OBJECTIVE—To analyze trends in health care costs in the decade after identification of diabetes, contrasting type 1 and 2 diabetes.

RESEARCH DESIGN AND METHODS—The Canadian National Diabetes Surveillance System criteria were applied to administrative databases to identify incident diabetes cases in 1992. Cases were categorized as type 1 or type 2 diabetes based on patterns of drug use. Per capita health care costs (in 2001 Canadian dollars) for five resource categories were estimated according to the type of diabetes, for the year before identification (1991) and 10 years after (1992–2001) identification of the cases.

RESULTS—We identified 156 type 1 and 3,469 type 2 incident cases of diabetes, from a population base of ∼950,000. The mean (±SD) age of case subjects at index was 61.2 ± 16.7 years, and 54% of subjects were male. Overall annual per capita health expenditures rose considerably in the year after identification of diabetes but then stabilized at a lower level for the next 9 years, ranging from $3,800 to$4,400. From 1992 to 2001, diabetic individuals used $137.1 million in health care resources, most of which (96%) was attributable to type 2 diabetes. The average 10-year cost per individual with diabetes was$37,820 ($33,684 per type 1 and$38,006 per type 2 diabetes case; adjusted P = 0.45).

CONCLUSIONS—Total health expenditures for diabetes are driven by the much larger prevalence of type 2 compared with type 1 diabetes. Policymakers need to acknowledge and allocate resources for diabetes prevention and management accordingly.

It is generally recognized that the costs of treating chronic medical conditions are rising for health care systems because of an aging population and the increasing prevalence of such conditions. Attempts to quantify the cost of these chronic medical conditions have often relied on models, synthesizing estimates using data from various sources, or surveys of relatively small samples of diseased individuals. Diabetes, for example, has been the focus of a large body of literature (1), with cost estimates usually based on modeled values (2,3).

Diabetes, however, is a difficult condition to study because it includes both type 1 and type 2 diabetes, two very different conditions in terms of pathophysiology and epidemiology (4). Furthermore, most cost-of-diabetes estimates have been derived from prevalence cohorts or cross-sectional analyses (1). Few studies have followed the health care costs of an incident cohort with diabetes over time, an approach where the natural history of the disease can be more accurately captured (1,57).

Estimating the progressive costs of diabetes is important for resource allocation, health policy, and perhaps even allocation of research funding. Accurate costing is important, given that a substantial portion of costs are not attributable directly to the condition itself but to accumulated complications and comorbidities (13,810). The purpose of this study was to analyze trends in actual health care use and estimated costs in the year before and 10 years after identification of diabetes in an incident cohort of individuals with diabetes and to contrast the cost of care for people identified with type 1 and 2 diabetes.

Ethical approval for this study was obtained from the Health Research Ethics Board of the University of Alberta. The Canadian National Diabetes Surveillance System (NDSS) criteria (4) were applied to Saskatchewan Health's administrative databases to identify incident cases of diabetes in 1992. Using the NDSS case definition, individuals are considered to have diabetes if they have two physician visits with a diagnosis of diabetes (ICD code 250) on 2 different days within any 730-day period or one hospitalization with a discharge diagnosis of diabetes (4).

Individuals were then categorized as having type 1 or type 2 diabetes based on patterns of drug use during the 11-year study period. Individuals who had insulin dispensed within 6 months of their index date and no oral agents at any point during the follow-up period were classified as having type 1 diabetes; all others were considered to have type 2 diabetes.

### Estimation of health care costs

Health care services (i.e., prescriptions, physician visits, hospitalizations, day surgeries, and dialysis) in each calendar year were identified from services recorded in the databases. The prescription database contains out-of-hospital medication claims for all beneficiaries eligible for provincial drug plan benefits. Annual prescription costs were obtained directly from each claim record and inflated to year 2001 dollars using the Consumer Price Index (11).

Annual physician costs were also obtained directly from service claim records in Saskatchewan Health's databases. Physician visits included visits to general practitioners, specialists, and physicians out of the province, as well as visits to other practitioners who provided insured services, such as optometrists.

