Economic Burden of Illness in Canada, 1993

[Previous][Table of Contents]

Indirect Costs

Mortality Costs

Methods

Using the human capital approach, mortality costs are represented by the current monetary value of future productivity lost due to premature mortality. The estimated cost or value to society of all deaths is the product of the number of deaths and the discounted value, with age and sex taken into account, of an individual's lifetime productivity.¹

The discounted present value of lifetime productivity (future labour force work and unpaid work) by age and sex is calculated using the lifetime productivity loss model (refer to Appendix 5).^2,3 The model sums productivity in the current year plus productivity expected in future years if the individual continues to live. This model accounts for life expectancy for different age and sex groups,⁴ earnings at successive ages,⁵ varying labour force participation rates,⁵ the value of unpaid work,⁶ the productivity growth rate ¹ and the appropriate discount rate to convert a stream of costs or benefits into present worth.¹

We apply a discount rate of 6% to both employment earnings and the value of unpaid work to reflect the present value of future productivity. A sensitivity analysis using discount rates varying from 2% to 10% provides a range of possible lifetime productivity losses (see Appendix 5). Productivity growth rates of 1% for labour force work and 0% for unpaid work are applied to account for future productivity gains. These rates reflect conservative estimates of projected and past experience in Canada and the United States.^1-3

The 1993 national average supplementary labour income rate ⁷ is applied to the discounted value of future labour force work to account for wage supplements (i.e. employer contributions to employee welfare, pensions [CPP/QPP], workers' compensation and unemployment insurance funds).

The value of future unpaid work is added to the adjusted value of future labour force work for each age group and sex. The resulting values are multiplied by the number of deaths in 1993 ⁸ for each diagnostic category, age group and sex to obtain final estimates of the present value of future productivity loss due to premature mortality.

Assumptions

• The 1991 survival pattern is assumed to reflect 1993 life expectancy.
• We use average annual earnings for all earners, rather than full year, full-time workers,⁹ to provide a more accurate estimate of future earnings that reflect existing labour variations (i.e. seasonal and part-time work).
• It is assumed that no earnings are earned between the ages of 0 and 14. Earning patterns for people aged 75-79, 80-84 and 85 or older are assumed to vary according to patterns for US residents in these age cohorts.^10,11
• The future pattern of earnings for an average individual within a sex group is assumed to follow the pattern reported by Statistics Canada in 1993.⁵ The average individual may expect his or her earnings to rise with age and experience in accordance with the cross-sectional data for 1993.
• People are assumed to be working and productive during their expected lifetime in accordance with the current patterns for their age group and sex.¹ No distinction is made between those in or those out of the labour force at the time of death.
• Average values of annual unpaid work for age 85 and over were inferred from the values for age 75 and over, using the US ratios for these age groups.¹¹
• Average values of unpaid work in 1993 are based on 1992 replacement cost generalist estimates of unpaid work by age group and sex, inflated to 1993 dollars.⁶

Limitations

• Survival probabilities are averaged across five-year time periods.⁴
• Average earnings are estimated across five-year age groups ⁵ and the value of unpaid work is estimated across ten-year age groups.⁶
• Employment income and the value of unpaid work for children under 15 are not available.
• Employment income and the value of unpaid work for ages 75 and over are not disaggregated due to sample size limitations.^5,6

Results

The present value of future productivity lost due to premature mortality in 1993 was an estimated $29.3 billion at a 6% discount rate. A sensitivity analysis produced productivity losses for premature mortality between $22.3 billion (using a discount rate of 10%) and $43.7 billion (using a discount rate of 2%). The figure in Appendix 5 illustrates the results of the sensitivity analysis.

Cancer, cardiovascular diseases and injuries had the highest mortality costs in 1993. These three diagnostic categories represented three quarters (75.1%) of the total mortality costs in 1993.

