Technical ReportPDF Available

Abstract and Figures

Given the success in increasing fire safety across Canada as a result of legislative and community-based initiatives and policies related to smoke detectors, there are clear benefits to expanding the scope of the Smoke Alarm Movement to incorporate carbon monoxide (CO) alarms. Previous research has demonstrated that accidental CO poisoning can be reduced through a combination of public education, emission controls, warning labels on products, and combination residential CO and smoke alarms (Hampson & Holm, 2017).
Content may be subject to copyright.
Carbon Monoxide Poisoning
Hospitalizations and Deaths in Canada
Irwin Cohen, Len Garis, Fahra Rajabali, Ian Pike
October 2017
The British Columbia Injury Research and Prevention Unit (BCIRPU) was established by the
Ministry of Health and the Minister’s Injury Prevention Advisory Committee in August 1997.
BCIRPU is housed within the Evidence to Innovation research theme at BC Children’s Hospital
(BCCH) and supported by the Provincial Health Services Authority (PHSA) and the University of
British Columbia (UBC). BCIRPU’s vision is, to be a leader in the production and transfer of injury
prevention knowledge and the integration of evidence-based injury prevention practices into the daily
lives of those at risk, those who care for them, and those with a mandate for public health and safety in
British Columbia.
Authors: Irwin Cohen, Len Garis, Fahra Rajabali, Ian Pike.
Reproduction, in its original form, is permitted for background use for private study, education
instruction and research, provided appropriate credit is given to the BC Injury Research and
Prevention Unit and the University of the Fraser Valley. Citation in editorial copy, for newsprint,
radio and television is permitted. The material may not be reproduced for commercial use or profit,
promotion, resale, or publication in whole or in part without written permission from the
University of the Fraser Valley.
For any questions regarding this report, contact:
BC Injury Research and Prevention Unit University of the Fraser Valley
F508-4480 Oak Street 33844 King Road
Vancouver, BC V6H 3V4 Abbotsford, BC V2S 7M8
Email: bcinjury1@cw.bc.ca Email: info@ufv.ca
Phone: (604) 875-3776 Phone: (604) 875 3776
Fax: (604) 875-3569
Web page: www.injuryresearch.bc.ca Web page: www.ufv.ca
Suggested Citation: Cohen, I, Garis, L, Rajabali F, Pike I. Carbon Monoxide Poisoning,
Hospitalizations and Deaths in Canada. A report by the BC Injury Research and Prevention Unit, for
the University of the Fraser Valley: Vancouver, BC. October, 2017.
1
Executive Summary
Carbon monoxide (CO) disorients its victims and is most dangerous when people are sleeping
and fail to wake up or realize they are at risk, as this gas has no colour, odour, or taste. The only
way to detect the presence of the deadly gas is to install a carbon monoxide alarm.
This research note provides an analysis of carbon monoxide poisoning hospitalizations and
deaths in Canada to demonstrate the need for specific legislation in each province and territory
to mandate the installation and maintenance of functioning CO alarms in every existing
residence.
According to Statistics Canada, between 2000 and 2013, in Canada, there were 4,990 deaths
associated to CO poisoning. This included 1,125 deaths where there were no other underlying
causes of death and 3,865 where there were other underlying causes of death.
In terms of the total number of deaths, 34.8% were for people between the ages of 25 and 44
years old, and an additional 40% were for those between 45 and 64 years old.
Quebec had the highest total number of carbon monoxide-related deaths (n = 1,445) followed
by Ontario (27.5 per cent), the Prairies (24.6 per cent), British Columbia and the Territories
(13.3 per cent), and the Maritimes (5.6 per cent).
In total, there were 3,027 hospitalizations related to CO poisoning in Canada between 2002 and
2016, and 14.3% of CO-related hospitalizations in Canada were for people 65 years old or older.
When missing data or unspecified data was removed from the analysis, 75% of CO-related
hospitalizations occurred as a result of CO poisoning originating in the home.
Ontario, Alberta, and British Columbia had the highest number of CO-related hospitalizations,
while the largest increases in per capita hospitalization between 2002 and 2016 were in
Saskatchewan (34.7 per cent), Manitoba (19.2 per cent), and British Columbia (7.6 per cent).
There are more than 300 CO-related deaths per year in Canada, and more than 200
hospitalizations per year in Canada.
Similar to the research on smoke alarms, CO alarms save lives and can reduce CO-related
deaths.
This research note recommends that all provincial governments either assist with regulating
the requirement for the retrospective mandatory installation of CO detectors in all residential
homes or legislate this requirement through building code amendments.
2
Introduction
Given the success in increasing fire safety across Canada as a result of legislative and community-
based initiatives and policies related to smoke detectors, there are clear benefits to expanding the
scope of the Smoke Alarm Movement to incorporate carbon monoxide (CO) alarms. Previous
research has demonstrated that accidental CO poisoning can be reduced through a combination of
public education, emission controls, warning labels on products, and combination residential CO
and smoke alarms (Hampson & Holm, 2017).