The databases contained the number of hospitalizations and day surgeries each year. The cost per hospitalization in 2001 Canadian dollars was determined by multiplying the resource intensity weight (12), calculated by the Canadian Institute of Health Information, by the funding per weighted case for the 2001–2002 fiscal year ($3,369.77) (12,13). The diagnoses most responsible for hospitalizations were identified from claims records. The cost per day surgery was estimated by multiplying the day procedure group weight by the same funding per weighted case. Fee-for-service billing codes were used to identify individuals who received dialysis. Annual dialysis costs were calculated by multiplying the proportion of each calendar year on either hemodialysis or peritoneal dialysis by an estimated annual cost of each dialysis modality (14). These latter cost estimates were based on a prospective observational study of patients who attended dialysis clinics in a neighboring province (14) and amounted to$52,719 per year of hemodialysis and $37,431 per year of peritoneal dialysis, in 2001 Canadian dollars. ### Analysis Per capita health care costs (in year 2001 Canadian dollars) for each resource category were calculated according to type of diabetes and type 1 and type 2 diabetes combined for the year before index (1991) and 10 years after identification of diabetes (1992–2001). Per capita health care costs for the five resource categories combined were also determined. These estimates were based on the number of active cases in each follow-up year. The cumulative and average costs for the entire decade of follow-up were determined based on the entire cohort. We compared the average per capita cumulative cost of care for type 1 and type 2 diabetes after adjusting for age, sex, and comorbidities using ANCOVA models. Analyses were performed using SPSS (version 13.0; SPSS, Chicago, IL). Using the NDSS criteria, 3,625 incident diabetes cases were identified for the 1992 calendar year. Individuals with type 1 (n = 156) and type 2 (n = 3,469) diabetes were similar in demographic characteristics, with the exception of age at index (Table 1). The average duration of follow-up was longer in the type 2 diabetic group, but overall mortality was similar. Hospitalizations accounted for the largest proportion of total per capita costs each year (Table 2). In the incident year (1992), per capita hospital costs were several times higher than the year prior to diagnosis (1991), particularly for individuals with type 1 diabetes, where costs rose from$738 in 1991 to $5,212 in 1992. Hospital costs decreased over the 10 years of follow-up for both type 1 and type 2 diabetes but were often higher in type 2 than type 1 diabetes (Table 2). From 1991 to 2001, cardiovascular disease was most frequently the reason for hospitalization. In the index year (1992), however, diabetes (ICD codes 250.0, 250.1, 250.2, 250.7, and 250.9) was the diagnosis responsible for more hospitalizations than was cardiovascular disease (14.2 vs. 13.7%). In the years after identification of diabetes, prescription costs initially decreased but then steadily increased for individuals with type 2 diabetes; costs more than doubled from$564 in 1992 to $1,171 in 2001. For type 1 diabetes, prescription costs increased ∼32% from$851 in 1992 to $1,126 in 2001. Physician costs did not increase to the same extent as prescription costs during the 10 years after index. An initial spike in physician costs was observed in the 1st year after identification, but annual per capita physician costs grew relatively slowly from 1993 to 2001 (Table 2). The largest increase in costs during the 10 years after identification of diabetes was seen for dialysis. From 1991 to 1994, there were no costs for dialysis for individuals with type 1 diabetes. By 2001, however, the per capita dialysis cost for type 1 diabetes had reached$1,098, which was almost five times the per capita cost for type 2 diabetes ($227) in the same year (Table 2). After 1996, dialysis costs in the type 1 diabetes cohort by far exceeded those of the type 2 diabetes cohort. Overall, ∼31.5% of dialysis costs were for peritoneal dialysis and 68.5% were for hemodialysis. Total per capita costs were substantially higher in the year after index than in any other year (Fig. 1). Per capita costs for type 1 diabetes were relatively constant for 3–6 years after index but began to rise in the last 4 years of follow-up with the increased onset of dialysis. For type 2 diabetes, total per capita costs increased from a low of$3,880 3 years after index to $4,441 10 years thereafter. Over the 10 years of follow-up, ∼$137.1 million in health care resources were used by 3,625 individuals with diabetes. The average cost per individual with diabetes was $37,820. The difference in average individual 10-year cumulative costs between type 1 ($33, 684) and type 2 ($38,006) diabetes was not statistically significant (adjusted P = 0.45). The vast majority of the total expenditures ($131.8 million or 96% of costs) were for the provision of care to individuals with type 2 diabetes (Fig. 2).

