Health at a Glance
Deaths involving sepsis in Canada

by Tanya Navaneelan, Sarah Alam, Paul A. Peters, and Owen Phillips

Release date: January 21, 2016

Start of text box 1

Highlights

  • In 2011, one in 18 deaths in Canada involved sepsis, a serious medical condition caused by an overwhelming immune response to an infection.
  • Deaths involving sepsis increased significantly between 2000 and 2007 and then remained stable between 2007 and 2011.
  • Deaths involving sepsis were higher among males than females, although between 2007 and 2011 the difference between the sexes narrowed.
  • Sepsis contributed to more than half of all deaths from infectious diseases in Canada during the 2009–2011 time period.

End of text box 1

Sepsis is a serious medical condition caused by an overwhelming immune response to an infection. The immune system sets off a series of reactions in the body including widespread inflammation, swelling and blood clotting.Note 1,Note 2 This results in impaired blood flow, which damages the body's organs by depriving them of nutrients and oxygen. If not treated quickly, sepsis can lead to multiple organ failure and death. It is estimated that 30% to 50% of people who develop sepsis die from it.Note 3,Note 4 Sepsis also has an economic burden by extending a patient's stay in the hospitalNote 4 and significantly increasing the cost of treatment.Note 5

Sepsis was the 12th leading cause of death in Canada in 2011, up from 15th place in 2000.Note 6,Note 7 Previous research shows that elderly people, males, and non-Caucasians have an increased risk of developing sepsis. People who develop sepsis often have other, pre-existing, conditions. The most common are human immunodeficiency virus (HIV), cancer, diabetes, cirrhosis, alcohol dependence, and bedsores.Note 8,Note 9

This article provides the first Canadian estimates of sepsis-associated deaths; that is deaths where sepsis was either an underlying or contributing cause of death. Previous national estimates of sepsis deaths using administrative data sources have only included deaths where sepsis was the underlying cause of death.

Start of text box 2

Underlying and contributing causes of death

The Vital Statistics Program at Statistics Canada collects data on deaths through its Canadian Vital Statistics – Deaths Database. Mortality rates are most often estimated using the underlying cause of death as reported on the death certificate. The underlying cause of death is defined as the disease or injury that initiated the series of events that led to death.Note 10

However, many deaths have multiple causes. That is, they are the result of not only the underlying cause but also other conditions. These other conditions are contributing causes of death – diseases or injuries which played a part in the death but were not reported as the underlying cause of death. For example, HIV may lead to a suppressed immune system which can lead to infection, ultimately causing sepsis. In this case, HIV may be reported as the underlying cause of death on the death certificate, while sepsis, the specific infection and the suppressed immune system may be reported as contributing causes.

Between 2000 and 2010, only 18% of all deaths in Canada had just an underlying cause reported on the death certificate. On the other hand, 25% of deaths had one reported contributing cause in addition to the underlying cause, 23% had two contributing causes, 16% had three contributing causes, 9% had four contributing causes and 9% had five or more contributing causes.Note 11

Only including the underlying causes of death does not allow for a full picture of all factors playing a role in the death. This is particularly true of sepsis for two reasons. First, sepsis is known to be a greater risk for the elderly and for other people with multiple chronic conditions that can make them more likely to get an infection. For these people, sepsis may be a condition that contributed to their death but it may not be listed as the underlying cause of death on the death certificate. Second, many cases of sepsis result from infections acquired in healthcare facilities. In these cases, the original reason for hospitalization, rather than sepsis, is often reported as the underlying cause of death on the death certificate.Note 8

Until recently, national multiple-causes-of-death data have only been available for limited use. However, the Vital Statistics – Deaths Database now includes multiple-causes-of-death data for all years starting in 2000.Note 12,Note 13

End of text box 2

Data presented in this article come from the Canadian Vital Statistics – Deaths Database (CVS:D) and are analysed over time and by sex, age, and underlying cause of death. Data from the years 2000 to 2011 are used when discussing changes over time. Three years of data (2009, 2010 and 2011) are combined when discussing differences by age or underlying cause of death to reduce the random variation that can occur with small numbers.

One in 18 deaths in Canada involve sepsis

In 2011, there were 2,515 deaths where sepsis was reported as the underlying cause.Note 14 In addition, there were 10,985 deaths where sepsis was reported as a contributing causeNote 15. This means that about one in 18 Canadian deaths involved sepsis that year (an age-standardized rate of 27.2 deaths per 100,000 people). This is similar to the level of sepsis-associated deaths found in England and the United States.Note 8,Note 16

Sepsis as a contributing cause of death increased between 2000 and 2007

In the early 2000s, the rate of sepsis-associated death (both underlying and contributing) increased significantly (Chart 1). Between 2000 and 2004, the sepsis-associated death rate increased by 4.1% annually.Note 17,Note 18 The rise was even greater between 2004 and 2007 when there was an increase of 10.6% annually. However, there was no significant change between 2007 and 2011.

Chart 1 - Sepsis-associated mortality, by type of cause of death

Description for Chart 1

The title of the graph is "Chart 1 Sepsis-associated mortality, by type of cause of death, age-standardized rates, Canada, 2000 to 2011."
This is a line chart.
There are in total 12 categories in the horizontal axis. The vertical axis starts at 0 and ends at 30 with ticks every 5 points.
There are 3 series in this graph.
The vertical axis is "deaths per 100,000."
The units of the horizontal axis are years from 2000 to 2011.
The title of series 1 is "Underlying and contributing."
The minimum value is 17.8 occurring in 2000.
The maximum value is 27.8 occurring in 2007.
The title of series 2 is "Contributing."
The minimum value is 14.0 occurring in 2000.
The maximum value is 23.2 occurring in 2007.
The title of series 3 is "Underlying."
The minimum value is 3.7 occurring in 2001.
The maximum value is 4.9 occurring in 2011.

Data table for Chart 1
Sepsis-associated mortality, by type of cause of death, age-standardized rates, Canada, 2000 to 2011
Table summary
This table displays the results of Sepsis-associated mortality, by type of cause of death, age-standardized rates, Canada, 2000 to 2011. Rate, Lower 95% Confidence interval and Upper 95% Confidence interval (appearing as column headers).
  Rate Lower 95% Confidence interval Upper 95% Confidence interval
Underlying and contributing
2000 17.8 17.4 18.3
2001 18.5 18.0 18.9
2002 18.9 18.5 19.4
2003 20.2 19.7 20.6
2004 20.9 20.4 21.4
2005 22.2 21.7 22.6
2006 26.4 25.9 26.9
2007 27.8 27.3 28.3
2008 27.7 27.2 28.2
2009 27.5 27.0 28.0
2010 27.2 26.7 27.6
2011 27.2 26.7 27.7
Contributing
2000 14.0 13.6 14.4
2001 14.8 14.4 15.2
2002 14.9 14.5 15.3
2003 16.0 15.6 16.4
2004 16.7 16.3 17.1
2005 17.9 17.5 18.3
2006 22.1 21.6 22.5
2007 23.2 22.8 23.7
2008 22.9 22.5 23.4
2009 22.7 22.3 23.2
2010 22.5 22.1 23.0
2011 22.2 21.8 22.6
Underlying
2000 3.8 3.6 4.0
2001 3.7 3.5 3.9
2002 4.0 3.8 4.2
2003 4.2 4.0 4.4
2004 4.1 3.9 4.3
2005 4.2 4.0 4.4
2006 4.4 4.2 4.6
2007 4.6 4.4 4.8
2008 4.8 4.6 5.0
2009 4.8 4.6 5.0
2010 4.7 4.5 4.9
2011 4.9 4.7 5.1

The increase in sepsis-associated deaths between 2000 and 2007 was mostly driven by increases in the recording of sepsis as a contributing cause of death. During this time period, the increase in the mortality rate when sepsis was a contributing cause of death (8.5%) was almost three times higher than when it was reported as the underlying cause (2.9%). While this increase in sepsis-associated mortality may reflect an actual increase in deaths involving sepsis, it may also be influenced by other factors including: changes in diagnosis methods and definitions, an increased awareness of sepsis by healthcare professionals, or changes in how deaths are classified.Note 19,Note 20

Males have a higher risk of sepsis-associated death than females

Males had a significantly higher age-standardized rate of sepsis-associated deaths, at 31.7 deaths per 100,000, compared with females (23.8 deaths per 100,000) in 2011. This trend, of higher sepsis-associated deaths in males, was present in all years between 2000 and 2011 (Chart 2).

