Diabetes on a cardiovascular ward: adherence to current recommendations.

Objectives: Improving diabetes and blood pressure control decreases the incidence and progression of microvascular disease. Likewise, screening for microvascular complications is beneficial in the early detection and treatment of these disorders. However, adherence to practice guidelines for screening and treatment in patients with diabetes is suboptimal. This study describes a group of patients with diabetes who were admitted to a cardiology service at an academic medical center.

Methods: Patient interview and chart review were used to determine glycemic control and compliance with practice guidelines.

Results: The mean hemoglobin A1c was 8.3%. Only 69% of patients received ophthalmologic examinations, and fewer were screened for nephropathy. Thirty-five percent of patients monitored home blood glucoses less than daily. Nearly 17% had no hemoglobin A1c or lipid checks during the 3 months before admission.

Conclusions: For a group of poorly controlled patients with diabetes who are at high risk for cardiovascular disease, adherence to practice guidelines and the level of diabetes control is inadequate.

Key Words: diabetes mellitus, hemoglobin A1c, standards of care 

**********

Diabetes affects approximately 18.2 million people in the United States and accounts for $92 billion in direct medical costs annually. An additional $40 billion dollars is accrued in indirect medical costs, including disability, work loss, and premature mortality for patients with diabetes mellitus. (1) Many of the chronic complications of diabetes are preventable. Data from the United Kingdom Prospective Diabetes Study (UKPDS) and the Diabetes Control and Complications Trial have clearly demonstrated that improved glucose control and aggressive treatment of blood pressure decrease the incidence and progression of microvascular disease in patients with diabetes. (2-5) Likewise, screening for diabetic retinopathy, nephropathy, and neuropathy have been shown to be beneficial in the early detection and treatment of these disorders. (6)

Although practice guidelines exist for the care of patients with diabetes, numerous studies have demonstrated that adherence to these guidelines is low. (7,8) The goal of this study is to determine the level of adherence to practice guidelines among patients with diabetes who are admitted to a cardiovascular inpatient service at a major academic medical center. Presumably, these patients will be among those at highest risk for both micro- and macrovascular complications of diabetes, as many already have established vascular disease.

Materials and Methods

Study Subjects

Patients hospitalized on the inpatient cardiology service at Duke University Medical Center were eligible to participate in the study. This service is a high-volume cardiology service with more than 8,000 admissions and 5,000 procedures annually. The daily roster of the inpatient service was reviewed and patients were recruited from March 17, 2003, to April 21, 2003, excluding weekends and holidays. All patients with an existing or newly established diagnosis of diabetes were approached for consent to participate; 613 patients were screened, of whom 185 had diabetes. Forty-nine of the patients with diabetes were not approached for consent because they were discharged from the hospital before consent could be obtained. Of the remaining 136 patients, 28 (21%) refused consent to participate and 8 (6%) were not competent to provide informed consent. Of the remaining 100 patients, one patient was admitted twice during this time period. Since each admission is an opportunity for intervention with respect to guideline adherence, both time points were retained for analysis. Therefore, the data set analyzed in this study includes 100 patients, representing 101 hospital admissions.

Data Collection

After informed consent was obtained, patients were interviewed and medical charts were reviewed to obtain demographic data, including age, height, weight, ethnicity, sex, level of education, and duration of diabetes. Patients were also asked to identify the health care provider responsible for their diabetes care and to indicate whether they performed home glucose monitoring. Patients were asked whether they had a personal history of, or had received any therapy for diabetic retinopathy, nephropathy, or neuropathy. Patient report and chart review were used to determine if the patient had received a dilated eye examination within 1 year, if the patient had any painful paresthesias in the extremities or had a history of foot ulcer or amputation, and if the patient had any documented renal disease. Although data documenting clinical symptoms of heart failure were not collected, the ejection fraction (EF) was collected when available. For the purposes of data analysis, an EF <35% was used as a surrogate marker of congestive heart failure. Finally, patient report and chart review were used to compile a list of current medications both at admission and at discharge from the hospital.

Statistical Analysis

Descriptive statistics including mean, standard deviation (SD), median, interquartile range (IQR), frequencies, proportions, and graphical displays were computed with the use of SAS Enterprise Guide software (Version 2.05.89; SAS Institute Inc, Cary, NC). All variables were examined by visual inspection of each variable's histogram to determine if parametric distributional assumptions were valid. The mean and SD are reported for all continuous variables. When parametric distributional assumptions were not valid, median and IQR were also reported for comparison. Frequencies of missing variables have been noted for all reported variables. Missing data were not included in the analysis. Where appropriate, 95% confidence intervals were calculated for continuous variables as an additional measure of the spread of the data.

Results

Demographics

Of the 613 patients originally screened for study eligibility, 185 (30%) had diabetes; 100 consecutive patients with diabetes who gave consent were included in the analysis. For the patient with two admissions during the study period, data were collected at each admission, resulting in n = 101 for the data set. Table 1 shows demographic characteristics of this patient population. Ninety patients (89%) had type 2 diabetes, whereas eight patients (8%) had type 1 diabetes, as determined by self-report and chart review. The type of diabetes could not be determined in three patients (3%). Eighty-five patients had a family history of diabetes (84%). Fifteen patients (15%) had no family history of diabetes, whereas one patient was adopted and did not know if he had a family history of diabetes. The mean duration of diabetes was 11 [+ or -] 9 years (median, 10; IQR, 12; 95% CI: 9.4, 13.0). Review of the histogram for this variable demonstrated a nonparametric distribution. Therefore, a subgroup analysis of those patients requiring insulin was performed. Among patients taking insulin, the mean duration was 15.8 [+ or -] 9.4 years (95% CI: 13.1, 18.5). Among those on no insulin, the mean duration was 6.7 [+ or -] 5.9 years, 95% CI: 6.9, 8.1). Among patients with an EF documented during this hospitalization (n = 79), the mean EF was 44 [+ or -] 17% (95% CI: 40, 48).

Microvascular Complications

Retinopathy. Seventy patients (69%) had had an eye examination within the last year. Among those who had an eye examination, 48 (69%) had no history of retinopathy. Twenty-two patients (31%) had retinopathy; 16 (23%) had required laser therapy in the past.

Neuropathy. Forty-five patients reported symptoms of neuropathy. Only 15 patients had received any pharmacologic treatment for neuropathy (drugs not specified).

Twenty-eight patients (28%) reported a history of renal disease related to their diabetes. Sixty-seven patients (66%) reported that they had no nephropathy, and six patients (6%) were unsure. Thirty-one patients (31%) carried the diagnosis of chronic renal insufficiency, and one patient had end-stage renal disease. However, on review of the medical record, 52 patients (51%) had no assessment of renal function other than serum creatinine.

Of the remaining 49 patients, 43 (88%) had a random urinalysis performed during hospitalization, providing a gross assessment of proteinuria. Thirty-one patients (63%) had no proteinuria on urinalysis, 5 (10%) had 1+ proteinuria, 1 (2%) had 2+ proteinuria, and six (12%) had 3+ proteinuria.

Only six patients (6%) had a documented microalbumin level within 3 months of hospitalization. Of those six patients, four demonstrated microalbuminuria. Only 20 patients (20%) could be definitively determined to have no overt clinical evidence of nephropathy, as defined by either a urinalysis without protein or no microalbuminuria.

Medications

The admission and discharge medication regimens for patients in this study are described in Table 2. At admission, 83 patients were taking aspirin, 71 were taking a [beta]-blocker, 67 were taking an angiotensin-converting enzyme inhibitor (ACEI), and 63 were taking a hydroxymethylglutaryl CoA reductase inhibitor (statin). Five patients were receiving glyburide and 50 were taking insulin. Among those taking insulin, the mean daily dose was 64 U (range, 12 to 170, SD [+ or -] 33, 95% CI: 55, 74). At discharge, 93 patients were taking an aspirin, 81 were taking a [beta]-blocker, 74 were taking an ACEI, and 68 were taking a statin. Only one patient had glyburide discontinued during hospitalization, and one patient was initiated on insulin therapy.

Drug Interactions

Among those patients with a current EF available (n = 79), six patients were taking a thiazolidinedione and 15 were on metformin. Two patients with an EF <35% were receiving thiazolidinedione and three were receiving metformin.

Of the 15 patients treated with metformin, none had a creatinine level >1.5 mg/dL.

Outpatient Care

Physician. Eight patients reported having no physician taking care of their diabetes. Table 3 shows the remaining distribution of care givers.

Blood Glucose Monitoring. Eighteen patients (18%) reported that they do not check home blood glucose readings. Forty-four (44%) check more than once daily, 22 (22%) check once daily, and 17 patients (17%) check only occasionally (less than once daily).

Laboratory Monitoring. Mean and 95% confidence interval, as well as median and IQR where appropriate, for collected laboratory data are shown in Table 4.

Hemoglobin A1c. Only 82 patients had a hemoglobin A1c (HbA1c) value documented within the 3 months before admission. The mean HbA1c was 8.3% (SD [+ or -] 2.0). However, the median HbA1c of 7.5% (IQR, 2.9) indicates some degree of right skew for this variable. Consequently, a subgroup analysis of those patients requiring insulin was done. Of the 51 patients not taking insulin, the mean HbA1c was 7.5% (range, 5.2 to 14; SD [+ or -] 1.8; 95% CI: 6.9, 8.0). Among the 50 patients taking insulin, the mean HbA1c was 9.0% (range, 6.1 to 12.8; SD [+ or -] 1.9; 95% CI: 8.4, 9.6).

Lipids. Only 83 patients had an assessment of lipid status documented within 3 months before admission. Of these, 10 patients had triglycerides greater than 400; therefore, low-density lipoprotein (LDL) values were not calculated. One other patient had only total cholesterol and high-density lipoprotein (HDL) values obtained. The mean lipid values for patients in this study were total cholesterol, 183 [+ or -] 79 mg/dL; triglycerides, 230 [+ or -] 402 mg/dL; HDL, 44 [+ or -] 15 mg/dL; and LDL 98 [+ or -] 44 mg/dL. Because values for total cholesterol, triglycerides, and LDL were not normally distributed, a subgroup analysis of patients requiring insulin and of patients receiving lipid-lowering medications was performed (data not shown). In both subgroup analyses, 95% confidence intervals for all lipid variables overlapped considerably and did not demonstrate a significant difference in these variables, based on group.

Renal. All patients had a documented serum creatinine during hospitalization. Mean serum creatinine was 1.4 [+ or -] 0.94 mg/dL. Data concerning proteinuria is documented above.

Discussion

The care of patients with diabetes is an ongoing endeavor, requiring continuing medical care, patient education, and a multidisciplinary approach. To assist in managing the complexities of diabetes care, the American Diabetes Association (ADA) has published guidelines for the standards of care for patients with diabetes, including routine home glucose monitoring, at least yearly screening for nephropathy, retinopathy, and neuropathy, HbA1c monitoring every 3 to 6 months, and at least annual screening for hyperlipidemia. (6) Adherence to these guidelines helps to reduce both the morbidity and mortality associated with diabetes.

The patients in this sample represent a group fairly typical of patients with diabetes in the United States. According to the ADA, 18.3% of people in the United States over age 60 have diabetes. In our population, 30% had diabetes. This may represent an increased burden of diabetes among patients admitted to the cardiology service of a tertiary care medical center and may also reflect the increased prevalence of diabetes in the area of the Southeast covered by Duke University's referral base. (9) The patients with diabetes are also fairly typical in that they represent an obese population (mean body mass index = 33) with many features of the metabolic syndrome (coexisting hypertension and hyperlipidemia).

Most of these patients could identify a medical professional responsible for their diabetes care, but 8% reported having no health care giver to help them with their diabetes. The preponderance of patients was cared for by generalists, with only 12% seeing an endocrinologist. This distribution is typical or slightly higher than for patients in the United States. (10) Unfortunately, many patients were not receiving routine health maintenance examinations designed to detect and treat the microvascular complications at an early stage of progression. For example, only 69% of patients in this study had an annual eye examination, and over 30% of those who had an examination reported having retinopathy. Yearly retinopathy screening has been demonstrated to be cost-effective from both medical and economic perspectives. (11) Early detection of retinopathy and treatment can prevent adverse outcomes. (12-14) Because diabetes is the leading cause of blindness in the United States, increased adherence to screening guidelines has the potential to reduce the morbidity associated with diabetes and improve quality of life.

The study's questions regarding nephropathy seem to indicate that this was the microvascular complication about which patients were least educated. Although 28 patients knew they had renal disease, 31 carried a diagnosis of chronic renal insufficiency, and 12 had documented proteinuria. More than half of the patients had no documented assessment of renal function other than a serum creatinine, and very few had a documented microalbumin level within 3 months before hospitalization. Not only is this of concern from a health-maintenance standpoint, but microalbuminuria has been shown to be an independent risk factor for cardiovascular events and all-cause death in patients with and without diabetes. (15) Therefore, it is particularly important to know whether microalbuminuria is present in this patient population.

Home blood glucose monitoring is the cornerstone of diabetes care. The ADA recommends that home blood glucose monitoring be included in the treatment plan for all patients treated with diabetes, but particularly in those patients in poor control and/or treated with insulin. (6) Only in patients with adequate blood glucose control treated with diet alone has the role of blood glucose monitoring not been defined. However, in our study population of poorly controlled patients with diabetes at high risk for cardiovascular complications, 18% of patients did not check home blood glucose at all, and 17% only monitored occasionally. Likewise, almost 20% of patients (n = 19) had neither a HbA1c nor a lipid profile documented within 3 months before admission or during the time of their hospitalization.

Hospitalization for a cardiovascular event is an ideal time for reevaluation of the treatment regimen. Most patients in this sample were receiving typical cardiovascular medications at admission, including aspirin, [beta]-blocker, ACEI, and a statin. Not surprisingly, rates of aspirin, [beta]-blocker, and ACEI use were increased at discharge. However, even at discharge, only 74% of patients were taking an ACEI and aspirin. Failure to provide both aspirin and ACEI for prophylaxis for cardiovascular and renal disease in patients with diabetes could be considered a medical error by the definitions set forth by the Institute of Medicine. (16) Likewise, although 85 patients had hyperlipidemia, only 63 patients were taking a statin at admission. This number was not substantially increased at discharge (n = 68). Given the demonstrable benefit of these agents, (17) it seems surprising that they are not more often used. Although we do not have data for all contraindications for the use of these medications, it is unlikely that all of the remaining patients were not candidates for statin, ACEI, and aspirin therapy.

The control of diabetes can also be reassessed at the time of hospitalization. The standard method for determining overall glucose control is by measuring HbA1c. In this population, only 82 of 100 patients had their HbA1c measured, either during the course of their hospitalization, or within 3 months prior. Of those patients who had a measured HbA1c, the mean was 8.3%, demonstrating an unsatisfactory level of control. Among those patients taking insulin, the duration of diabetes was substantially longer (mean, 15.8 years; 95% CI: 13.1, 18.4, compared with mean, 6.7 years; 95% CI: 5.1, 8.4 for those not taking insulin) and HbA1c markedly higher (mean, 9.0; 95% CI: 8.4, 9.6 compared with mean, 7.5; 95% CI: 6.9, 8.1). Neither group demonstrated adequate glucose control. Despite this fact, diabetes medications were not changed appreciably from admission to discharge. Furthermore, the increased HbA1c and duration of diabetes in the insulin group highlights the current practice pattern of using insulin as a late addition to the diabetes regimen when, presumably, all other options have failed. (18)

Because of the numerous potential adverse effects of many oral antidiabetic agents in patients with cardiac and renal disease, continuous review of the medical regimen is warranted. In this study, 66% of patients were taking an oral agent at admission. Among the patients receiving oral agents, 51% received a sulfonylurea, 22% received metformin, 21% received a thiazolidinedione, 4% received a meglitinide, and 1% received acarbose. Although there are few current data on the distribution of oral agent use in the United States, an analysis of Merck-Medco Managed Care Information data from October 1997 to March 1999 showed that 66% of initial oral agent therapy was sulfonylurea, 24% metformin, 7% thiazolidinedione, 2% meglitinide, and 1% [alpha]-glucosidase inhibitor. (19) This probably underestimates the use of thiazolidinediones, because currently available agents were not available. Therefore, the distribution of oral agents in this study is similar to known data for the US population.

Previously reported data indicate continued use of metformin in a substantial number of patients who had renal insufficiency (67%) and use of metformin and thiazolidinediones in patients with congestive heart failure (4.4% and 16.1%, respectively). (20,21) In our patient sample, use of these medications in patients with established contraindications was limited. No patients with a creatinine level greater than 1.5 mg/dL were discharged on metformin. Likewise, of the patients with a known EF of less than 35%, only two were discharged on a thiazolidinedione, and three were discharged on metformin. Whether these incidents represent true contraindications is not known because data on clinical heart failure were not collected.

A recent examination of the use of oral antidiabetic agents in the United States determined that sulfonylureas represent 35% of the market share for patients with diabetes. (22) Similarly, more than one third of our patient population (34%) was taking a sulfonylurea at admission. Although concern about the safety of sulfonylurea therapy in patients with both diabetes and cardiovascular disease was raised by the University Group Diabetes Program study, other large scale studies, including the UKPDS, have not supported the notion that sulfonylurea therapy entails an excess risk of cardiovascular mortality. (23,24) However, there remains some concern that glyburide, currently the most widely used sulfonylurea in the United States, may have deleterious effects on ischemic preconditioning, a phenomenon allowing the myocardium to become conditioned to transitory ischemia and providing a protective effect. (25,26) Therefore, many experts in the field advocate that glyburide be avoided entirely, particularly in patients with cardiovascular disease. (27) In our patient population, 15% (n = 5) of patients receiving sulfonylurea therapy were taking glyburide. Only one of those patients had the drug discontinued at discharge. Twenty-four percent (n = 8) of patients taking sulfonylurea were receiving glimepiride, a sulfonylurea that has been shown not to have any effect on ischemic preconditioning, (28) and 62% (n = 21) were taking glipizide, the effects of which on ischemic preconditioning have not been studied.

Conclusion

This study is limited by a number of factors. A referral bias is almost certainly present, since Duke University Medical Center is a tertiary care hospital. However, the demographics and characteristics of our sample are not dissimilar to those known for the general population with diabetes in the United States. It is likely that our population exhibits an admission rate bias, whereby patients with diabetes are more likely than those without diabetes to be admitted to the hospital for the evaluation of chest pain. Likewise, it seems plausible that those patients admitted may have more severe disease, both diabetic and cardiovascular, than patients treated on an outpatient basis. There may also be a form of nonresponder bias introduced by the number of patients who were unavailable for consent because they had already been discharged. The direction of this bias is not completely clear; these patients may have had a shorter hospital stay, implying less severe illness, or they may have been receiving more inpatient evaluations, causing their absence from the ward. Thus, they may have been more severely ill than those patients recruited into the study. Finally, there may be a limited component of recall bias. Not all patients in this sample had received previous care at Duke University Hospital. For those patients, the study was dependent on chart records from another institution sent with the patient and the patient's own recollection of which screening tests and diagnoses had been documented. Wherever possible, patient recall was confirmed by review of data from our hospital and from outside institutions. However, such data were not always obtainable.

Despite these limitations, this sample is likely to be representative of the level of diabetes care at tertiary care institutions and, based on its similarities to larger samples of patients with diabetes, is probably generalizable to other populations. These data demonstrate that even in patients at highest risk for adverse coronary events, adherence to guidelines for standards of diabetes care is inadequate. The cause of this deficiency remains unclear. Possibilities include lack of information regarding the guidelines, or inability to achieve guideline goals due to the complexities of medical care in this population. Recent recognition of the importance of glucose control in hospitalized patients has prompted many institutions to develop protocols for the acute treatment of diabetes. (29,30) Several studies have demonstrated that multidisciplinary teams, including endocrinologists, diabetes educators, nurse practitioners, and nutritionists, can be effective in improving both the cost of care and patient satisfaction with treatment. (31) Recent recommendations have also called for the creation of hospital-wide committees to form institutional guidelines based on national standards for the treatment of patients with diabetes. (32) Further research is necessary to determine whether improved adherence to diabetes treatment guidelines will have a significant impact on long-term patient outcomes.

The nice thing about being a celebrity is that if you bore people they
think it's their fault.
--Henry Kissinger

Table 1. Patient demographic characteristics (a)

                           Mean (SD)             Range

Age (yr)                   61 ([+ or -]12)       32-84
BMI (kg/[m.sup.2])         33 ([+ or -]7.74)     18-63
Duration of DM (yr)        Mean: 11 ([+ or -]9)   0-43
                           Median: 10 (IQR: 12)
Race
  White                    56% (n = 57)
  Black                    33% (n = 33)
  Native American           8% (n = 8)
  Other                     3% (n = 3)
Sex
  Male                     57% (n = 58)
  Female                   43% (n = 43)
Level of education
  Less than high school    44% (n = 44)
  High school              40% (n = 40)
  College                  13% (n = 13)
  Graduate degree           4% (n = 4)
Comorbid illnesses
  HTN                      86% (n = 87)
  CHF                      55% (n = 55)
  History of stroke        21% (n = 21)
  PVD                      36% (n = 36)
  Hyperlipidemia           84% (n = 85)
  Tobacco use
    Current                22% (n = 22)
    Ever                   73% (n = 74)
  Coronary artery disease  85% (n = 85)

(a) SD, standard deviation; BMI, body mass index; DM, diabetes mellitus;
IQR, interquartile range; HTN, hypertension; CHF, congestive heart
failure; PVD, peripheral vascular disease.

Table 2. Medications (a)

                   Admission     Discharge

ASA                82% (n = 83)  94% (n = 93)
[beta]-blocker     70% (n = 71)  82% (n = 81)
ACEI               66% (n = 67)  74% (n = 74)
ARB                10% (n = 10)   7% (n = 7)
Lipid lowering
  Statin           62% (n = 63)  69% (n = 68)
  Fibrate           3% (n = 3)    4% (n = 4)
Sulfonylurea       34% (n = 34)  33% (n = 33)
  Glimepiride       8% (n = 8)    9% (n = 9)
  Glipizide        21% (n = 21)  20% (n = 20)
  Glyburide         5% (n = 5)    4% (n = 4)
Metformin          15% (n = 15)  10% (n = 10)
Meglitinide         3% (n = 3)    1% (n = 1)
  Repaglinide       2% (n = 2)    1% (n = 1)
  Nateglinide       1% (n = 1)    0
Acarbose            1% (n = 1)    0
Thiazolidinedione  14% (n = 14)  13% (n = 13)
  Rosiglitazone     7% (n = 7)    5% (n = 5)
  Pioglitazone      7% (n = 7)    8% (n = 8)
Insulin            51% (n = 51)  52% (n = 53)

(a) ASA, aspirin; ACEI, angiotensin converting enzyme inhibitor; ARB,
angiotensin receptor blocker; statin, hydroxymethylglutaryl CoA
reductase.

Table 3. Diabetes health care giver (a)

Care Giver         N   %

Internal Medicine  36  46
Family Medicine    28  28
Endocrinologist    12  12
NP/PA               2   2
Other               5   5

(a) NP, nurse practitioner; PA, physician's assistant.

Table 4. Laboratory monitoring (a)

Variable     N    Mean (SD)               95% CI

HbA1c (%)     82  8.3 (2.0)                 7.8, 8.7
                  Median: 7.5 (IQR: 2.9)
Cr (mg/dL)   101  1.41 (0.9)                1.2, 1.6
TC (mg/dL)    83  183 (79)                166, 200
                  Median: 171 (IQR: 73)
TG (mg/dL)    82  229 (403)               141, 318
                  Median: 126 (IQR: 150)
HDL (mg/dL)   83  44 (15)                  41, 48
LDL (mg/dL)   72  98 (44)                  88, 108
                  Median: 86 (IQR: 55)
EF (%)        79  44 (17)                  40, 48

(a) HbA1c, hemoglobin A1c; IQR, interquartile range; Cr, serum
creatinine; TC, total serum cholesterol; TG, triglycerides; HDL, high
density lipoprotein; LDL, low density lipoprotein; EF, ejection
fraction.

Acknowledgments

The authors thank Jeanne Kimpel, RN, for her help with data collection.

Accepted June 2, 2004.

References

1. Hogan P, Dall T, Nikolov P. American Diabetes Association: economic costs of diabetes in the US in 2002. Diabetes Care 2003;26:917-932.

2. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-986.

3. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulfonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352:837-853.

4. Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol 1995;75:894-903.

5. Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UK-PDS 35): prospective observational study. BMJ 2000;321:405-412.

6. American Diabetes Association. Standards of medical care for patients with diabetes mellitus. Diabetes Care 2003;26(Suppl 1):S33-S50.

7. Beckles GL, Engelgau MM, Narayan KM, et al. Population-based assessment of the level of care among adults with diabetes in the US. Diabetes Care 1998;21:1432-1438.

8. Harris MI. Medical care for patients with diabetes: epidemiologic aspects. Ann Intern Med 1996;124:117-122.

9. Mokdad AH, Bowman BA, Ford ES, et al. The continuing epidemics of obesity and diabetes in the United States. JAMA 2001;286:1195-2000.

10. Janes GR. Ambulatory medical care for diabetes, in Harris MI, Cowie CC, Ster MP, et al. (eds): Diabetes in America. 2nd ed. National Institutes of Health, Publication No. 95-1468, Bethesda, MD, 1995, pp 541-551.

11. Klonoff DC, Schwartz DM. An economic analysis of interventions for diabetes. Diabetes Care 2000;23:390-404.

12. Javitt JC, Aiello LP. Cost-effectiveness of detecting and treating diabetic retinopathy. Ann Intern Med 1996;124:164-169.

13. Vijan S, Hofer TP, Hayward RA. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med 1997;127:788-795.

14. Vijan S, Hofer TP, Hayward RA. Cost-utility analysis of screening intervals for diabetic retinopathy in patients with type 2 diabetes mellitus. JAMA 2000;283:889-896.

15. Gerstein HC, Mann JF, Yi Q, et al. Albuminuria and risk of cardiovascular events, death, and heart failure in diabetic and nondiabetic individuals. JAMA 2001;286:421-426.

16. Richardson WC, Berwick DM, Bisgard JC, et al: To err is human: building a safer health system. Institute of Medicine Report 1999; November. Retrieved February 19, 2004 from http://www.nap.edu/books/0309068371/html

17. Heart Protection Study Collaborative Group. MRC/BHF heart protection study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomized placebo-controlled trial. Lancet 2002;360:7-22.

18. Fertig BJ, Simmons DA, Martin DB. Therapy for diabetes, in Harris MI, Cowie CC, Ster MP, et al (eds): Diabetes in America. 2nd ed. National Institutes of Health. Publication No. 95-1468, Bethesda, MD, 1995, pp 519-540.

19. Boccuzzi SJ, Wogen J, Fox J, et al. Utilization of oral hypoglycemic agents in a drug-insured US population. Diabetes Care 2001;24:1411-1415.

20. Emslie-Smith AM, Boyle DI, Evans JM, et al. Contraindications to metformin therapy in patients with Type 2 diabetes: a population-based study of adherence to prescribing guidelines. Diabet Med 2001;18:483-488.

21. Masoudi FA, Wang Y, Inzucchi SE, et al. Metformin and thiazolidinedione use in Medicare patients with heart failure. JAMA 2003;290:81-85.

22. Wysowski DK, Armstrong G, Governale L. Rapid increase in the use of oral antidiabetic drugs in the United States, 1990-2001. Diabetes Care 2003;26:1852-1855.

23. Meinert CL, Knatterud GL, Prout TE, et al: A study of the effects of hypoglycemic agents on vascular complications in patients with adult-onset diabetes, II: mortality results. Diabetes 1970;19(Suppl 2):789-830.

24. United Kingdom Prospective Diabetes Study Group. Intensive blood-glucose control with sulfonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352:837-853.

25. Deutsch E, Berger M, Kussmaul WG, et al. Adaptation to ischemia during percutaneous transluminal coronary angioplasty: clinical, hemodynamic, and metabolic features. Circulation 1990;82:2044-2051.

26. Kloner RA, Yellon D. Does ischemic preconditioning occur in patients? J Am Coll Cardiol 1994;24:1133-1142.

27. Riddle MC. Editorial: sulfonylureas differ in effects on ischemic preconditioning: is it time to retire glyburide? J Clin Endocrinol Metab 2003;88:528-530.

28. Lee T-M, Chou T-F. Impairment of myocardial protection in type 2 diabetic patients. J Clin Endocrinol Metab 2003;88:531-537.

29. Van den Berghe G, Wouters M, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-1367.

30. Malmberg K, Ryden L, Efendic S, et al. Randomized trial of insulin-glucose infusion followed by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction (DIGAMI Study): effects on mortality at 1 year. J Am Coll Cardiol 1995;26:57-65.

31. Koproski J, Pretto Z, Poretsky L. Effects of an intervention by a diabetes team in hospitalized patients with diabetes. Diabetes Care 1997;20:1553-1555.

32. Clement S, Braithwaite SS, Magee MF, et al. Management of diabetes and hyperglycemia in hospitals. Diabetes Care 2004;27:553-591.

RELATED ARTICLE: Key Points

* Standards of care for patients with diabetes are based on data showing decreased morbidity and mortality rates.

* In a population with significant vascular disease, risk factor modification is paramount.

* Adherence to standard of care guidelines in this patient population remains inadequate.

M. Angelyn Bethel, MD, John Alexander, MD, Jim Lane, PHD, Christina Barkauskas, AB, and Mark N. Feinglos, MD, CM

From the Division of Endocrinology, Metabolism, and Nutrition, the Division of Cardiology, and the Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, and Duke University School of Medicine, Durham, NC.

All work for this publication was done in the Division of Endocrinology, Metabolism, and Nutrition at Duke University Medical Center. The Institutional Review Board at Duke University Medical Center reviewed the study plan and approved all aspects of this study.

Reprint requests to Dr. Mark N. Feinglos, Duke University School of Medicine, Box 3921 Duke University Medical Center, Durham, NC 27710. Email: [e-mail]