We analyzed five categories of health care costs in an incident cohort with diabetes over an 11-year period, including the year before and the decade after identification of diabetes. Total health care costs increased markedly in the incident year but then stabilized at a lower level, with only minor increases over the ensuing decade. Trends in overall per capita costs were similar in type 1 and type 2 diabetes, with the exception of dialysis, where costs increased more dramatically for type 1 diabetes, resulting in slightly higher per capita costs in later years of our follow-up.

Interestingly, only a very small proportion (4%) of the total \$137 million spent over the decade after onset of diabetes was for individuals with type 1 diabetes. Type 2 diabetes accounted for >96% of total costs (i.e., the same proportion of their population distribution). As the number of type 2 diabetes cases will likely continue to outweigh the number of type 1 diabetes cases, type 2 diabetes is clearly the driver of the public health burden for diabetes (15).

We had anticipated that overall per capita health care costs for individuals with diabetes would increase over the 10 years after identification of diabetes, as a result of disease progression, accumulation of comorbidities, and deterioration of health status. This, however, was not the case. After a large increase in costs in the incident year, costs were relatively stable, despite increases in prescription, day surgery, and dialysis costs. Decreased hospital costs over the study period, which offset cost increases in the other categories, may reflect changes in hospitalization rates, length of stay, case mix, and service intensity (12,16,17). It is also possible that decreased hospital costs in the later years of follow-up reflect a survival bias in which healthier individuals survived to the end of the follow-up period.

Prescription medications were the second largest cost category after hospitalizations during most years of follow-up. Medication costs increased considerably from 1991 to 2001 for both type 1 and type 2 diabetes. This trend could be the result of 1) increased use of medications overall, 2) increased use of more expensive disease treatment options for cardiovascular risk reduction, and 3) use of newer, more expensive antidiabetic agents in the later years of follow-up (18). Increasing prescription costs should not be viewed entirely as problematic, however, given that proven effective treatments in this population are reportedly underused (19,20).

Other studies have explored costs before and after diagnosis of diabetes. Brown and colleagues, for example, estimated the incremental medical costs for people with type 2 diabetes for 8 years before (6) and after (5) diagnosis by using health records from one HMO in the U.S. In the estimation of costs of care after diagnosis, however, these authors used a cumulative incidence cohort approach rather than a closed incident cohort identified in a single year because of the relatively small sample size available in the HMO. This limited their evaluation of cumulative costs and reduced confidence in the costs in the longer term after diagnosis. Regardless of differences between studies and between Canadian and U.S. health care systems, results of both studies indicate relatively large increases in total health care costs in the year that diabetes is diagnosed (attributable to increased hospital expenditures) and substantial increases in prescription expenditures in the years after the diagnosis (5).

Several limitations should be noted when interpreting these data. First, although there is good evidence of validity of the NDSS criteria for identifying diabetes cases, some cases may have been missed (21). In addition, we could not examine the Registered Indian population because prescription data were not available. Because ∼10% of Saskatchewan's population with diabetes are Registered Indians, it is unlikely that overall trends would be affected drastically by inclusion of their data, although the relative proportion of type 2 diabetes and costs would be even more than 96%.

Resource use and cost comparisons between cohorts in this study were limited to five categories of direct costs. We were unable to capture resource use and costs for diabetes educators, dietitians, and podiatrists, as well as the auxiliary costs of transplants and laboratory services after 1993, because these are managed under regional health budgets. Before 1993, funding of laboratory services was reflected in fee-for-service physician billings, but these were no longer included after this year. We were also unable to capture other expenditures such as emergency department services or directly capture dialysis costs, although we are confident that the prospective costing data we used to estimate the latter is both valid and accurate (14). It should also be noted that day surgery and hospital costs were not obtained directly from Saskatchewan Health records but were estimated by applying costs from external sources to records obtained directly from Saskatchewan Health. Furthermore, the same estimation approach was used for all subjects, so there was no bias in the comparison between type 1 and type 2 diabetic patients.

The algorithm used in this study to distinguish between type 1 and type 2 diabetes has not been previously validated. Using our algorithm, it is possible that some individuals with late-onset type 1 diabetes would be misclassified as having type 2 diabetes. We chose not to use age at index as a criterion for classification and depended only on treatment patterns. The proportion of incident cases categorized as type 1 diabetes was, however, similar to the proportion observed in U.S. population-based data from 1990 to 1992 (22,23), where ∼4.9% of incident cases were categorized as type 1 diabetes. Although some potential for misclassification existed, we believe it would be trivial.