Recent amendments to the Ontario Fire Protection and Prevention Act, 1997, have mandated that
CO alarms be installed in all residences with a fuel-burning appliance or an attached garage.
Specifically, on October 15, 2014, the Ontario Government formally enacted The Hawkins-Gignac
Act (Bill 18) making CO alarms mandatory in all Ontario homes at risk of CO. This revision to the
Ontario Fire Code superseded any existing municipal by-laws, ensuring that CO alarms are installed
outside sleeping areas in residential dwellings with CO producing sources. Ontario’s CO alarm law
provides a consistent level of protection to all Ontarians, and should serve as a model for the rest of
Canada.
In addition to addressing the issue of CO in fire codes retrospectively to certify that all current
dwellings have functioning CO detectors, building codes are also an effective way of ensuring that
CO detectors are installed in all constructions moving forward. This objective was supported by the
National Fire Protection Association, the Fire Fighters Association of Ontario, members of the
medical community, and carbon monoxide survivor groups and their families (Garis, Clare, &
Hughan, 2015).
Currently, there is no requirement in British Columbia to address the risk of carbon monoxide
poisoning deaths and injuries in residential dwellings with mandatory CO detector regulations in
existing homes, which is preventable by having functioning CO detectors in all residential dwellings.
However, legislation similar to that of Ontario is extremely important for all communities.
According to the United States Center for Disease Control, 5,149 deaths resulted from unintentional
CO poisoning between 1999 and 2010, an average of 430 deaths per year (Centre for Disease
Control, 2014). Moreover, the average annual death rate from CO poisoning is three times higher
for males compared to females, at 0.22 per 100,000 males and 0.07 per 100,000 females (Garis,
Clare, & Hughan, 2015). As with fire-related fatalities, the death rates from CO poisoning were
highest among those aged 65 years and over (Garis, Clare, & Hughes, 2015). CO can be emitted if
fuel-burning devices are improperly installed or poorly maintained. Given this, vents and flues must
be free of debris and not cracked or clogged. Carbon monoxide can originate from gas, oil, or
propane furnaces, water heaters, clothes dryers, space heaters, gas ovens, and wood burning or gas
fireplaces. Prolonged exposure to carbon monoxide can lead to brain damage and death.
Carbon monoxide disorients its victims and is most dangerous when people are sleeping and fail to
wake up or realize they are at risk, as this gas has no colour, odour, or taste. The only way to detect
the presence of the deadly gas is to install a carbon monoxide alarm. While CO alarms do not reduce
the requirement to use fuel-burning appliances in a safe manner, installing and ensuring that CO
alarms are functioning as intended are, therefore, critical to reducing the risks that lead to
unintentional CO poisoning. This research note provides an analysis of carbon monoxide poisoning
3
hospitalizations and deaths in Canada to demonstrate the need for specific legislation in each
province and territory to mandate the installation and maintenance of functioning CO alarms in
every existing residence.
NATIONAL BUILDING CODE
A building code (also referred to as building control or building regulations) is a set of regulations
that specify the standards for constructed objects, such as buildings and non-building structures.
Buildings must conform to the code to obtain planning permission, usually from a local
government. The main purpose of building codes is to protect public health, safety, and general
welfare as they relate to the construction and occupancy of buildings and structures. The building
code becomes the law of a particular jurisdiction when formally enacted by the appropriate
governmental or private authority.
Building codes are generally intended to be applied by architects, engineers, interior designers,
constructors and regulators, but are also used for various purposes by safety inspectors,
environmental scientists, real estate developers, sub-contractors, manufacturers of building
products and materials, insurance companies, facility managers, tenants, and others. Importantly, in
Canada, the National Building Code is the model building code that forms the basis for all of the
provincial building codes. Some jurisdictions create their own code based on the National Building
Code, and other jurisdictions have adopted national standards often with supplementary laws or
regulations to the requirements in the National Building Code.
According to the National Building Code, the requirement for a CO alarm applies to every building
that contains a residential occupancy, a care occupancy with individual suites, or a care occupancy
containing sleeping rooms not within a suite, and that also contains a fuel-burning appliance, or a
storage garage. Here, CO devices are to be equipped with an integral alarm that satisfies the
audibility requirements of the policy, have no disconnect switch between the overcurrent device
and the CO alarm where the CO alarm is powered by the electrical system serving the suite, and
must be mechanically fixed at a height above the floor as recommended by the manufacturer.
Moreover, where a fuel-burning appliance is installed in a suite of residential occupancy or in a
suite of care occupancy, a CO alarm must be installed inside each bedroom, or outside each
bedroom, within five meters of each bedroom door, measured following corridors and doorways.
