of "hidden homelessness" may have been falsely classified to the "people with housing" group.
10. It's important to note that even though an individual may be housed, they could be experiencing housing insecurity. It was not possible to examine other factors, such as unstable housing, poor housing quality or overcrowding.

Data on intention

11. Poisonings are classified as accidental unless there is clear documentation of intentional self-harm or undetermined intention. This may reflect an overrepresentation of accidental poisonings.
12. Determining intention can rely on self-reported information which patients may not be willing to disclose due to various reasons, or unable to due to disability or death.

Data on mental health conditions

13. Co-diagnoses of mental health conditions in this report do not reflect the overall prevalence of mental health conditions among those hospitalized for substance-related poisonings, but rather mental health conditions that were recorded/deemed relevant to the patient's stay in hospital. Mental health conditions can go undiagnosed, or take time to diagnose, and these instances would not be captured in this data.

Data suppression

Values representing less than five hospitalizations are suppressed according to CIHI's privacy guidelines.