Hospital data sheds light on patients experiencing homelessness

The population experiencing homelessness has become increasingly diverse, with more women, families and youth experiencing homelessness.Reference1 To better understand the characteristics and needs of patients experiencing homelessness, health system leaders can use hospital data that identifies PEH via code Z59.0. Understanding the number and clinical characteristics of PEH at a hospital level can influence the development of services tailored to their needs.

As health care providers, we recognize the importance of accurate data for understanding an individual’s care journey across sectors and organizations, especially when addressing homelessness. For example, hospital and service utilization data has been essential to understanding emergency department patterns for those living without stable housing in our community, enabling us to improve care for marginalized people through different service delivery models. Through collaborations like our partnership with London Cares, we’ve leveraged data to help individuals in our community access supportive housing and comprehensive 24/7 health and social support services. — Brad Campbell, Corporate Hospital Administrative Executive, London Health Sciences Centre

Nearly 30,000 hospitalizations included code Z59.0 Homelessness in 2022–2023

It is likely that more PEH could have been identified. Factors affecting the recording of ICD-10-CA code Z59.0 include the following:

• Patients not identifying themselves as experiencing homelessness due to a concern about stigma or for other reasons
• Variation in the processes hospitals use to ensure that health care providers ask patients about their housing status.Reference2

Given these factors, there are opportunities to increase consistency in recording this code and to improve its utility.Reference2, Reference3

Patients experiencing homelessness stayed in hospital twice as long, with double the costs, compared with the national average 

In addition to longer lengths of stay, hospitalizations for PEH were more likely to include alternate level of care (ALC) days, compared with the Canadian average (11.9% versus 6.2%).Reference6 Note that both percentages exclude Quebec.

Substance use and mental health disorders were frequent reasons for hospitalization  

Analysis of the top 10 reasons for hospitalizations suggests that more than a quarter are related to substance use disorders (SUDs). SUDs were identified as the main reason for hospitalization in 18% of hospitalizations. Other top 10 diagnoses such as cellulitis (7%) and poisoning by narcotics (including opioids) and psychodysleptics (2%) are health consequences of specific types of substance use.Reference7 Patients may also have had secondary diagnoses of SUDs that were not considered in the analysis.

Most types of mental health and substance use hospitalizations are associated with longer-than-average hospital stays and repeat admissions.Reference4, Reference8, Reference9 Reference10

Admission via the emergency department was very common

Our analysis supports previously reported findings that patients experiencing homelessness often have complex health needs. Compared with other patients, they are more likely to rely on hospitals for health care.Reference11 Many PEH face barriers to accessing other types of health care services, including primary care.Reference11, Reference12 Getting appropriate care in the community may prevent emergency departments visits and hospital stays.

Most patients or people experiencing homelessness were male and were younger adults

We compared the demographic characteristics of PEH from hospital administrative data (the Hospital Morbidity Database) with the following non-hospital data sources on homelessness: the Homelessness Counts project, the Point-in-Time Count and the National Shelter Study.Reference13, Reference14, Reference15 For more information on these data sources, please refer to the About the data section.

The data sources generally showed a similar profile in that most patients or people experiencing homelessness were male and more than half were younger adults. Some differences in the age distribution were noted. For example, children and youth were more likely to be identified in the 3 non-hospital data sources than in hospitalization data.

Indigenous Peoples are disproportionately impacted by homelessness. While 5% of the population identified as Indigenous in the 2021 Census,Reference16 at least 31% of people experiencing homelessness identified as Indigenous in the non-hospital data sources.Reference14, Reference15, Reference17  Indigenous identity data was not available in CIHI’s hospitalization data. Starting in 2022–2023, hospitals have the option to collect and submit data on patients’ Indigenous identity and racialized group. Enhanced collection of these socio-demographic data elements will enable better understanding of specific groups such as PEH in the future.  

CIHI is committed to working in partnership with First Nations, Inuit and Métis Peoples to address their health data priorities and advance data sovereignty. To learn more, visit First Nations, Inuit and Métis Health.

Demographic characteristics of people experiencing homelessness: Comparing data sources

Notes 
* Depends on the data source. Hospitalization data is reported as “male” or “female” based on the variable “Recorded sex or gender.” Gender identity was collected for the Point-in-Time Count survey, the National Shelter Study and the Homelessness Counts project. 
† Refers to the category “gender diverse.”
‡ Refers to people who identified as a gender other than male or female.
— Not available. 
Percentages may not add to 100 due to rounding.

Sources 
Hospital Morbidity Database, 2022–2023, Canadian Institute for Health Information.
Homelessness Counts, 2021, Lawson Health Research Institute.Reference17 
Point-in-Time Count, Survey component, 2020 to 2022, Infrastructure Canada.Reference14
National Shelter Study, 2022, Infrastructure Canada.Reference15

Hospital Morbidity Database (CIHI) 

CIHI analysis used acute care records from the Hospital Morbidity Database, which includes data from all Canadian provinces and territories. Records for other types of inpatient care such as psychiatric or rehabilitation were not included. Homelessness was identified in the data using a standard code (ICD-10-CA Z59.0 Homelessness). The definition of Z59.0 covers a range of physical living situations that include overnight shelters and staying in temporary accommodations.Reference18

Some PEH received care in designated adult inpatient mental health beds in Ontario. These services were reported to the Ontario Mental Health Reporting System (OMHRS) and were not included in this study as OMHRS does not use Z59.0 as its primary identifier of PEH.

The methodology used in this analysis is described in the Notes to readers tab in the data tables included in the Featured material section.

Homelessness Counts (Lawson Health Research Institute)

Homelessness Counts is a federally funded project launched in 2021 to improve understanding of how many people in Canada are homeless and who they are.Reference13 The research team collected data from 28 communities across every province and territory in Canada. They interviewed 400 people experiencing homelessness and held focus groups with 200 homelessness-related service providers.Reference19

A key finding was that data gaps in rural areas contribute to underestimates of homelessness, particularly for Indigenous Peoples. Of the people surveyed in rural/remote areas, 83.6% identified as Indigenous compared with 28.6% in urban areas.Reference17 The project also identified additional priority subgroups within PEH populations in Canada, and the intersections of homelessness with dementia,Reference20 accidental opioid-related overdose death,Reference21 COVID-19 testing and infection,Reference22, Reference23 and COVID-19 vaccine coverage.Reference24 Other subgroups included youth, seniors, women, families, people who use substances, 2SLGBTQ+, students, and adults or young people with developmental or cognitive disabilities.Reference19 People experiencing neurodiversity ranged from youth with autism to seniors with dementias such as Alzheimer disease.

Point-in-Time Counts (Infrastructure Canada)

Nationally coordinated Point-in-Time Counts of homelessness in Canadian communities are supported by Infrastructure Canada. The third nationally coordinated Point-in-Time Count was conducted between 2020 and 2022.Reference14

The Point-in-Time Counts have 2 components:

• The enumeration (i.e., count) covers 3 types of locations: unsheltered, sheltered (e.g., emergency shelters; hotel or motel rooms provided as shelter spaces) and transitional housing.
• The survey component aims to reach people experiencing homelessness across the community, including people in health and correctional institutions, and those who are temporarily staying with others without access to a safe and permanent residence.Reference14

National Shelter Study (Infrastructure Canada)

The National Shelter Study is an ongoing analysis of trends in homelessness shelter use in Canada.Reference15 It uses data from the Homeless Individuals and Families Information System, and from provincial and municipal partners that have data-sharing agreements with the Government of Canada. In 2022, approximately 50% of emergency shelters in Canada were included in the analysis, representing about 60% of Canada’s emergency shelter beds.

National Shelter Study reports provide information on the number of shelter users, their demographic characteristics and trends in shelter use. In 2022, an estimated 105,655 people experienced homelessness at an emergency shelter. On an average night, there were approximately 16,248 people staying in shelters.Reference15

CIHI acknowledges Dr. Cheryl Forchuk of Parkwood Institute Research as a co-author on this story, and the staff of the Homelessness Policy Directorate at Infrastructure Canada for their valuable review.   

• Better quality hospital data for identifying patients experiencing homelessness 
• Codes, Classifications, and Terminologies
• First Nations, Inuit and Métis health 
• Health equity and population health
• Hospital Morbidity Database 
• A review of Canadian homelessness data, 2023
• About Reaching Home: Canada's Homelessness Strategy
   

Back to Homelessness and hospital use