Where a fuel-burning appliance is installed in a service room that is not in a suite of residential
occupancy nor in a suite of care occupancy, a CO alarm must be installed either inside each
bedroom, or if outside, within five meters of each bedroom door, measured following corridors and
doorways, in every suite of residential occupancy or suite of care occupancy that shares a
wall/ceiling assembly with the service room, and in the service room. For each suite of residential
occupancy or suite of care occupancy that shares a wall or floor/ceiling assembly with a storage
garage, or that is adjacent to an attic or crawl space to which the storage garage is also adjacent, a
CO alarm must be installed inside each bedroom, or outside each bedroom, within five meters of
each bedroom door, measured following corridors and doorways. Where CO alarms are installed in
a house with a secondary suite including their common spaces, the CO alarms shall be wired so that
the activation of any one CO alarm causes all CO alarms within the house with a secondary suite,
including their common spaces, to sound. As will be demonstrated below, Ontario, Alberta, and
4
British Columbia are the leading provinces for the number of hospitalizations, while all three are in
the top four provinces for fatalities related to carbon monoxide poisoning. As such, the codes for
these three provinces will be reviewed.
BRITISH COLUMBIA
In British Columbia, on December 20, 2012, the British Columbia Building and Plumbing Code or “BC
Building Code 2012” came into force and provided an objective-based code that identified the
minimum standard within the Province of BC for buildings. It is a regulation of the Local
Government Act and is substantially based on the model National Building Code of Canada 2010 and
the model National Plumbing Code of Canada 2010. While its references to CO alarms are exactly the
same as those found in the National Building Code of Canada, it is important to note that Canadian
building codes typically no longer apply once a building is occupied, unless the building is
undergoing alteration or change of use, or being demolished. Still, the BC Building Code 2012
requires that an enclosed storage garage shall have a mechanical ventilation system designed to
limit the concentration of carbon monoxide to not more than 100 parts per million parts of air, and
where CO alarms are installed in a house with a secondary suite, including their common spaces,
the CO alarms shall be wired so that the activation of any one CO alarms causes all CO alarms within
the house with a secondary suite, including their common spaces, to sound.
ALBERTA
The Alberta Building Code (2014) sets out technical provisions for the design and construction of
new buildings. It also applies to the alteration, change of use, and demolition of existing buildings.
The Alberta Building Code (2014) complements the Alberta Fire Code (2014), and both are
indispensable for building officials, educators, and professionals in the construction industry. Safety
Services and the Safety Codes Council develop Alberta Building Code STANDATA jointly. Some are
issued under the authority of the Code or the Safety Codes Act as province-wide variances or
interpretations. Others are information bulletins that provide general advice on related matters.
The Alberta Building Code (2006) contains provisions requiring for CO alarms to be installed in all
new residential construction containing a fuel-burning appliance, a solid-fuel-burning appliance, or
a storage garage. Where a fuel-burning appliance is installed in a suite of residential occupancy, a
CO alarm must be installed inside each bedroom or outside each bedroom within five meters of
each bedroom door. Where a fuel-burning appliance is installed in a service room that is not in a
suite of residential occupancy, each suite that shares a wall or a floor/ceiling assembly with the
service room shall have a carbon monoxide alarm installed inside each bedroom or outside each
bedroom within five meters of each bedroom door, and in the service room. Where a solid-fuel-
burning appliance is installed in a suite of residential occupancy, a carbon monoxide alarm shall be
installed on or near the ceiling in the room containing the solid-fuel-burning appliance. Each suite
that shares a wall or floor/ceiling assembly with a storage garage or that is adjacent to an attic or
crawl space to which the storage garage is also adjacent, a CO alarm must be installed inside each
bedroom or outside each bedroom within five meters of each bedroom door.
According to the Alberta Building Code (2006), CO alarms must meet the requirements of the
Canadian Standards Association’s CAN/CSA Standard 6.19 “Residential Carbon Monoxide Alarming
Devices.” Labels found on certified CO alarms provide assurance that the alarm was tested and that
5
it conforms to established safety standards. As well as being a certified product, CO alarms must be
mechanically fixed to a surface at a height recommended by the manufacturer, and have no
disconnect switch between the overcurrent device and the CO alarm when the CO alarm is powered
by the dwelling unit’s electrical system. Both battery operated alarms and alarms that are
connected to the dwelling unit’s electrical system are acceptable. The average lifespan of CO alarms
varies. There are several models on the market with many different features, such as indicators to
let the user know when they need to be replaced or power supply backups.
ONTARIO
The 2012 Building Code Compendium of Ontario is administered by the Building and Development
Branch of the Ministry of Municipal Affairs and Housing. Prior to the enactment of the first Ontario
Building Code Act in 1974, individual municipalities were responsible for developing their own
building codes, resulting in a fragmented and potentially confusing regulatory environment. The
introduction of a provincial Building Code Act and a provincial Building Code addressed this
problem by providing for uniform construction standards across Ontario.
With respect to CO, where a fuel-burning appliance is installed in a suite of residential occupancy, a
CO alarm shall be installed adjacent to each sleeping area in the suite. Where a fuel-burning
appliance is installed in a service room that is not in a suite of residential occupancy, a CO alarm
shall be installed adjacent to each sleeping area in every suite of residential occupancy that is
adjacent to the service room, and in the service room. Where a storage garage is located in a
building containing a residential occupancy, a CO alarm shall be installed adjacent to each sleeping
area in every suite of residential occupancy that is adjacent to the storage garage.
Moreover, CO alarms must be permanently connected to an electrical circuit and shall have no
disconnect switch between the overcurrent device and the CO alarm, be wired so that its activation
will activate all CO alarms within the suite, where located within a suite of residential occupancy, be
equipped with an alarm that is audible within bedrooms when the intervening doors are closed,
where located in a suite of residential occupancy, and conform to CAN/CSA-6.19, “Residential
Carbon Monoxide Alarming Devices”, or UL 2034, “Single and Multiple Station Carbon Monoxide
Alarms”. Where the building is not supplied with electrical power, CO alarms are permitted to be
battery operated.
Purpose of Research Note
The purpose of this research note is to draw attention to the number of carbon monoxide
poisonings and deaths in Canada, which is, to some degree, preventable through the
implementation of legislation requiring all new constructions to have installed and functioning CO
alarms as the current built environment does not have this requirement. In effect, this research
note is a call for retrospective regulation to require the built environment to install CO alarms in all
existing residences.
6
Current Study
In addition to building code modifications, it is important for governments to create public
campaigns focused on educating people about the dangers of carbon monoxide exposure, the
benefits of having properly installed and functioning CO alarms, and what to do if their CO alarm is
activated.
In light of national and provincial building codes related to the prevention of CO poisoning in
Canada, this report provides statistical analyses on hospitalizations and deaths in which CO
poisoning was either the main diagnosis or played a part in the final diagnosis. This report uses
data provided by Statistics Canada’s Canadian Vital Statistics Death Database, the Canadian
Institute for Health Information’s Discharge Abstract Database, and Statistics Canada’s Table 051-
0001 - Estimates of Population for the analyses. The data on hospitalizations comprises the years
2002/2003 to 2015/2016, while the data on deaths is from 2000 to 2013.
DATA LIMITATIONS
For the mortality data, data by province was not available. Death counts are provided by geographic
region. Multiple cause data was not available for Saskatchewan in 2000-2002 and the Yukon in
2000 and part of 2002. Data for the Territories was included with British Columbia. Data for deaths
due to CO poisoning occurring in homes was not available.
For the hospitalization data, data for Quebec was not available. Due to small counts, hospitalization
data for Yukon, Nunuvat and Northwest Territories were excluded from the figures.
Data Analysis
CARBON MONOXIDE RELATED DEATHS
To understand the number of deaths in Canada related to CO poisoning, Statistics Canada’s
Canadian Vital Statistics Death Database, Multiple Cause Data was used. According to Statistics
Canada, the data represents deaths of Canadian residents and non-residents that occurred in
Canada that mentioned CO anywhere on the medical certificate of death, regardless of any
underlying causes of death. This includes cases where the underlying cause of death was accidental
or intentional poisoning, or an assault by and exposure to CO. Moreover, Statistics Canada rounds
the number of deaths to five if the total number is under five to meet their internal confidentiality
requirements. Of note, Statistics Canada states that missing information on the multiple causes of
death might contribute to an underestimation of CO-related deaths.
According to Statistics Canada, between 2000 and 2013, in Canada, there were 4,990 deaths
associated to CO poisoning. This included 1,125 deaths where there were no other underlying
causes of death and 3,865 where there were other underlying causes of death. In effect, over those
14 years, there was an average of 80 deaths per year exclusively attributed to CO poisoning, and an
additional 276 deaths per year in which CO poisoning, in addition to other underlying causes of
death, played a role in a person’s death. Statistics Canada also provided age ranges for CO-related
deaths. As demonstrated in Figure 1, in terms of the total number of deaths, 34.8% were for people
between the ages of 25 and 44 years old, and an additional 40% were for those between 45 and 64
7
years old. Still, 15% were for people 65 years old and older, and 10.2% were for those 24 years old
and younger. When considering just those instances where there were no other underlying causes
of death, 7.6% were for youth 14 years old and younger, and an additional 8.4% were for victims
between the ages of 15 and 24 years old. Similarly, 9.8% of deaths occurred among people between
the ages of 65 and 74 years old, 8.9% of deaths were among those between 75 and 84 years old, and
4% of all carbon monoxide deaths in Canada were among people 85 years old or older (see Figure
1).
FIGURE 1: AGE RANGES FOR CARBON MONOXIDE-RELATED DEATHS IN CANADA BETWEEN 2000 2013 (N =
4,990)
Statistics Canada grouped Canada’s provinces and Territories into five large geographic areas when
considering deaths related to CO poisoning. These groupings are the Maritime Provinces, Quebec,
Ontario, the Prairies, and British Columbia and the Territories. Figure 2 provides the data on deaths
related to CO poisoning based on these geographic groupings between 2000 and 2013. Quebec had
the highest total number of carbon monoxide-related deaths (n = 1,445) that accounted for 29% of
all carbon monoxide-related deaths in Canada. This was followed by Ontario (27.5 per cent) and the
Prairies (24.6 per cent). British Columbia and the Territories accounted for 13.3% of all carbon
monoxide-related deaths, while the Maritimes accounted for 5.6%. Of note, when considering those
deaths exclusively attributed to CO poisoning, 38.2% occurred in Quebec, 32% in Ontario, 13.3% in
British Columbia and the Territories, 11.1% in the Prairies, and 5.3% in the Maritimes.
0
85
95
320
370
110
100
45
0
15
315
1415
1625
285
160
50
0
100
410
1735
1995
395
260
95
Under 1 year
1 to 14 years
15 to 24 years
25 to 44 years
45 to 64 years
65 to 74 years
75 to 84 years
85 years and over
Total Other Underlying Causes of Death Carbon Monoxide Poisoning
8
FIGURE 2: CARBON MONOXIDE-RELATED DEATHS IN CANADA BY GEOGRAPHIC REGION BETWEEN 2000 2013
(N = 4,990)
The geographic distribution of carbon monoxide-related deaths in Canada between 2000 and 2013
is presented in Figure 3.
FIGURE 3: DEATHS FROM CARBON MONOXIDE POISONING IN CANADA (2000 TO 2013)
60
150 125
360 430
220
515
1105
1010 1015
280
665
1230
1370 1445
Maritimes BC and Territories Prairies Ontario Quebec
Carbon Monoxide Poisoning Other Underlying Causes of Death Total
9
CANADIAN DATA ON HOSPITALIZATIONS RELATED WHOLLY OR PARTIALLY TO CARBON
MONOXIDE POISONING
There was very little variation year over year in Canada for hospitalizations related to CO poisoning
between 2002/2003 and 2015/2016 in the Discharge Abstract Database from the Canadian
Institute for Health Information. According to the Canadian Institute for Health Information, the
hospitalization counts presented in this report represent acute hospitalizations in Canada with
mention of carbon monoxide anywhere in the discharge diagnosis, regardless of the most
responsible diagnosis for the hospitalizations. The category “CO as the Main Diagnosis” refers to
acute hospitalizations where CO poisoning is the most responsible diagnosis, which describes the
most significant condition of the patient during hospitalization.
In total, there were 3,027 hospitalizations over the 14 years or, on average, 216 per year. Over that
time period, 54.8% of hospitalizations had CO poisoning as the main diagnosis, while the remaining
45.2% had CO as part of the diagnosis. In effect, on average, of the 216 hospitalizations related to
CO poisoning, 118.5 cases resulted in CO poisoning being the main diagnosis. Of note, there were
271 hospitalizations in 2002/2003 and 234 hospitalizations in 2015/2016, which accounts for a
13.7% decrease over the 14 years. The year by year change for all CO-related hospitalizations in
Canada, those in which CO poisoning was the main diagnosis and those where CO poisoning was
part of a diagnosis is represented in Figure 4.
FIGURE 4: TOTAL NUMBER OF CO-RELATED HOSPITALIZATIONS IN CANADA BETWEEN 2002/03 AND 2015/16
(N = 3,027)
In terms of the gender and age of those hospitalized, of the 3,027 cases over the 14 years, 74.2% of
patients were male and those 49 years old and younger made up 48.7% of the CO-related
hospitalizations. Figure 5 presents the distribution of age ranges for all CO-related hospitalizations
over the 14-year time period. Of note, there was very little difference by age group between
0
100
200
300
400
500
600
CO Main Diagnosis CO as Part of Any Diagnosis Total
10
hospitalizations where CO was the main diagnosis and CO was a part of any diagnosis. Moreover,
only 14.3% of CO-related hospitalizations in Canada over the 14 years were for people 65 years old
or older.
FIGURE 5: AGE DISTRIBUTION OF CO-RELATED HOSPITALIZATIONS IN CANADA BETWEEN 2002/03 AND 2015/16
(N = 3,027)
Demonstrating the benefits of mandating CO alarms in all residences, 63% of all co-related
hospitalizations were the result of a poisoning that occurred in the home. Moreover, when missing
data or unspecified data was removed from the analysis, 75% of CO-related hospitalizations
occurred as a result of CO poisoning originating in the home. As demonstrated in Figure 6, there
was little variation year over year in the number of hospitalizations for CO poisoning that
originated in the home. In fact, the number of hospitalizations ranged from a high of 168 in
2004/2005 to a low of 114 in 2007/2008; a difference of 54 hospitalizations. And, between
2002/2003 and 2015/2016, the number of hospitalizations decreased by just 5.8% (see Figure 6).
In other words, residences were consistently the primary location that people suffered CO
poisonings that resulted in a hospitalization.
0 50 100 150 200 250 300 350 400
<1
1-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90+
Total CO as Part of Any Diagnosis CO as Main Diagnosis
11
FIGURE 6: NUMBER OF CARBON MONOXIDE RELATED HOSPITALIZATIONS WHERE THE POISONING OCCURRED IN
A HOME BY YEAR (N = 1,908)
Figure 7 presents the age distribution in Canada for CO-related hospitalizations where the
poisoning occurred in a home between 2002/2003 and 2015/2016. As demonstrated by Figure 7,
those between the ages of 30 years old and 54 years old made up 49.5% of CO-related hospital
admissions where the poisoning occurred in a home. Of note, only 17.2% of hospitalizations were
for people 65 years old or older.
FIGURE 7: AGE DISTRIBUTION OF CARBON MONOXIDE RELATED HOSPITALIZATIONS WHERE THE POISONING
OCCURRED IN A HOME BETWEEN 2002/2003 AND 2015/2016 (N = 1,908)
155
146
168
140
115
114
123
140
123
138
122
153
125
146
2002/03
2003/04
2004/05
2005/06
2006/07
2007/08
2008/09
2009/10
2010/11
2011/12
2012/13
2013/14
2014/15
2015/16
12
30
23
33
50
102
101
133
158
229
206
218
178
106
80
58
73
63
43
12
<1
1-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90+
12
PROVINCIAL DATA ON HOSPITALIZATIONS RELATED WHOLLY OR PARTIALLY TO CARBON
MONOXIDE POISONING
One clear way of understanding the frequency and rate of CO poisoning is by examining
hospitalizations in which CO poisoning was either the main diagnosis or was a contributing factor
in the final diagnosis. As demonstrated by Figure 8, 77.6% of all hospitalizations related to CO
poisoning in Canada between 2002 and 2016 were in Ontario (32.7 per cent), British Columbia
(23.4 per cent), and Alberta (21.5 per cent). In considering these three provinces, a majority of
hospitalizations for both Ontario (58.2 per cent) and British Columbia (58.8 per cent) were cases in
which CO poisoning was the main diagnosis. However, in Alberta, in only a minority (44.5 per cent)
of hospitalizations was CO poisoning the main diagnosis. It should be noted that CO poisoning as
the main diagnosis was also in the majority of cases for Newfoundland (66.3 per cent; n = 89),
Prince Edward Island (69.6 per cent; n = 23), Quebec (64.7 per cent; n = 17), and Saskatchewan
(65.8 per cent; n = 202). Regardless of whether the final diagnosis was exclusively CO poisoning or
CO poisoning was part of a diagnosis, the large number of hospitalizations suggests that additional
public education about the harms of CO and the need for functioning residential CO alarms is
merited.
FIGURE 8: TOTAL NUMBER OF CARBON MONOXIDE HOSPITALIZATIONS BY PROVINCE (2002 2016)
7
43
30
56
71
69
361
292
415
16
27
59
52
54
133
289
416
577
23
70
89
108
125
202
650
708
992
PE
NS
NL
MB
NB
SK
AB
BC
ON
Total CO as Main Diagnosis Co as Part of any Diagnosis
13
Another way to consider the hospitalization data is to control for population size. The data
presented in Figure 9 is the rate of hospitalization related wholly or in part to CO poisoning per
100,000 in each province. As demonstrated in Figure 6, while Alberta had the highest per capita
rate of hospitalizations (1.27 per 100,000 people), this was closely followed by Newfoundland
(1.22) and New Brunswick (1.19). Of note, Ontario’s per capita rate was only 0.55, suggesting that
their large contribution to the overall number of hospitalizations was due to their larger population
size, rather than a disproportionate number of CO-related hospitalizations.
FIGURE 9: TOTAL RATE OF CARBON MONOXIDE HOSPITALIZATIONS BY PROVINCE (2002 2016)
In terms of change over time, for those provinces in which data was available, on the raw number of
hospitalizations related wholly or in part to CO poisoning, there was no clear pattern across
provinces as some provinces saw increases between 2002/2003 and 2015/2016, while others saw
decreases (see Figure 10). However, for the three provinces with the largest number of
hospitalizations, Ontario had a decrease in hospitalizations over the 14 years of 15.6%, while
Alberta had a 40% decrease over the same time period. Conversely, British Columbia experienced a
23.7% increase in the number of CO-related hospitalizations over the same time period. It is
interesting to note that Ontario had a 31% increase in CO-related hospitalizations from 2014 to
2016, which might be the result of the legislation change mandating the installation and
maintenance of CO alarms in residences.
0.06
0.53
0.55
0.63
1.15
1.16
1.19
1.22
1.27
SK
NS
ON
MB
BC
PE
NB
NL
AB
14
FIGURE 10: NUMBER OF CARBON MONOXIDE HOSPITALIZATIONS BY PROVINCE (2002 2016)
When controlling for population sizes, again, some provinces saw increases in their per capita rate
of CO-related hospitalizations, while other provinces saw decreases (see Figure 11). In terms of
increases, the largest increase was found in Saskatchewan (34.7 per cent) between 2002/2003 and
2015/2016. This was followed by Manitoba (19.2 per cent), and British Columbia (7.6 per cent).
Conversely, the province with the largest decrease in CO-related hospitalizations over the same
time period was Newfoundland (70.5 per cent) followed by New Brunswick (63.4 per cent), Alberta
(55.1 per cent), Nova Scotia (44.0 per cent), and Ontario (25.7 per cent). Again, with respect to
Ontario, between 2014 and 2016, there was a 30.9% increase in the number of CO-related
hospitalizations.