One possible reason why males are more likely to have sepsis involved in their death is that they may be more likely than females to develop sepsis in the first place.Note 16,Note 21 This difference may be related to hormonal and genetic differences between males and females or differences in the presence of other health conditions.Note 22,Note 23

Between 2000 and 2007, sepsis-associated deaths increased significantly for males and females separately, with males having a higher rate of sepsis-associated deaths than females for all years (Chart 2). Since 2007, however, the gap between the sexes has started to close. Between 2007 and 2011, the sepsis-associated mortality rate among males decreased significantly, declining by 1.7% per year, while there was no significant change in the death rate for females during these years. As a result, the difference in the rates between males and females fell from a gap of 10.5 deaths per 100,000 in 2007, to 7.9 per 100,000 in 2011.

Chart 2 - Sepsis-associated mortality

Description for Chart 2

The title of the graph is "Chart 2 Sepsis-associated mortality, age-standardized rates, by sex, Canada, 2000 to 2011."
This is a line chart.
There are in total 12 categories in the horizontal axis. The vertical axis starts at 0 and ends at 40 with ticks every 5 points.
There are 2 series in this graph.
The vertical axis is "deaths per 100,000."
The units of the horizontal axis are years from 2000 to 2011.
The title of series 1 is "Male."
The minimum value is 21.8 occurring in 2000.
The maximum value is 33.9 occurring in 2007.
The title of series 2 is "Female."
The minimum value is 15.0 occurring in 2000.
The maximum value is 23.8 occurring in 2011.

Data table for Chart 2
Sepsis-associated mortality, age-standardized rates, by sex, Canada, 2000 to 2011
Table summary
This table displays the results of Sepsis-associated mortality, age-standardized rates, by sex, Canada, 2000 to 2011. Rate, Lower 95% Confidence interval and Upper 95% Confidence interval (appearing as column headers).
  Rate Lower 95%
Confidence interval
Upper 95%
Confidence interval
Male
2000 21.8 21.0 22.6
2001 22.5 21.7 23.2
2002 23.1 22.3 23.8
2003 24.4 23.6 25.2
2004 25.0 24.2 25.8
2005 26.8 26.0 27.5
2006 32.0 31.2 32.9
2007 33.9 33.0 34.7
2008 33.3 32.5 34.2
2009 32.6 31.8 33.4
2010 32.1 31.3 32.9
2011 31.7 30.9 32.5
Female
2000 15.0 14.5 15.6
2001 15.6 15.0 16.1
2002 15.9 15.4 16.5
2003 17.2 16.6 17.7
2004 17.7 17.2 18.3
2005 18.7 18.2 19.3
2006 22.3 21.7 22.9
2007 23.4 22.8 24.1
2008 23.7 23.1 24.3
2009 23.7 23.1 24.3
2010 23.4 22.8 24.0
2011 23.8 23.2 24.4

Sepsis-associated deaths increase with age

During the time period 2009–2011, the rate of sepsis-associated deaths increased by age among those 5 and older (Table 1). Deaths involving sepsis were highest among the oldest age group (85 and older) at 587.8 deaths per 100,000 and lowest between ages 5 and 19 (ranging from 0.7 to 0.8 deaths per 100,000). Infants under the age of 1 had a death rate of 33.0 deaths per 100,000, higher than the rates for those aged 1 to 59 years. The pattern of sepsis-associated deaths parallels that of deaths from all causes in Canada, where deaths are highest among the elderly and relatively high in those under the age of one.Note 24

Sepsis contributed to deaths in every age group, from newborn babies to the elderly. Among those who died by age 30, sepsis contributed to a relatively similar percentage of mortality, varying between 4.0% and 6.2% of all deaths. (Table 1). However among children (those who died before the age of 15) sepsis contributed to a greater percentage of mortality, varying between 6.7% and 8.2% of all deaths. Adolescents and young adults (ages 15 to 29) were the least likely to have sepsis contribute to their deaths, with 2.0% to 3.3% of all deaths involving sepsis. While children were more likely to have sepsis contribute to their death than adults, older people experienced higher overall mortality than young people, resulting in higher rates of sepsis-associated mortality.