Last, the time period of data used to eliminate prevalent cases of diabetes from the analysis was relatively short. It is conceivable that some prevalent diabetes cases would appear to be incident in 1992 because they may not have met the NDSS criteria in 1989 or 1990. This would explain, in part, the older-than-expected average age at index of individuals with type 1 diabetes and why two case subjects with type 2 diabetes received dialysis before identification of the disease. Further, as a chronic medical condition, even 10 years of follow-up might be considered a limited time horizon for the estimation of costs of care for diabetes. Nevertheless, to our knowledge, this study represents the largest cohort with the longest follow-up of health care use and costs published to date.

Information on the relative burden of health conditions, in epidemiologic or economic terms, is of interest to health policymakers when making decisions regarding the allocation of resources for the management of prevalent and expected cases and to justify health promotion and disease prevention programs. It has also been suggested that public health should be the first consideration in allocation of research funding (24). If this were the case, research funding for type 2 diabetes should far outweigh funding available for type 1 diabetes.

In summary, although we had anticipated that per capita costs associated with caring for individuals with diabetes would increase as the disease progressed over the decade after diagnosis, decreasing hospitalization costs kept total per capita costs relatively stable. Total health care expenditures for diabetes are driven by the much larger prevalence of type 2 compared with type 1 diabetes and highlight the relative public health burden of both conditions.

Figure 1—

Per capita overall costs according to type of diabetes, for case subjects who remained active in each year of the study. Values in table are mean ± SD.

Figure 1—

Per capita overall costs according to type of diabetes, for case subjects who remained active in each year of the study. Values in table are mean ± SD.

Close modal
Figure 2—

Cumulative health care costs according to type of diabetes.

Figure 2—

Cumulative health care costs according to type of diabetes.

Close modal
Table 1—

Demographic characteristics of 1992 incident cohort (n = 3,625)

Type 1 diabetesType 2 diabetesTotal
n 156 (4.3) 3,469 (95.7) 3,625
Age (years) 49.5 ± 22.1 61.7 ± 16.2* 61.2 ± 16.7
Follow-up (years) 6.5 ± 3.6 7.7 ± 2.9* 7.7 ± 3.0
Sex (male) 92 (59.0) 1,854 (53.4) 1,946 (53.7)
Residence (rural) 76 (48.7) 1,619 (46.7) 1,695 (48.6)
Data source for NDSS case definition (hospital file) 24 (15.4) 402 (11.6) 426 (11.7)
Deaths, 1991–2001 48 (30.8) 1,066 (30.7) 1,114 (30.7)
Type 1 diabetesType 2 diabetesTotal
n 156 (4.3) 3,469 (95.7) 3,625
Age (years) 49.5 ± 22.1 61.7 ± 16.2* 61.2 ± 16.7
Follow-up (years) 6.5 ± 3.6 7.7 ± 2.9* 7.7 ± 3.0
Sex (male) 92 (59.0) 1,854 (53.4) 1,946 (53.7)
Residence (rural) 76 (48.7) 1,619 (46.7) 1,695 (48.6)
Data source for NDSS case definition (hospital file) 24 (15.4) 402 (11.6) 426 (11.7)
Deaths, 1991–2001 48 (30.8) 1,066 (30.7) 1,114 (30.7)

Data are mean ± SD or n (%). *P < 0.001.

Table 2—

Per capita costs according to resource category and type of diabetes

Prescriptions
Physicians
Hospitalizations
Type 1 diabetesType 2 diabetesTotalType 1 diabetesType 2 diabetesTotalType 1 diabetes
1991 586 ± 784 424 ± 640 431 ± 648 311 ± 502 491 ± 762 484 ± 753 738 ± 2,924
1992 851 ± 875 564 ± 709 577 ± 719 1,047 ± 1,392 867 ± 1,029 875 ± 1,047 5,212 ± 11,510
1993 632 ± 902 540 ± 657 544 ± 669 613 ± 831 649 ± 889 647 ± 886 3,739 ± 10,288
1994 582 ± 594 541 ± 660 543 ± 658 515 ± 1,019 580 ± 863 578 ± 869 2,308 ± 7,104
1995 696 ± 772 575 ± 685 579 ± 688 538 ± 865 580 ± 873 578 ± 873 2,104 ± 5,869
1996 637 ± 577 642 ± 751 642 ± 745 528 ± 1617 589 ± 822 586 ± 864 1,825 ± 6,765
1997 738 ± 756 753 ± 876 752 ± 871 568 ± 1,273 595 ± 827 594 ± 847 1,357 ± 4,347
1998 787 ± 906 823 ± 898 821 ± 898 578 ± 1,266 620 ± 912 618 ± 926 2,281 ± 12,200
1999 898 ± 1,089 922 ± 1,014 921 ± 1,017 639 ± 1,288 646 ± 910 646 ± 926 2,023 ± 7,417
2000 1,024 ± 1,256 1,039 ± 1,076 1,038 ± 1,083 702 ± 1,481 661 ± 1,014 662 ± 1,034 2,341 ± 7,045
2001 1,126 ± 1,297 1,171 ± 1,190 1,169 ± 1,194 598 ± 1,207 718 ± 1,045 714 ± 1,051 1,386 ± 4,972
1,648 ± 6,693 1,609 ± 6,578 61 ± 250 73 ± 277 72 ± 276 0 ± 0 14 ± 665 14 ± 651
5,788 ± 41,262 5,763 ± 40,435 58 ± 200 101 ± 398 99 ± 391 0 ± 0 23 ± 787 22 ± 770
2,767 ± 8,721 2,809 ± 8,795 87 ± 317 109 ± 398 108 ± 395 0 ± 0 46 ± 1,325 44 ± 1,296
2,547 ± 13,230 2,538 ± 13,050 107 ± 353 124 ± 441 124 ± 438 0 ± 0 87 ± 2,167 83 ± 2,126
2,835 ± 20,453 2,808 ± 20,100 188 ± 651 119 ± 407 121 ± 418 90 ± 989 74 ± 1,671 74 ± 1,650
2,864 ± 25,422 2,826 ± 24,986 144 ± 433 138 ± 457 138 ± 456 467 ± 4,959 44 ± 1,421 60 ± 1,687
3,066 ± 31,745 3,005 ± 31,183 104 ± 361 142 ± 361 140 ± 496 780 ± 5,816 40 ± 1,269 67 ± 1,665
2,351 ± 11,279 2,348 ± 11,310 125 ± 453 180 ± 624 178 ± 619 949 ± 6,533 52 ± 1,560 83 ± 1,960
2,001 ± 5,894 2,002 ± 5,952 127 ± 418 166 ± 525 164 ± 522 1,102 ± 5,905 93 ± 1,920 129 ± 2,189
2,279 ± 8,336 2,281 ± 8,293 136 ± 583 151 ± 464 150 ± 469 866 ± 8,127 159 ± 2,640 184 ± 3,005
2,163 ± 7,110 2,136 ± 7,048 142 ± 411 163 ± 552 162 ± 547 1,098 ± 9,944 227 ± 3,108 257 ± 3,566
Prescriptions
Physicians
Hospitalizations
Type 1 diabetesType 2 diabetesTotalType 1 diabetesType 2 diabetesTotalType 1 diabetes
1991 586 ± 784 424 ± 640 431 ± 648 311 ± 502 491 ± 762 484 ± 753 738 ± 2,924
1992 851 ± 875 564 ± 709 577 ± 719 1,047 ± 1,392 867 ± 1,029 875 ± 1,047 5,212 ± 11,510
1993 632 ± 902 540 ± 657 544 ± 669 613 ± 831 649 ± 889 647 ± 886 3,739 ± 10,288
1994 582 ± 594 541 ± 660 543 ± 658 515 ± 1,019 580 ± 863 578 ± 869 2,308 ± 7,104
1995 696 ± 772 575 ± 685 579 ± 688 538 ± 865 580 ± 873 578 ± 873 2,104 ± 5,869
1996 637 ± 577 642 ± 751 642 ± 745 528 ± 1617 589 ± 822 586 ± 864 1,825 ± 6,765
1997 738 ± 756 753 ± 876 752 ± 871 568 ± 1,273 595 ± 827 594 ± 847 1,357 ± 4,347
1998 787 ± 906 823 ± 898 821 ± 898 578 ± 1,266 620 ± 912 618 ± 926 2,281 ± 12,200
1999 898 ± 1,089 922 ± 1,014 921 ± 1,017 639 ± 1,288 646 ± 910 646 ± 926 2,023 ± 7,417
2000 1,024 ± 1,256 1,039 ± 1,076 1,038 ± 1,083 702 ± 1,481 661 ± 1,014 662 ± 1,034 2,341 ± 7,045
2001 1,126 ± 1,297 1,171 ± 1,190 1,169 ± 1,194 598 ± 1,207 718 ± 1,045 714 ± 1,051 1,386 ± 4,972
1,648 ± 6,693 1,609 ± 6,578 61 ± 250 73 ± 277 72 ± 276 0 ± 0 14 ± 665 14 ± 651
5,788 ± 41,262 5,763 ± 40,435 58 ± 200 101 ± 398 99 ± 391 0 ± 0 23 ± 787 22 ± 770
2,767 ± 8,721 2,809 ± 8,795 87 ± 317 109 ± 398 108 ± 395 0 ± 0 46 ± 1,325 44 ± 1,296
2,547 ± 13,230 2,538 ± 13,050 107 ± 353 124 ± 441 124 ± 438 0 ± 0 87 ± 2,167 83 ± 2,126
2,835 ± 20,453 2,808 ± 20,100 188 ± 651 119 ± 407 121 ± 418 90 ± 989 74 ± 1,671 74 ± 1,650
2,864 ± 25,422 2,826 ± 24,986 144 ± 433 138 ± 457 138 ± 456 467 ± 4,959 44 ± 1,421 60 ± 1,687
3,066 ± 31,745 3,005 ± 31,183 104 ± 361 142 ± 361 140 ± 496 780 ± 5,816 40 ± 1,269 67 ± 1,665
2,351 ± 11,279 2,348 ± 11,310 125 ± 453 180 ± 624 178 ± 619 949 ± 6,533 52 ± 1,560 83 ± 1,960
2,001 ± 5,894 2,002 ± 5,952 127 ± 418 166 ± 525 164 ± 522 1,102 ± 5,905 93 ± 1,920 129 ± 2,189
2,279 ± 8,336 2,281 ± 8,293 136 ± 583 151 ± 464 150 ± 469 866 ± 8,127 159 ± 2,640 184 ± 3,005
2,163 ± 7,110 2,136 ± 7,048 142 ± 411 163 ± 552 162 ± 547 1,098 ± 9,944 227 ± 3,108 257 ± 3,566

Data are mean ± SD. For case subjects who remained active in each year of the study.

J.A.J. is a Health Scholar with the Alberta Heritage Foundation for Medical Research (AHFMR) and is a Canada Research Chair in Diabetes Health Outcomes. He is also Chair of a New Emerging Team grant to the Alliance for Canadian Health Outcomes Research in Diabetes, sponsored by the Canadian Diabetes Association, the Heart and Stroke Foundation of Canada, The Kidney Foundation of Canada, and the Canadian Institutes of Health Research. S.R.M. is a Population Health Investigator of AHFMR and a New Investigator of the Canadian Institutes of Health Research.

We acknowledge Mary Rose Stang and Winanne Downey for their contribution to this project. This study is based in part on deidentified data provided by the Saskatchewan Department of Health. The interpretation and conclusions contained herein do not necessarily represent those of the Government of Saskatchewan or the Saskatchewan Department of Health.

1
Ettaro L, Songer TJ, Zhang P, Engelgau MM: Cost-of-illness studies in diabetes mellitus.
Pharmacoeconomics
22
:
149
–164,
2004
2
Hogan P, Dall T, Nikolov P: Economic costs of diabetes in the U.S. in
2002
.
Diabetes Care
26
:
917
–932, 2003
3
Dawson KG, Gomes D, Gerstein H, Blanchard JF, Kahler KH: The economic cost of diabetes in Canada,
1998
.
Diabetes Care
25
:
1303
–1307, 2002
4
Responding to the Challenge of Diabetes in Canada: First Report of the National Diabetes Surveillence System (NDSS)
2003
5
Brown JB, Nichols GA, Glauber HS, Bakst AW: Type 2 diabetes: incremental medical care costs during the first 8 years after diagnosis.
Diabetes Care
22
:
1116
–1124,
1999
6
Nichols GA, Glauber HS, Brown JB: Type 2 diabetes: incremental medical care costs during the 8 years preceding diagnosis.
Diabetes Care
23
:
1654
–1659,
2000
7
Olsen BS, Sjolie A, Hougaard P, Johannesen J, Borch-Johnsen K, Marinelli K, Thorsteinsson B, Pramming S, Mortensen HB: A 6-year nationwide cohort study of glycaemic control in young people with type 1 diabetes: risk markers for the development of retinopathy, nephropathy and neuropathy: Danish Study Group of Diabetes in Childhood.
J Diabetes Complications
14
:
295
–300,
2000
8
Bjork S: The cost of diabetes and diabetes care.
Diabetes Res Clin Pract
54(Suppl. 1)
:
S13
–S18,
2001
9
Caro JJ, Ward AJ, O'Brien JA: Lifetime costs of complications resulting from type 2 diabetes in the U.S.
Diabetes Care
25
:
476
–481,
2002
10
Simpson SH, Corabian P, Jacobs P, Johnson JA: The cost of major comorbidity in people with diabetes mellitus.
CMAJ
168
:
1661
–1667,
2003
11
Statistics Canada: Consumer price index (CPI), 2001 basket content, annual [article online],
2006
. Available from www.statscan.ca. Accessed 20 September 2006
12
Pohar SL, Simpson SH, Majumdar SR, Jacobs P, Ohinmaa A, Osei W, Stang M, Downey W, Johnson JA:
Epidemiologic and Cost Trends in Diabetes in Saskatchewan, 1991 to 2001
2006
(Institute of Health Economics Working Paper 05-06)
13
Johnson JA, Simpson SH, Jacobs P, Downey W, Beck P, Osei W:
Cost of Health Care for People With Diabetes Mellitus in Saskatchewan 1991 to 1996
. Edmonton, Alliance for Canadian Health Outcomes Research in Diabetes, 2002 (Institute of Health Economics Working Paper 02-02)
14
Lee H, Manns B, Taub K, Ghali WA, Dean S, Johnson D, Donaldson C: Cost analysis of ongoing care of patients with end-stage renal disease: the impact of dialysis modality and dialysis access.
Am J Kidney Dis
40
:
611
–622,
2002
15
Engelgau MM, Geiss LS, Saaddine JB, Boyle JP, Benjamin SM, Gregg EW, Tierney EF, Rios-Burrows N, Mokdad AH, Ford ES, Imperatore G, Narayan KM: The evolving diabetes burden in the United States.
Ann Intern Med
140
:
945
–950,
2004
16
Pohar SL, Majumdar SR, Jacobs P, Johnson JA:
Healthcare Utilization and Direct Healthcare Costs of Diabetes in Urban and Rural Saskatchewan, 1991–2001
2006
(Institute of Health Economics Working Paper WP 06-01)
17
Liu L, Hader J, Brossart B, White R, Lewis S: Impact of rural hospital closures in Saskatchewan, Canada.
Soc Sci Med
52
:
1793
–1804,
2001
18
Morgan SG, Bassett KL, Wright JM, Evans RG, Barer ML, Caetano PA, Black CD: “Breakthrough” drugs and growth in expenditure on prescription drugs in Canada.
BMJ
331
:
815
–816,
2005
19
Klinke JA, Johnson JA, Guirguis LM, Toth EL, Lee TK, Lewanczuk RZ, Majumdar SR: Underuse of aspirin in type 2 diabetes mellitus: prevalence and correlates of therapy in rural Canada.
Clin Ther
26
:
439
–446,
2004
20
Brown LC, Johnson JA, Majumdar SR, Tsuyuki RT, McAlister FA: Evidence of suboptimal management of cardiovascular risk in patients with type 2 diabetes mellitus and symptomatic atherosclerosis.
CMAJ
171
:
1189
–1192,
2004
21
Hux JE, Ivis F, Flintoft V, Bica A: Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm.
Diabetes Care
25
:
512
–516,
2002
22
LaPorte RE, Matsushima M, Chang YF: Prevalence and incidence of insulin-dependent diabetes [article online],
1995
. Available from http://diabetes.niddk.nih.gov/dm/pubs/america/pdf/chapter3.pdf. Accessed 13 December 2005
23
Kenny SJ, Aubert RE, Geiss LS: Prevalence and incidence of non-insulin dependent diabetes [article online],
1995
. Available from diabetes.niddk.nih.gov/dm/pubs/america/pdf/chapter4.pdf. Accessed 13 December 2005
24
Gross CP, Anderson GF, Powe NR: The relation between funding by the National Institutes of Health and the burden of disease.
N Engl J Med
340
:
1881
–1887,
1999

A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.