0
50
100
150
200
250
300
AL BC MB NB NL NS ON SK
15
FIGURE 11: RATE OF CARBON MONOXIDE HOSPITALIZATIONS BY PROVINCE (2002 2016)
When considering the raw number of CO-related hospitalizations where the poisoning occurred in a
home by province and territory in Canada over the 14-year time period, as expected, the largest
number occurred in Ontario, followed by British Columbia, and Alberta (see Figure 12). In fact,
these three provinces accounted for 79.7% of all the hospitalizations. The importance of CO
detectors in the home is evidenced by the finding that, when removing missing or unspecified data
from the analysis, for Ontario, 76.6% of all CO-related hospitalizations were the result of CO
poisoning that occurred in the home. Similarly, the results for British Columbia was 73.7%, and
75% of CO-related hospitalizations in Alberta were the result of a CO poisoning that occurred in the
home.
0.00
2.00
4.00
6.00
8.00
10.00
12.00
AL BC MB NB NL NS ON SK
16
FIGURE 12: TOTAL NUMBER OF CARBON MONOXIDE RELATED HOSPITALIZATIONS IN CANADA BY PROVINCE
WHERE THE POISONING OCCURRED IN A HOME (N = 1,908)
Figure 13 provides a visual representation of the distribution of hospitalizations from carbon
monoxide poisonings in Canada between 2002/2003 and 2015/2016.
FIGURE 13: HOSPITALIZATIONS FROM CARBON MONOXIDE POISONING IN CANADA (2002/2003 TO 2015/2016)
Yukon, Nunavut, NWT
PE
NL
NS
MB
NB
SK
AB
BC
ON
17
Conclusion and Recommendation
This research note has provided evidence that, on average, there are more than 300 CO-related
deaths per year in Canada, and more than 200 hospitalizations per year in Canada. Moreover, the
vast majority of carbon monoxide poisonings occur in the home. Similar to the research on smoke
alarms, CO alarms save lives and can reduce CO-related deaths. However, unlike the research on
fire-related fatalities, which demonstrated a disproportionate risk for the elderly, three-quarters of
the CO-related deaths in Canada were among those between the ages of 25 years old and 64 years
old. However, if one only considers those deaths where there were no other underlying causes of
death other than CO poisoning, the proportion decreases to 61.3%.
The province of Ontario has already mandated the installation and maintenance of CO alarms in all
residential homes, retrospectively. While there has not been enough time to evaluate the full effect
of this policy to date, in the short time since the implementation of the policy, hospitalizations for
carbon monoxide poisoning in Ontario has increased slightly. What is needed is to determine
whether this increase in hospitalizations has a corresponding decrease in fatalities. Moreover, the
limited research in the United States (Hampson & Holm, 2017) has demonstrated that a
combination of legislation and public education can increase the number of CO alarms in homes,
which can decrease CO-related deaths. This is important because, in several provinces, such as
British Columbia and Saskatchewan, the per capita number of hospitalizations for CO poisoning has
increased in recent years.
While this research note could not find any research detailing the extent to which CO alarms are
installed and functioning in homes across Canada, similar to fire alarms in the past, it is likely that
there is a lot of variation. However, like smoke alarms, it is possible for every community to make
progress on a commitment to ensure that every home has a functioning CO alarm. The first step
towards this goal is to mandate this requirement in retrospective legislation throughout Canada.
Again, given the success associated with smoke alarms, there is little excuse for not adequately
addressing this issue. As such, the recommendation of this research note is that all provincial
governments either assist with regulating the requirement for the retrospective mandatory
installation of CO detectors in all residential homes or legislate this requirement through
building code amendments.
References
[1] Canadian Institute for Health Information, Discharge Abstract Database.
[2] Center for Disease Control. QuickStats: Average Annual Number of Deaths and Death Rates from
Unintentional, NonFire-Related Carbon Monoxide Poisoning,*† by Sex and Age Group United
States, 19992010. Morbidity and Mortality Weekly Report (MMWR) 2014 [cited 2015 20
August]; Available from:
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6303a6.htm
[3] Garis, L., Clare, J., & Hughan, S. (2015). Smoke Alarms Work, But Not Forever: Revisited.
Successes and Ongoing Challenges from BC’s Working Smoke Alarm Campaign. Centre for Public
Safety and Criminal Justice Research, University of the Fraser Valley.
18
[4] Hampson, N. B. & Holm, J. R. (2017). Compliance with Washington State’s Requirement for
Residential Carbon Monoxide Alarms. Preventive Medicine Reports. Vol. 5, pp. 232 235.
[5] Statistics Canada. Table 051-0001. Estimates of Population, by age group and sex for July 1,
Canada, Provinces, and Territories, annual (accessed June 23, 2017).
Acknowledgements
The authors would like to thank Kidde Canada for its ongoing support in smoke and CO detection
technology.
Author Biographical Information
Dr. Irwin M. Cohen is an Associate Professor in the School of Criminology and Criminal Justice at the
University of the Fraser Valley (UFV) and the Director of the Centre for Public Safety and Criminal Justice
Research at UFV. Contact him at Irwin.cohen@ufv.ca.
Len Garis is the Fire Chief for the City of Surrey, British Columbia, an Adjunct Professor in the
School of Criminology and Criminal Justice & Associate to the Centre for Social Research at the
University of the Fraser Valley (UFV), a member of the Affiliated Research Faculty at John Jay
College of Criminal Justice in New York, and a faculty member of the Institute of Canadian Urban
Research Studies at Simon Fraser University. Contact him at Len.Garis@ufv.ca
Fahra Rajabali holds an MSc in Health Information Science and has been a Researcher with the BC
Injury Research and Prevention Unit since 2000. Contact her at frajabali@bcchr.ca
Dr. Ian Pike is Professor of Pediatrics at UBC; Investigator and Co-Lead of the Evidence to
Innovation Research Theme at the Research Institute at BC Children’s Hospital; Director of the BC
Injury Research and Prevention Unit, and Co-executive Director, The Community Against
Preventable Injuries. Dr. Pike has given over 30 continuing education sessions to physicians,
nurses, public health, and safety professionals, and has over 50 peer-reviewed journal articles, and
numerous invited plenary and professional presentations. Contact him at ipike@bcchr.ca
Article
Full-text available
Fire-related mortality and morbidity has previously been demonstrated to be significantly higher among Indigenous people in Canada. However, several gaps in literature remain. This report examines age-standardized mortality and hospitalization rates relating to fire, burns and carbon monoxide poisoning among First Nations people, Métis and Inuit and compares it to rates among non-Indigenous people, living in private households.  The fire-related mortality rate was 1.6, 0.6 and 5.3 deaths per 100,000 person-years at risk among First Nations people, Métis and Inuit, respectively. In comparison, it was 0.3 among non-Indigenous people.  The hospitalization rate from injuries associated with fires was 7.5, 2.8 and 8.8 hospitalizations per 100,000 person-years at risk. The rate among non-Indigenous people was 1.7.  The mortality rate for burns was 1.0 death per 100,000 person-years at risk among First Nations people and 0.2 among non-Indigenous people.  The hospitalization rate for burns was 13.9, 5.0, 13.5 hospitalizations per 100,000 person-years at risk. The rate among non-Indigenous people was 4.3.  CO mortality rate was 0.5 deaths per 100,000 person-years at risk among First Nations people, 0.7 among Métis and 0.6 among non-Indigenous people.  The study indicates that disparities continue to exist among First Nations people, Métis and Inuit and non-Indigenous people. Contributing factors, which are not examined in the study, such as inadequate housing, lack of smoke detectors, underfunding for fire services on Indigenous communities, poverty, lack of legislation mandating adherence to building and fire codes on reserve need to be considered in interpreting the disparities in mortality and morbidity.
Article
Full-text available
Carbon monoxide (CO) poisoning is responsible for significant morbidity and mortality in the US. In response, a majority of states have passed legislation in recent years requiring the installation of residential CO alarms. There is, however, no published information evaluating compliance with such laws. Employees of a Seattle medical center were surveyed in 2008 regarding home use of CO and smoke alarms. Washington State enacted legislation requiring residential CO alarms by all residences by January 1, 2013. The survey was repeated in mid-2016 to evaluate compliance. In 2016, a total of 354 employees completed the survey and their responses were compared to an equal number of 2008 survey respondents matched by home ownership and ZIP code. Residential CO alarm use rose from 37% to 78% (p < 0.0001). Among homeowners, 78% had alarms while 80% of renters had them. Homeowners with the highest compliance (96%) had purchased their homes since January 1, 2013 while those with the lowest compliance (73%) had purchased them earlier. A majority (79%) of renters without alarms reported the reason was that their landlord did not provide one, a violation of the law. Only one-half to two-thirds of all equipped homes had the required number of either CO or smoke alarms. Use of residential CO alarms increased significantly in this study population three years after law required them. Areas for further improvement include education of landlords, tenants, and longtime homeowners about the law, as well as public education regarding the number of CO and smoke alarms needed.
QuickStats: Average Annual Number of Deaths and Death Rates from Unintentional, Non-Fire-Related Carbon Monoxide Poisoning,* † by Sex and Age Group United States
  • Center For Disease
  • Control
Center for Disease Control. QuickStats: Average Annual Number of Deaths and Death Rates from Unintentional, Non-Fire-Related Carbon Monoxide Poisoning,* † by Sex and Age Group United States, 1999-2010. Morbidity and Mortality Weekly Report (MMWR) 2014 [cited 2015 20 August]; Available from: http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6303a6.htm
Smoke Alarms Work, But Not Forever: Revisited. Successes and Ongoing Challenges from BC's Working Smoke Alarm Campaign
  • L Garis
  • J Clare
  • S Hughan
Garis, L., Clare, J., & Hughan, S. (2015). Smoke Alarms Work, But Not Forever: Revisited. Successes and Ongoing Challenges from BC's Working Smoke Alarm Campaign. Centre for Public Safety and Criminal Justice Research, University of the Fraser Valley.
Table 051-0001. Estimates of Population, by age group and sex for July 1, Canada, Provinces, and Territories, annual
Statistics Canada. Table 051-0001. Estimates of Population, by age group and sex for July 1, Canada, Provinces, and Territories, annual (accessed June 23, 2017).