Coronary heart disease ($4.6 billion) accounted for two thirds (61.8%) of the mortality costs for cardiovascular diseases. Diabetes was responsible for almost three quarters (73.0%, $559 million) of the mortality costs of endocrine and related disorders. Chronic bronchitis, emphysema and asthma ($734 million) represented just over half (52.6%) of the mortality costs of respiratory diseases.

[Figures]

Morbidity Costs Due to Long-term Disability

Methods

The value of productivity lost to long-term disability is calculated for the household and institutionalized population.

The National Population Health Survey (NPHS) household component ¹ provides the number of people who reported a long-term disability by diagnostic category, age group and sex. These figures are adjusted for severity and annual average length of long-term disability, applying distributions from the Quebec Health and Social Survey (QHSS).^2,3 Weights are assigned to account for lost productivity at different levels of long-term disability ^4-7 (see Appendix 6).

The NPHS institutional component ⁸ provides the number of people living in Canadian long-term health care facilities by age and sex. These figures, multiplied by the distribution of long-term disability in institutions according to diagnostic category, age group and sex ⁹ and the annual average length of stay in institutions,¹⁰ provide estimates of the number of people in long-term health care facilities by diagnostic category, age group and sex. Weights are applied to account for productivity loss at different levels of long-term disability.^4-7

The adjusted figures for long-term disability by diagnostic category, age group and sex for the household and institutionalized populations are summed. These figures, multiplied by the 1993 annual average value of labour force work,¹¹ adjusted for wage supplements ¹² and unpaid work,¹³ are used to estimate the total value of productivity lost to long-term disability by diagnostic category, age group and sex.

Assumptions

• The distribution of severity levels and the annual average length of long-term disability obtained from the QHSS are assumed to represent those of the Canadian household population.
• The estimated annual average length of stay in institutions from NPHS cross-sectional data is assumed to represent the annual average length of stay consistently over the year 1993.
• Annual average labour force earnings and the value of unpaid work are assumed to be the same for people in and not in the labour force. People living in institutions are assumed to have the same potential annual average labour force earnings and value of unpaid work as the household population.
• Labour force earnings have been adjusted for wage supplements. The 1993 supplementary labour income rate (13.78%) is assumed to be the same for various age and sex groups.
• The annual average value of unpaid work for youths aged 12-14 is not available and is assumed to be half that for young people aged 15-19.
• Loss of productivity to long-term disability by diagnostic category is based on the main health problem (by ICD-9 code or V-code) obtained from the NPHS.
• The 9% of respondents to the NPHS (institutional component)⁸ who did not answer the long-term disability question are assumed to have the same prevalence of long-term disability as people who did respond.

Limitations

• The QHSS ^2,3 is used to estimate the length and severity of long-term disabilities since the NPHS household component ¹ did not collect this information.
• Only two age groups (less than 65, 65 and older) were available from the NPHS institutional component due to small numbers within some diagnostic categories. The estimated costs for youths aged 12-14, from the NPHS household component, should be treated with caution because of small numbers in this age group.
• Indirect costs are slightly underestimated since the NPHS ^1,8 excludes people residing at home or in health facilities on Indian reserves, on Canadian Forces bases or in some remote areas.
• Productivity lost to long-term disability is calculated for people aged 12 and over. We cannot estimate the value of school work and unpaid work lost to long-term disability for children under age 12.
• Using the number of people living in institutions in 1995 to estimate the indirect costs due to long-term disability for 1993 may introduce bias.
• The NPHS uses V-codes developed by Statistics Canada to code musculoskeletal diseases. Injuries are included in this coding system. We apply the 1986 distribution of long-term disability costs for musculoskeletal diseases and injuries ¹⁴ to the 1993 cost of musculoskeletal diseases to split out the 1993 cost of injuries.

Results

The estimated value of productivity lost due to long-term disability in 1993 was $38.3 billion, based on weights of 0.9 for "very severe," and, for the household population only, 0.1 for "minor limitations" (see Appendix 6). The institutionalized population represented $3.1 billion of this total.

A sensitivity analysis, using a range of weights, produced productivity losses from $31.8 billion (0.8; 0.0) to $44.7 billion (1.0; 0.2).

Musculoskeletal diseases, the leading cause of long-term disability, accounted for $13.5 billion, roughly one third (35.2%) of long-term disability costs. Nervous system and sense organ diseases (14.9%, $5.7 billion) and cardiovascular diseases (11.8%, $4.5 billion) also had high disability costs.

Arthritis and musculoskeletal disorders of the back and spine each represented a third of the cost of all musculoskeletal diseases ($4.4 billion; $4.3 billion). Sight and hearing disorders made up 36.6% of the cost of nervous system and sense organ disorders ($1.2 billion; $943 million). Asthma and bronchitis/emphysema ($1.8 billion; $424 million) accounted for 85.3% of the cost of long-term disability due to respiratory diseases.

[Figures]

Morbidity Costs Due to Short-term Disability

Methods

The National Population Health Survey ¹ provides the average number of days of short-term disability by age and sex for two levels of severity: "days in bed" or "days of reduced major activity." Weights are assigned to these levels to account for the loss of productivity at different severity levels of short-term disability: 0.9 for "days in bed" and 0.5 for "days of reduced major activity"^2-5 (see Appendix 6). A sensitivity analysis of productivity losses using weights of 0.8 and 1.0 for "days in bed" has also been conducted.

The adjusted values are applied to the general population ⁶ to estimate total annual days of productivity lost due to short-term disability by age and sex. Annual days of productivity lost due to short-term disability by diagnostic category, age group and sex are generated by applying these values to the distribution of days lost due to short-term disability by diagnostic category, age group and sex obtained from the Quebec Health and Social Survey (QHSS).⁷

The number of annual days of productivity lost according to diagnostic category, age group and sex is then multiplied by an average value per day of labour force work,⁸ adjusted for wage supplements ⁹ and unpaid work,¹⁰ to estimate the value of productivity lost to short-term disability by diagnostic category, age and sex. Labour force earnings and the value of unpaid work are available by age and sex.

Assumptions

• The distribution of days lost to short-term disability by diagnostic category, age group and sex obtained from the QHSS is assumed to reflect the distribution for the Canadian population. Quebec represents 25% of the Canadian population.¹¹
• It is assumed that there are 250 days of labour force activity and 365 days of unpaid work per year.
• Average daily earnings and the value of unpaid work by age and sex are applied similarly to people in the labour force and those not in the labour force.
• The average daily value of unpaid work for youths aged 12-14 is not available. Unpaid work for this age group is assumed to be half that for youths aged 15-19.
• Labour force earnings have been adjusted for wage supplements. The 1993 supplementary labour income rate (13.78%) is assumed to be the same for various age and sex groups.

Limitations

• Productivity lost to short-term disability is underestimated because the NPHS does not include people living on Indian reserves, or on Canadian Forces bases or children under age 12.
• The National Population Health Survey household component ¹ does not provide short-term disability by diagnostic category. Applying the distribution of short-term disability by diagnostic category obtained from the QHSS may introduce bias.
• We may underestimate the economic impact of short-term disability by excluding people with long-term disabilities from our calculation.

Results

The total productivity lost due to short-term disability was an estimated $17.5 billion in 1993. The leading cause of short-term disability was respiratory diseases ($4.4 billion, 24.8%), followed by injuries ($3.2 billion, 18.5%), musculoskeletal diseases ($1.8 billion, 10.0%) and digestive diseases (1.2 billion, 6.5%).

A sensitivity analysis produced productivity losses for short-term disability of $16.8 billion to $18.3 billion, using respective weights of 0.8 and 1.0 for the severity level "days in bed" (see Appendix 6).

[Figures]

[Previous][Table of Contents]