Table 1
Age-specific sepsis-associated mortality rates, Canada, 2009 to 2011
Table summary
This table displays the results of Age-specific sepsis-associated mortality rates, Canada, 2009 to 2011. The information is grouped by Age group (appearing as row headers), Sepsis-associated mortality, calculated using Rate per 100,000 people and % of all deaths units of measure (appearing as column headers).
Age group Sepsis-associated mortality
Rate per 100,000 people % of all deaths
All ages 38.4 5.5
Under 1 33.0 6.7
1 to 4 1.2 6.9
5 to 9 0.7 8.2
10 to 14 0.8 6.7
15 to 19 0.8 2.2
20 to 24 1.1 2.0
25 to 29 1.7 3.3
30 to 34 2.5 4.0
35 to 39 4.4 5.3
40 to 44 6.0 4.8
45 to 49 10.1 5.1
50 to 54 17.0 5.3
55 to 59 27.3 5.5
60 to 64 45.4 5.9
65 to 69 72.4 6.0
70 to 74 117.7 6.1
75 to 79 199.2 6.2
80 to 84 323.0 5.9
85 and over 587.8 4.7

Sepsis contributes to more than one-half of all deaths from infectious disease

About 5.5% of all deaths in Canada during 2009–2011 were sepsis-associated.Note 15 Of the underlying causes of death that involved sepsis the most common was diseases of the skin and tissue under the skin (subcutaneous tissue) (61.8%), followed by infectious diseases (53.4%) (see Appendix A).Note 25,Note 26 Sepsis was also involved in over 10.0% of all deaths from diseases of the: digestive system, musculoskeletal system, and urinary system or genital organs.Note 27

Deaths from infectious diseases provide a good example of how sepsis can be both an underlying and contributing cause of death. During 2009–2011, 53.4% of all infectious disease deaths involved sepsis (Table 2). In 82.2% of those deaths, sepsis was the underlying cause of death; while in the other 17.8%, sepsis was a contributing cause of death.

Of the 17.8% of infectious disease deaths where sepsis was a contributing cause there were a number of underlying causes of death. For example, sepsis was involved in almost a quarter of all deaths from bacterial infections of the intestinesNote 28 (e.g., e. coli and c. difficile); and over 10.0% of all deaths from: tuberculosis, viral infections involving skin lesionsNote 29 (e.g., chickenpox, herpes zoster), HIV and viral hepatitis.

Table 2
Sepsis-associated mortality, by type of infectious disease death, Canada, 2009 to 2011
Table summary
This table displays the results of Sepsis-associated mortality, by type of infectious disease death, Canada, 2009 to 2011. The information is grouped by Underlying cause of death (appearing as row headers), Sepsis-associated mortality, calculated using % of all deaths units of measure (appearing as column headers).
Underlying cause of death Sepsis-associated mortality
% of all deaths
Infectious and parasitic diseases 53.4
Sepsis 100.0
Intestinal bacterial infections 23.3
Tuberculosis 15.0
Viral infections involving skin lesions 12.2
HIV 11.3
Viral hepatitis 10.8
Other underlying cause 7.7

Conclusion

Of all of the Canadians who died in 2011, sepsis played a role in the death of one in 18. Sepsis was more likely to be reported as a contributing cause rather than the underlying cause of death itself. Including both underlying and contributing causes of death when estimating sepsis-associated deaths, allows for a more complete picture of how sepsis plays a role in overall mortality in Canada. It also means that sepsis-associated deaths can be examined using data from the Vital Statistics Program, providing an alternative to special clinical studies and hospitalization-based analyses.

As the Canadian population ages, sepsis is expected to continue to be a major contributor to mortality and an important health care issue.

Start of text box 3

Data source, method and definitions

Data source

The statistics presented in this article are derived from the multiple-causes-of-death data, produced as part of the Canadian Vital Statistics – Death Database (CVS:D) at Statistics Canada.Note 23 This dataset includes the underlying and contributing causes of death for deaths that occurred since 2000. Further information on the recording and coding of contributing cause of death information from the death certificate is available in the Vital Statistics User Guide.Note 13 The CVS:D includes demographic and cause of death information for all deaths from all provincial and territorial vital statistics registries in Canada. Prior to 2010, some data were collected on Canadian residents who died in some American states; these deaths were excluded from this analysis. Starting with the 2010 reference year, data on Canadian residents who died in American states are no longer collected.

Method

Trends in age-standardized incidence and death rates over time were determined using the Joinpoint Regression Program (v.4.0.4) distributed by the SEER (Surveillance Epidemiology and End Results) program.Note 30 All trend analyses were performed using age-standardized rates.

To allow for more valid comparisons over time and between populations, age-standardized rates are used to control for differences in population age-structure. The age-standardized rate is the rate that would occur if the population of interest had the same age-structure as the standard population. The standard population used in this article is the July 1, 1991 Canadian population.

Age-standardized rates are used throughout this paper except when rates are given for a specific age group. Rates for individual age groups are called age-specific rates and have not been age-standardized.

Definitions

Death data were classified according to the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10).Note 10 There are no codes that directly identify sepsis in the ICD-10. Following previous research into sepsis deaths that used administrative databases, this study used ICD codes for septicaemia as a proxy for sepsis.Note 8,Note 16,Note 21 As there is no standardized coding for sepsis the definition used in this article may not match that used in other studies.

Sepsis was defined by the following codes:
Table summary
This table displays the results of Sepsis was defined by the following codes:. The information is grouped by International Statistical Classification of Diseases and Related Health Problems, Tenth Revision Chapter (appearing as row headers), International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes (appearing as column headers).
ICD-10 Chapter ICD-10 codes
Infectious and parasitic diseases A02.1, A20.2, A20.7, A31.7, A22.7, A26.7, A32.7, A39.2, A39.4, A40, A41, A42.7
Pregnancy, childbirth and the puerperium O85
Conditions originating in the perinatal period P36, P37.2
Symptoms, signs and abnormal clinical and laboratory findings R57.8

The underlying cause of death was defined as the cause of death reported on the lowest line of Part I (Disease or condition directly leading to death) of the death certificate. Contributing causes were defined as any causes listed on the other lines of Part I and any causes listed in Part II (other significant conditions contributing to the death).

Limitations

Multiple-cause-of-death data was not available for parts of the years 2000 to 2002 for Saskatchewan and Yukon. While the total proportion of missing data was small (2.8% of all deaths between 2000 and 2010),Note 13 an analysis of the data found that the frequency of missing data was not evenly distributed within demographic variables. Therefore there is a potential for bias.

This study relies on the accuracy of the information reported on the death certificate. Unlike leading causes of death, data on contributing causes of death are subjectively documented and may be dependent upon the knowledge, expertise and available time of the person filling out the death certificate.Note 19

As there are no codes for sepsis in the ICD-10 this study used septicaemia codes as a proxy for sepsis. While this approach has been validated for ICD-9 and applied to ICD-10 in other studies, it has not yet been validated for ICD-10.Note 8,Note 16,Note 21 This may have led to misclassification and a possible underestimation of the rate of sepsis-associated death.Note 8

End of text box 3

Start of text box 4

Owen Phillips is an analyst with the Health Statistics Division. Tanya Navaneelan and Sarah Alam were with the Health Statistics Division and Paul Peters was with the Health Analysis Division.

End of text box 4
Appendix A
Sepsis-associated mortality, by selected underlying cause of death, Canada, 2009-2011
Table summary
This table displays the results of Sepsis-associated mortality, by selected underlying cause of death, Canada, 2009-2011. The information is grouped by Underlying cause of death (appearing as row headers), Sepsis recorded as underlying cause of death, Sepsis recorded as contributing cause of death and Sepsis-associated mortality, calculated using number of deaths and % of all deaths units of measure (appearing as column headers).
Underlying cause of death Sepsis recorded as underlying cause of death Sepsis recorded as contributing cause of death Sepsis-associated mortality
number of deaths % of all deaths
Diseases of the skin and tissue 0 735 61.8
Infectious and parasitic diseases 6,805 1,505 53.4
Diseases of the digestive system 0 5,270 18.3
Diseases of the urinary system and genital organs 0 2,695 16.9
Diseases of the muscles, skeletal system and connective tissue 0 730 15.9
Conditions originating during the period immediately before and after birth 125 125 7.6
Diseases of the respiratory system 0 3,840 5.9
Endocrine, nutritional and metabolic diseases (e.g. diabetes, metabolic syndromes) 0 1,715 5.9
Development disorder of the embryo or fetus (congenital anomalies) 0 155 5.8
Mental and behavioural disorders 0 1,500 4.0
Diseases of the nervous system 0 1,315 3.7
Cancer 0 7,140 3.2
Diseases of the circulatory system 0 4,655 2.3
External causes (e.g. accidents, homicide, suicide) 0 865 1.8
All causes of death 7,035 32,310 5.5

Start of text box 5

Related material for this article

Related articles

Mortality from diabetes mellitus, 2004 to 2008: A multiple-cause-of-death analysis

Deaths, by cause, Chapter 1: Certain infectious and parasitic diseases

Data sources

Canadian Vital Statistics – Deaths Database

Bibliographic references

How to cite this article:
Navaneelan T, S. Alam, P.A. Peters, and O. Phillips. 2015. "Deaths involving sepsis in Canada". Health at a Glance. Statistics Canada catalogue no. 82-624-X.

End of text box 5


Date modified: