Considerable caution is required when considering estimates of healthcare costs and comparisons between them can only validly be made when they have been assembled using the same methods and the same assumptions. As with all cost of illness studies, the methods used are either 'top down' or 'bottom up'.
In the 'top down' approach, aggregate data at national or local level of treatments, healthcare utilization etc are used together with unit costs for the appropriate item. In the 'bottom up' approach, affected individuals are identified, their treatments and healthcare utilization events recorded and unit costs used to calculate the totals required. Each method has its own advantages and disadvantages.
Africa
No recent country specific estimates in Africa have been published since 19921Chale SS, Swai AB, Mujinja PG, McLarty DG. Must diabetes be a fatal disease in Africa? Study of costs of treatment. BMJ 1992; 304:1215-1218.. Cost of illness data from African countries is badly needed.
Eastern Mediterranean and Middle East
Data are also extremely sparse for the countries in the Eastern Mediterranean and Middle East Region. There appears to be nothing published since the study in Egypt2Arab M. Diabetes mellitus in Egypt. World Health Statistics Quarterly 1992; 45:334-337. and Tunisia3Rekik M, Abid M, Hachicha J, Abbes R, Moujahed M, Jarraya A. Direct cost of the ambulatory management of diabetes at the outpatient clinic of the National Social Security Fund of Sfax. Bulletin of the World Health Organization 1994;72:611-614.. It is worth noting that, in the Tunisian study, the total annual cost of medication and outpatient care for people with diabetes was 2.6 times that for people without diabetes (US$179 versus US$68) and that a clear relationship was found between higher costs and the presence of 'degenerative complications'.
Europe
Diabetes healthcare costs in Sweden have been extensively studied for many years. A study4Norlund A, Apelqvist J, Bitzen PO, Nyberg P, Schersten B. Cost of illness of adult diabetes mellitus underestimated if comorbidity is not considered. J Intern Med 2001; 250:57-65. estimated that 28% of the extra cost of diabetes is attributable to the costs of healthcare (the remainder are indirect costs). This amounted to an estimated SEK9,548 (US$1,278) per person per year.
Another study5Björk S. The cost of diabetes and diabetes care. Diabetes Res Clin Pract 2001; 54 Suppl 1:S13-S18. estimated that three times the healthcare resources are being spent on diabetes complications compared with that spent on diabetes control while yet another study6Jönsson P, Marke LA, Nystrom L, Wall S, Ostman J. Excess costs of medical care 1 and 8 years after diagnosis of diabetes: estimates from young and middle-aged incidence cohorts in Sweden. Diabetes Res Clin Pract 2000; 50 Suppl 1:35-47. made the important observation that excess costs in the first year after the diagnosis of diabetes in young adults (15-34 years at diagnosis) are considerably greater than those incurred seven years later ie eight years after diagnosis. Annual excess costs at these two time points for men were US$4,743 and US$2,010 respectively while, for women, the equivalent figures were US$4,976 and US$2,734. The cost profile during the natural history of the condition in any one person with diabetes seems, therefore, to be 'U' or 'J' shaped, with, immediately after diagnosis, relatively high costs which subsequently fall and then rise again with the onset of complications.
From France7Detournay B, Fagnani F, Phillippo M, Pribil C, Charles MA, Sermet C, Basdevant A, Eschwege E. Obesity morbidity and healthcare costs in France: an analysis of the 1991-1992 Medical Care Household Survey. Int J Obes Relat Metab Disord 2000; 24:151-155. comes the observation that medical care costs for people with diabetes are around EUR3,048 (US$3,726) per person per year (twice the average medical care consumption in the French population) while in the Netherlands8van Os N, Niessen LW, Koopmanschap MA, van der Lei J. Detailed analysis of the societal costs of diabetes mellitus. Ned Tijdschr Geneeskd 2000; 29:842-846., diabetes healthcare costs have been estimated to be a modest 2.5% of the total healthcare budget.
Using a number of complementary data sources, several cost estimates for diabetes and its complications in the same defined United Kingdom (UK) population have been published9Currie CJ, Kraus D, Morgan CL, Gill L, Stott NCH, Peters JR. NHS Acute Sector Expenditure for Diabetes: the Present, Future, and Excess In-patient Cost of Care. Diabet Med 1997; 14:686-692.. In 1997 the authors published one of the few estimates of future costs of diabetes. More recently, they have been involved with others in developing a computer-based model for testing the effects of demographic, epidemiological and therapeutic changes on these future costs10Bagust A, Hopkinson PK, Maier W, Currie CJ. An economic model of the long-term healthcare burden of Type II diabetes. Diabetologia 2001; 44:2140-2155., thus linking descriptive estimates of the cost of diabetes with cost-effectiveness information from studies such as the UK Prospective Diabetes Study (UKPDS).
The findings of the CODE-2 (Cost of Diabetes in Europe - Type 2) and T2ARDIS (Type 2 Diabetes Accounting for a Major Resource Demand In Society) studies agree that the financial burden of type 2 diabetes in the United Kingdom is just under 5% of the nation's healthcare budget in 1998, that there is a strong relationship between hospital costs and the presence of complications, and that the economic and psychosocial burden of diabetes extends not only to affected individuals but also to their carers11Williams R, Gillam S, Murphy M, Holmes J, Pringle M, Bootle S, Bottomley J, Baxter H, Chandler F. The True Costs of Type 2 Diabetes in the UK - Findings from T2ARDIS and CODE-2 UK. Monograph of studies supported by GlaxoSmithKline. GlaxoSmithKline UK, Uxbridge, 2002..
North America
The most recent peer-reviewed estimate of the annual direct cost of diabetes in the USA is the American Diabetes Associations 2002 estimate of US$91.8 billion12American Diabetes Association. Economic Costs of Diabetes in the U.S. in 2002. Diabetes Care 2003; 26:917-932.. This compares with the previous estimate for 1997 of US$44 billion13American Diabetes Association. Economic Consequences of Diabetes Mellitus in the U.S. in 1997. Diabetes Care 1998; 21:296-309.. This is a total figure covering all aspects of diabetes and its complications.
The US literature also has a number of estimates which break down the total cost figure into estimates for specific, individual complications. For example, one study14OBrien JA, Shomphe LA, Kavanagh PL, Raggio G, Caro JJ. Direct Medical Costs of Complications Resulting from Type 2 Diabetes in the US. Diabetes Care 1998;21:1122-1128. estimated the 'event cost' of a single myocardial infarction in a person with diabetes to be US$27,630 (1996 prices). This leads on to a 'state cost' (ie the annual additional cost of care following such an event) of US$2,185 per annum. They point out that some early complications, such as the presence of microalbuminuria (the presence of small traces of protein in the urine) are low cost, estimated as a US$14 per annum 'state cost', but, if left undetected and untreated, lead to extremely high later costs, in this case the US$53,660 per annum 'state cost' of end-stage renal failure. In a similar fashion, for people with diabetes receiving their healthcare from one health maintenance organization (HMO), a major cardiovascular disease event eg myocardial infarction was estimated to increase the cost of care by 360%15Brown JB, Pedula KL, Bakst AW. The Progressive Cost of Complications in Type 2 Diabetes Mellitus. Arch Intern Med 1999; 159:1873-1880.. End-stage renal failure had the effect of increasing costs by 771%.
Recent work in Canada suggested that the total direct healthcare cost of diabetes was US$3.5 billion in 1998 prices16Dawson KD, Gomes D, Gerstein H, Blanchard JF, Kahler KH. The Economic Cost of Diabetes in Canada, 1998. Diabetes Care 2002; 25:1303-1307.. Comparison with the calculated estimates in the table (see Data tables - North American Region) of 4.7 billion international dollars (R=2) and 8.7 billion international dollars (R=3) suggests that this may be an underestimate (or the calculated figures an overestimate). The empirical Canadian work emphasized the dominant contribution of cardiovascular disease in people with diabetes (35% of direct and indirect costs) and declared that "the cost of preventive treatment is insignificant compared with the downstream costs of failure to adequately treat the disease".
South and Central America
The direct healthcare costs were estimated as US$10.69 billion for 25 countries in Latin America and the Caribbean area in 2000 (see Data table)17Barceló A, Aedo C, Rajpathak S, Robles S. The cost of diabetes in Latin America and the Caribbean. Bulletin of the World Health Organization: the International Journal of Public Health 2003; 81(1):19-27.. The contributions of the various items to the direct cost total are shown in the figure below, while the specific contributors to the cost of diabetic complications are shown in the next figure. Indirect costs were also calculated and were around five times the direct cost total.
The method employed in this study is the 'top down' method. The authors used population prevalence estimates for diabetes, treatment and healthcare utilization figures taken both from individual studies and routinely collected data. These were then multiplied by unit costs for insulin, oral hypoglycaemics and other items. Like all top-down studies, the accuracy of the results is dependent on the validity of the assumptions made. Nevertheless, this work represents a considerable advance in pulling together comparable data from Latin American and Caribbean countries not previously studied (see Map).
South-East Asia
In India, estimates have been made of the out-of-pocket expenses resulting from diabetes. This is the direct healthcare impact on the person with diabetes who may need to spend a considerable proportion of the family income in order to receive treatment at a chosen location eg outside the public sector. In a diabetes centre in Chenai, families in the poorest section of the population may have to spend as much as 25% of their income in order to obtain the care of their choice18Shobhana R, Rao PR , Lavanya A, Williams R, Vijay V, Ramachandran A. Expenditure on healthcare incurred by diabetic subjects in a developing country - a study from southern India. Diabetes Res Clin Pract 2000; 48:37-42..
Western Pacific
A similar theme, of the personal direct costs, was developed in a study19Simmons D, Peng A, Cecil A, Gatland B. The personal costs of diabetes: a significant barrier to care in South Auckland. N Z Med J 1999; 112:383-385. for patients resident in an inner suburb of Auckland, New Zealand. Not only were these patients spending a significant proportion of their income on diabetes care but between a fifth and a half (depending on income and ethnic origin) of those taking part reported that personal costs had an inhibitory effect on self-monitoring of blood glucose, self medication and even on insulin therapy amongst those who needed it.
Taiwan has been estimated to spend 11.5% of its healthcare budget on the treatment of people with diabetes. This comes from a national extrapolation of Bureau of National Health Insurance claims between July 1997 and July 199820Lin T, Chou P, Lai M, Tsai S, Tai T. Direct cost-of-illness of patients with diabetes mellitus in Taiwan. Diabetes Res Clin Pract 2001; 54 Suppl 1:43-46.. The average cost of care for someone with diabetes in this system was 4.3 times higher than that for a person without diabetes.