Report 3: Economic Burden of Illness in Canada, 2005–2008 – EBIC physician care expenditures, 2005–2008

Report 3: EBIC Physician Care Expenditures, 2005-2008

1. Background

Physician care expenditures include fee-for-service payments made by provincial/territorial medical care insurance plans to physicians in private practice, as well as alternative forms of payment (salaries, sessional, capitation) made to physicians (5). Fees for services rendered in hospitals are also included in the physician care expenditures component when the provincial/territorial medical insurance plans make payments directly to the physicians. While previous versions of EBIC utilized provincial and territorial physician fee-for-service claims data, this was not feasible for EBIC 2005-2008. As an alternative, Manitoba’s publicly available fee-for-service physician care expenditure totals, by sex and ICD-9 chapter, were used along with EBIC 2000 data to distribute NHEX physician cost totals across EBIC categories (diagnostic category/subcategory, sex, age group and province/territory) for each year of analysis (5,33). Information on physician care expenditures for services remunerated by alternative payment methods was not available by diagnostic category for any province or territory. This report describes the data sources and methods used to derive the 2005-2008 physician care expenditure estimates. It also presents and discusses the results and limitations of the data used.

2. Data Sources

Publicly available Manitoba fee-for-service physician care expenditure totals by sex and ICD-9 chapter were obtained from the Manitoba Health Annual Statistics (43-46). Unpublished EBIC 2000 physician care expenditure data were obtained from PHAC. NHEX province/territory annual physician cost totals were obtained from CIHI (5).

3. Methods

The distribution of Manitoba's fee-for-service physician care expenditures according to ICD-9 chapter, and sex, was used to distribute NHEX provincial/territorial physician cost totals by ICD-9 chapter and sex for all provinces/territories.Footnote 61 For jurisdictions other than Manitoba, the jurisdictional cost distribution used to distribute the NHEX cost totals was obtained by multiplying Manitoba's cost per capita for each ICD-9 chapter and sex group by the appropriate provincial/territorial population count.

For all provinces/territories, the sex-specific ICD-9 chapter physician care expenditure totals were then distributed across individual ICD-9 codes and age groups using the EBIC 2000 cost distributions specific to each province/territory and ICD chapter. For all analyses, the EBIC 2000 cost distribution was adjusted for differences in population between the year 2000 and the year of analysis.Footnote 62 In PHAC's EBIC 2000 database, Manitoba and Newfoundland did not have records for the ICD-9 chapter 'Factors Influencing Health Status and Contact with Health Services'. Physician care expenditure totals for these chapters were distributed across individual ICD codes and age groups using the EBIC 2000 national (excluding Manitoba and Newfoundland) cost distribution for the chapter, after adjusting for differences in population. Costs by ICD code were then grouped into the EBIC diagnostic categories and subcategories according to the groupings described in Appendix C.

4. ResultsFootnote 63

4.1 Expenditures by Diagnostic Category

Table 8 illustrates EBIC physician care expenditures by diagnostic category for the years 2005-2008. In 2008, the five diagnostic categories with the largest expenditures were cardiovascular diseases ($2.4 billion, 9.9%), neuropsychiatric conditions ($2.3 billion, 9.9%), musculoskeletal diseases ($2.0 billion, 8.4%), genitourinary diseases ($1.6 billion, 6.8%) and injuries ($1.4 billion, 6.0%).

4.2 Expenditures by Diagnostic Category and Sex

Table 9 illustrates EBIC 2008 physician care expenditures by diagnostic category and sex. In 2008, 41.3% ($9.8 billion) and 58.7% ($13.9 billion) of expenditures were attributable to males and females respectively. The three diagnostic categories with the largest expenditure for males were cardiovascular diseases ($1.2 billion), neuropsychiatric conditions ($0.9 billion) and musculoskeletal diseases ($0.8 billion). For females, the three diagnostic categories with the largest expenditure were neuropsychiatric conditions ($1.4 billion), musculoskeletal diseases ($1.2 billion) and genitourinary diseases ($1.1 billion).

The five diagnostic categories with the largest difference in cost distribution across the sexes were genitourinary diseases (31.5% male, 68.5% female), nutritional deficiencies (34.0% male, 66.0% female), certain infectious and parasitic diseases (36.2% male, 63.8% female), other neoplasms (36.7% male, 63.3% female) and endocrine disorders (38.1% male, 61.9% female).Footnote 64

4.3 Expenditures by Diagnostic Category and Age Group

Figure 18 illustrates EBIC 2008 physician care expenditures for each age group. Individuals aged 0-14 years incurred the lowest percentage of physician care expenditures (9.1%) and individuals aged 35-54 years the highest (26.8%). Additionally, individuals aged 55 years and older accounted for approximately 44% of total EBIC physician care expenditures.

Figure 19 illustrates EBIC 2008 physician care expenditures by diagnostic category and age group for the five most costly diagnostic categories. Expenditures of the highlighted diagnostic categories were highest for individuals aged 35-54 years, except in the cardiovascular diseases category.

5. Discussion and Limitations

Record-level fee-for-service claims data were not obtained from the respective provinces/territories and therefore could not be used to estimate EBIC 2005-2008 physician care expenditures. Access to record-level claims data offers high value to the EBIC publication, as these records contain information on physician care expenditures by ICD code, sex, age group and province/territory. In the absence of claims data, assumptions had to be made using the available data, and the results may not reflect the true distribution of physician care expenditures across EBIC categories.

Manitoba was the only province/territory with publicly available physician care expenditures by diagnostic category (specifically, by ICD-9 chapter and sex). These cost distributions were used to estimate the sex-ICD chapter specific cost distributions for all other provinces/territories. There are several limitations to these methods. Primarily, Manitoba's costs per capita may not be an appropriate method to cost for other provinces/territories. The costs per capita were estimated by sex but not by age group, given that the Manitoba ICD chapter expenditure totals were only available by sex. Thus, differences in the age distribution of Manitoba's population and that of the other provinces/territories could not be adjusted for. Furthermore, using Manitoba's cost per capita distribution to cost for other provinces/territories does not consider that the prevalence of certain health conditions may vary among jurisdictions. Also, in 2008, Manitoba represented a very small percentage (3.6%) of the overall Canadian population and may have had very different distributions of urban/rural, aboriginal and other minority populations. All these factors may make Manitoba's cost per capita distribution an inaccurate proxy for the other provinces/territories.

Table 10 illustrates the comparison between the EBIC 2000 cost distribution by diagnostic category for Manitoba and Ontario. The three diagnostic categories with the largest difference (in absolute value) were neuropsychiatric conditions (6.5%), symptoms, signs and ill-defined conditions (3.0%) and injuries (2.6%). For the remaining categories, 13 are associated with a difference of 1% or less and 5 are associated with a difference of greater than 1% and less than or equal to 2%. Overall, Table 10 shows that for most diagnostic categories (85.7%), Manitoba's 2000 cost distribution was similar to Ontario's. The significance of a magnitude of difference may vary with personal opinion, with some individuals considering a difference of less than 1% or 2% to be significant. Although the EBIC 2000 cost distribution for Manitoba may be considered to closely reflect that of Ontario's, this may not be the case for the years 2005-2008, and using Manitoba's cost distribution to produce estimates for Ontario (and other provinces/territories) may have resulted in inaccuracies. Ontario was used as the comparator in Table 10; however, larger or smaller differences may exist for other provinces/territories.

Fee-for-service physician care expenditures by ICD code were not publicly available for any province/territory. As the ICD code groupings for the EBIC diagnostic categories are different from those in the ICD chapters, it was necessary to distribute costs for an ICD chapter across ICD codes, so that costs could then be re-grouped into the EBIC diagnostic categories. Additionally, it was necessary to distribute costs within an ICD chapter across ICD codes, so that costs by EBIC diagnostic subcategory could be obtained. As mentioned in the methods section of this report, ICD chapter costs were distributed across ICD codes using each respective jurisdiction’s EBIC 2000 cost distribution (after adjustment for changes in population). Table 11 illustrates the difference between a Manitoba 2008 cost distribution obtained by adjusting EBIC 2000 Manitoba data for population changes and Manitoba's 2008 cost distribution taken from the Manitoba Health Annual Statistics publication. The largest difference (in absolute value) is for the diagnostic category ‘injuries’ (3.3%). For the remaining diagnostic categories, 16 were associated with a difference of 1% or less, 2 were associated with a difference of greater than 1% and equal to or less than 2%, and 2 were associated with a difference of greater than 2% and equal to or less than 4%. Table 11 shows that adjusted EBIC 2000 data may be a reasonable proxy for other years (at least for Manitoba in 2008). Although, adjusted EBIC 2000 estimates may not produce substantial differences in the cost distribution by EBIC diagnostic category, differences (in absolute value) may be larger at the diagnostic subcategory level. Using EBIC 2000 data to distribute costs across ICD code does not take into consideration that the distribution of costs within an ICD chapter may be different across the years, even after adjustment for population changes. Costs were also distributed across age groups using the EBIC 2000 cost distribution specific to province/territory, sex and ICD chapter; similar limitations follow from these methods.

Record-level data or aggregated data by diagnostic category, sex and age group for alternative physician payment methods (salaries, sessional, capitation) were not available. CIHI’s National Physician Database (NPDB) showed that in 2008 the total national clinical payments made to physicians remunerated on a fee-for-service basis were 73.1% of total clinical payments, a decrease from 89.3% in 1999 (47).Footnote 65 As physician remuneration by fee-for-service method declines, ways of capturing the services by cost, patient diagnosis and demographic information when physicians are paid by other methods become more important. Shadow billing and/or physician surveys for services remunerated using alternative payment methods could produce valuable information, especially if physician remuneration methods influence the treatment that physicians provide to their patients and/or if patient characteristics (e.g. chronic illness, age) vary with payment method. The fee-for-service cost distributions used in this report may not accurately represent the cost distributions across EBIC category for physician services remunerated using alternative methods. In the absence of individual claims data for all provinces and territories, serious limitations exist when looking at specific diseases, especially those whose prevalence may change in response to exposure (e.g. food-borne illness outbreaks). The estimates in this report are not sensitive enough to capture these fluctuations in specific diseases across years and provinces/territories. Therefore, comparisons of EBIC physician care expenditure estimates across the years 2005-2008 and across provinces/territories are not recommended.

6. Conclusion

EBIC 2005-2008 physician care expenditures were attributed by EBIC categories for fee-for-service remuneration using publicly available Manitoba data, in conjunction with EBIC 2000 data. The three diagnostic categories with the highest expenditures were cardiovascular diseases ($2.4 billion, 9.9%), neuropsychiatric conditions ($2.3 billion, 9.9%), musculoskeletal diseases ($2.0 billion, 8.4%).   Males accounted for 41.3% of 2008 physician care expenditures while females accounted for 58.7%. EBIC 2008 physician care expenditures were lowest and highest for individuals aged 0-14 years (9.1%) and 35-54 years (26.8%) respectively.

The estimates in this report offer value in that they can be added to other EBIC cost components to obtain an estimate of the economic burden of illness and injury in Canada. EBIC 2005-2008 physician care expenditures should not be compared across years or provinces/territories. Given the limitations mentioned in this report, these expenditures may not accurately represent the cost burden by illness and injury, especially when more disaggregated disease categories are examined (e.g. EBIC diagnostic subcategories).



Page details

Date modified: