This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Variations in the Care of Elderly Persons With Diabetes Among Endocrinologists, General Internists, and Geriatricians

^a Section of General Internal Medicine, Department of Medicine
^b Diabetes Research and Training Center
^c Pritzker School of Medicine, University of Chicago, Illinois

Marshall H. Chin, University of Chicago, Section of General Internal Medicine, 5841 South Maryland Avenue, MC 2007, Chicago, IL 60637 E-mail: mchin{at}medicine.bsd.uchicago.edu.

Decision Editor: William B. Ershler, MD

	Abstract

Top Abstract Methods Results Discussion References

 
Background. The American Diabetes Association (ADA) clinical practice recommendations have been widely promoted, but they lack a geriatric-specific approach to care. We aimed to determine the style of care that endocrinologists, general internists, and geriatricians provided to their elderly patients with diabetes and to what extent these medical professionals adhered to the ADA standards.

Methods. We performed a retrospective cohort study of a stratified sample of 531 diabetic patients aged 65 years and older from the endocrinology, general internal medicine, and geriatrics clinics of an urban academic medical center.

Results. Patients of geriatricians were older, had higher comorbidity, and were more likely to be demented. The average number of diabetic complications was similar across the specialties, although patients of endocrinologists had higher prevalence of neuropathy and retinopathy compared with patients of geriatricians. Endocrinologists were more likely to use insulin, multiple types of insulin, and combined oral hypoglycemic and insulin therapies. Most patients had hemoglobin A1c measured, and average values were similarly high across specialties at 8.6%. Blood pressures were above 130/85 mm Hg in 85% of the patients. All specialties rarely measured urine microalbumin; geriatricians seldom performed fractionated cholesterol tests, and ophthalmology visits occurred in only half of the patients.

Conclusion. Endocrinologists had the most aggressive, complex diabetes treatment regimens, although geriatricians had older patients with more dementia and lower prevalence of microvascular complications. Average hemoglobin A1c levels and blood pressures were higher than recommended among patients of all three specialties. Screening for diabetic complications and hyperlipidemia was lower than advised.

AMONG the approximately 7.8 million people diagnosed with diabetes in 1993, 41% were 65 years or older ⁽¹⁾. Yet no definitive controlled studies have been conducted to provide guidance on the best glycemic management approach for these elderly patients. The United Kingdom Prospective Diabetes Study demonstrated that tight glucose control reduced the risk of microvascular complications in patients with type 2 diabetes who had a median age of 54 years at the time of enrollment in the study ⁽²⁾. However, it is unclear to what extent these findings are applicable to the geriatric population because elderly patients often have a shorter lifespan, more comorbidities, and complicating factors such as diminished cognitive status.

Nonetheless, several diabetes practice guidelines have been developed to help improve care. Probably the most influential diabetes guidelines in the United States are those of the American Diabetes Association (ADA). For example, these guidelines have affected the performance measures of the Diabetes Quality Improvement Project ⁽³⁾, a joint effort of the ADA, the Foundation for Accountability, the Health Care Financing Administration, the National Committee on Quality Assurance, the American Academy of Family Physicians, the American College of Physicians, and the Veterans Administration. In turn, the performance measures have been incorporated into the standards used by the peer review organizations that evaluate the quality of care within the Medicare program. Therefore, the ADA standards impact the care provided to elderly persons with diabetes through a variety of means.

However, perhaps reflecting the dearth of scientific evidence in elderly populations, the clinical practice recommendations of the ADA give scant attention to aging issues and diabetes. For example, with regard to setting treatment goals for type 2 diabetes, the guidelines only recommend that health providers consider "other patient factors that may increase risk or decrease benefit (e.g., advanced age, end-stage renal disease, advanced cardiovascular or cerebrovascular disease, or other coexisting diseases that will materially shorten life expectancy)" ⁽⁴⁾. In practice, optimal care of the elderly patient is dependent on clinical judgment, extrapolation of scientific evidence from younger populations, and, hopefully, patient preferences.

The effect of different physician practice styles on the care of elderly diabetic patients, in particular the role of specialty training, has been underexamined. In a survey of physicians who cared for Medicare beneficiaries, internists were more likely than family practitioners and general practitioners to report performing peripheral neurological and circulatory examinations, and serum creatinine tests ⁽⁵⁾. A Medicare claims study found that compared with general practitioners, internists or family practitioners were more likely to measure glycosylated hemoglobin in their elderly patients ⁽⁶⁾. A recent study of elderly diabetic patients in the Medicare Current Beneficiary Survey that also used claims data, determined that endocrinologists were more likely to adhere to process of care standards, such as measurement of glycosylated hemoglobin and referral for ophthalmological screening, than were internists, family practitioners, and general practitioners ⁽⁷⁾. However, no information has been available regarding the aggressiveness of the medication regimens, nor has any study compared diabetes care by geriatricians with that of other physicians.

Our goal, therefore, was to explore differences in the care of elderly diabetic patients by physicians in the following three specialties, each with relative strengths: (i) endocrinologists—physicians who should be most knowledgeable in the disease-specific care of these patients; (ii) general internists—physicians who treat a variety of illnesses across a wide age range; and (iii) geriatricians—physicians with the most training in aging-related issues. Our intention in this study is not to deem one style of care necessarily better than another but to describe differences in patient populations, processes of care, intensity of medical management, and resource utilization and to determine to what extent each specialty adhered to the guidelines of the ADA.

	Methods

Top Abstract Methods Results Discussion References

 
Patient Population
We studied 531 diabetic patients aged 65 years and older who attended the endocrinology, general internal medicine, or geriatrics clinic at the University of Chicago Hospitals, an urban academic medical center. No overriding philosophy of geriatric diabetes care dominates the institution; thus, we hypothesized that the practice patterns in the three clinics would differ. Using the administrative billing system, we initially identified 200 patients randomly chosen from each of the clinics with outpatient International Classification of Diseases-9-Clinical Modification (ICD-9-CM) codes of 250.00–250.91 as primary or secondary diagnoses between September 1996 and September 1997 ⁽⁸⁾. To confirm the diagnosis of diabetes, patients either had to be treated with insulin or oral hypoglycemic agents (87%), have a documented glycosylated hemoglobin greater than 7.5% (5%), or have a physician-recorded diagnosis of diabetes on the chart (8%). The few patients who qualified via a physician diagnosis of diabetes were equally distributed across the three specialties. If a patient attended more than one of the clinics (25%), then he or she was assigned to the specialty of the clinic they visited most frequently between January 1996 and March 1998. The eight patients who had equal numbers of visits to two of the clinics were assigned in hierarchical order to endocrinology, geriatrics, and general internal medicine.

Data Collection
The primary ADA standards we used were measurement of hemoglobin A1c, urine microalbumin, and fractionated cholesterol; ophthalmology visit; hemoglobin A1c value less than 8%; and blood pressure less than 130/85 mm Hg ⁽⁴⁾. From the medical records, we recorded type of diabetes, diabetic complications, comorbidities, medications, insulin dosing, and blood pressure. We used the Charlson Comorbidity Index, a weighted summary of comorbid conditions validated to predict death, as a summary measure of the burden of comorbid illness ⁽⁹⁾. Because diabetes was the study condition, we removed diabetes from the Charlson Comorbidity Index. We obtained the most recent laboratory values from the medical records and laboratory computer system. Demographic information such as date of birth, gender, race, and insurance status was obtained from the administrative billing system or medical records. Referrals to ophthalmology and podiatry were obtained from the chart and administrative database. Costs, outpatient visits, and hospitalizations from January 1996 to March 1998 were acquired from the administrative database.

Statistical Analysis
We compared patient characteristics and the use of resources by employing the chi-square test for categorical variables and the t test or Wilcoxon rank sum test for continuous variables. We performed analyses stratified by specialty, comparing the specialties two at a time. To determine if specialty was correlated with whether a patient was on insulin therapy, we used logistic regression analysis, adjusting for patient's age, gender, race, diabetic complications, hypertension, dementia, and Charlson Comorbidity Index score. With linear regression, we analyzed whether specialty was correlated with the hemoglobin A1c value as well as total costs, adjusting for the same covariates. Because total costs were not normally distributed, we examined the log transformation in the linear regression analyses.

We also stratified costs by their relation to diabetes. We defined diabetes-related costs as those with primary ICD-9-CM diagnostic codes of diabetes mellitus (250.x), other disorders of pancreatic internal secretion including hypoglycemia (251.x), and diabetic complications (renal: 250.4, 581, 583, 585; ophthalmic: 250.5, 362, 365, 366, 369; neurological: 250.6, 337, 357, 354, 355, 358.1, 713.5; and peripheral circulatory: 250.7, 440, 441, 442, 443, 785.4). Because heart failure is a common and costly condition among elderly persons with diabetes ⁽¹⁰⁾, we also stratified costs by their relation to heart failure (428.x). We also analyzed whether outliers accounted for any differences in costs across the specialties. We performed a subanalysis that excluded patients who saw more than one of the three targeted specialties; we also compared patients with and without dementia.

	Results

Top Abstract Methods Results Discussion References

 
Patient Characteristics
More than 95% of the patients had type 2 diabetes, and most were African American (Table 1 ). Virtually all patients had Medicare insurance, and approximately 80% also had secondary insurance. Patients of geriatricians were older and more likely to be female. They also had more comorbidity, higher systolic blood pressures, and were more apt to be demented. The overall number of diabetic complications was similar across the different specialties. However, patients of endocrinologists had higher prevalence of neuropathy and retinopathy compared with patients of geriatricians. Approximately 85% of patients of all specialties had systolic blood pressure greater than 130 mm Hg, 20% had diastolic blood pressure greater than 85 mm Hg, and 85% had either systolic or diastolic elevated blood pressure. Approximately one-quarter of the patients had systolic blood pressure of 160 mm Hg or greater.

Resource Utilization and Costs
Patients of endocrinologists had fewer hospitalizations than patients of general internists, and they had fewer visits to the primary specialty clinic than patients of generalist physicians (Table 5 ). In multivariable analyses, patients of endocrinologists had lower total costs compared with patients of general internists or geriatricians. Independent correlates of higher total costs were more diabetic complications, increased Charlson Comorbidity Index score, African American race, and not experiencing dementia. Male gender tended ( p < .10) to be associated with higher total costs. Geriatricians and general internists had a few expensive outlier patients. When we excluded these patients from the analyses, the differences in costs across specialties diminished but remained statistically significant.

Dementia Subanalysis
The 49 demented patients were less likely than nondemented patients to receive measurement of HDL cholesterol (16% vs 39%, ) and mammograms (21% vs 46%, ). However, average hemoglobin A1c values were similarly high at approximately 8.6%. In multivariable analyses, demented patients had lower total and outpatient costs than nondemented patients.

Subanalyses Excluding Patients Seeing Multiple Specialties
Subanalyses excluding patients who saw more than one of the three specialties were similar to the main analyses.

	Discussion

Top Abstract Methods Results Discussion References

 
Even when we limited our study population to those patients 65 years and older, we found that the types of patients and treatment styles of endocrinologists, general internists, and geriatricians differed. In general, geriatricians treated older patients with more comorbidity, more dementia, and fewer microvascular complications, and endocrinologists used the most complex, aggressive hypoglycemic treatment regimens. Patients of all specialties frequently received care that was less aggressive than recommended by the ADA regarding glycemic control, blood pressure treatment, and screening for diabetic complications and hyperlipidemia.

Even though endocrinologists used more aggressive insulin regimens, average hemoglobin A1c values across the specialties were similarly high at approximately 8.6%. This result could reflect incomplete case-mix adjustment. Patients of endocrinologists may have more severe diabetes. However, the similar level of glucose control across the specialties, despite variation in the aggressiveness of treatment style, could also reflect a predilection of endocrinologists to use insulin sooner in the course of illness than would geriatricians or general internists, or it could reflect equal effectiveness of the different regimens. In any case, glycemic control was frequently loose across all specialties. In addition, the elevated hemoglobin A1c values in patients with and without dementia may reflect a relatively unaggressive approach to glycemic control in general among elderly persons with diabetes.

The care of elderly diabetic patients raises difficult management issues ⁽¹¹⁾. With regard to tightness of glycemic control, most clinicians would agree that at a minimum, elderly patients should be kept free of the more immediate effects of hyperglycemia, including hyperosmolar symptoms, and perhaps diminished cognition and impaired leukocyte function ^(12)(13). However, dietary restrictions, insulin shots, and oral hypoglycemic agent treatment can adversely affect a patient's quality of life in the present. Therefore, it would be ideal to know which elderly patients are likely to benefit from more aggressive glucose control. A particularly important clinical challenge is that elderly diabetic patients are a heterogeneous group. Some have newly diagnosed diabetes, are asymptomatic, and are unlikely to develop complications from their disease. Others may have had diabetes for several years with moderate glucose control and are at risk for diabetic complications. Still others may have had longstanding disease with severe complications. It is uncertain whether the variations in practice styles we found across specialties reflect the most appropriate individualization of care for the patients.

Whereas younger patients have many years in which to develop complications from diabetes, elderly persons are more likely to die sooner and from unrelated diseases. Although we are unaware of any randomized controlled trials of glycemic control in the geriatric population ^(14)(15), decision analytical policy models using data extrapolated from younger populations suggest diminishing benefit from glycemic control in patients with later onset of diabetes and lower starting values of hemoglobin A1c ^(16)(17)(18). Moreover, available evidence from the Systolic Hypertension in the Elderly Program ⁽¹⁹⁾, in conjunction with the United Kingdom Prospective Diabetes Study ⁽²⁰⁾, indicates that aggressive treatment of hypertension may be equally or perhaps more important than tight glycemic control in prevention of major cardiovascular events, microvascular complications, and diabetes-related deaths in geriatric patients with diabetes. In addition, the recent Heart Outcomes Prevention Evaluation Study found that the angiotensin-converting enzyme inhibitor ramipril reduced the composite endpoint of myocardial infarction, stroke, or cardiovascular death in patients without heart failure who had diabetes plus one other cardiovascular risk factor ⁽²¹⁾. More than half of the patients in this trial were 65 years or older at enrollment. Therefore, an evidence-based framework for treating elderly diabetic persons needs to consider the special benefits of glycemic control in patients with relatively high hemoglobin A1c values, stress the importance of hypertension control, incorporate the use of angiotensin-converting enzyme inhibitors in appropriate patients, and perhaps most importantly, focus on overall quality of life.

Patients of endocrinologists had lower total costs than patients of general internists or geriatricians. The lower costs were manifest in the nondiabetes-coded visits. In contrast, Greenfield and colleagues found that endocrinologists had increased resource utilization and similar outcomes compared with generalist physicians ^(22)(23), but their diabetic study population included many nongeriatric patients. Therefore, comparison to our study is difficult. We suspect that the increased comorbidity of the patients of generalist physicians may have led to higher costs than those of patients of endocrinologists in our study. For example, heart failure–associated costs were higher for patients of geriatricians than for patients of endocrinologists.

Our study is limited in that it was performed at one urban academic medical center. In addition, the charts frequently missed information on the duration of diabetes. Therefore, we were unable to study that variable. Some illnesses such as dementia may have been underdocumented or underdiagnosed by physicians, and it is possible that the specialties vary in how consistently they list diabetic complications. We were also unable to determine if physicians had attempted maximal oral hypoglycemic therapy before switching to insulin regimens. In addition, we could not assess patient preferences from the medical records, and patients probably chose their particular physician based on nonrandom features. We may not have adjusted fully for case-mix differences among the specialties. Therefore, interpretation of some analyses, such as hemoglobin A1c values, is difficult. Although we could capture inpatient and outpatient costs, we did not have home care and pharmacy costs in our database, nor costs from outside the system. We also urge caution in interpreting the assignment of costs to diabetes-related and non–diabetes-related services. ICD-9-CM codes from administrative databases have limited validity as markers of the primary reason for an encounter, and it is difficult to assign causation for conditions with multifactorial etiologies such as cardiovascular disease.

However, our exploratory study is the first to compare the care of elderly diabetic patients by endocrinologists, general internists, and geriatricians. Given that more than 40% of patients with diabetes in the United States are 65 years or older, it is critical to examine how different specialties vary in their treatment of these patients. The fundamental question, ideally answered by a multisite prospective trial, is what style of care is most likely to prevent diabetic complications and improve quality of life and functioning in older persons? Other pressing research questions include whether physicians truly individualize care to their elderly patients or whether they are heavily influenced by the treatment paradigm of their specialty. What information is given to patients about risks and benefits of different treatment approaches? How will patients act in a shared physician–patient participatory relationship? What are ethical ways of approaching the care of patients with cognitive impairment ⁽²⁴⁾? What specific clinical problems should be emphasized in treatment? For example, should clinicians concentrate on glycemic control, hypertension, inhibition of angiotensin-converting enzyme, comorbidities, or functional problems in their elderly diabetic patients? The clinical practice recommendations of the ADA would benefit by viewing elderly diabetic patients through a more flexible geriatric framework in which care is individualized. The treatment approach that is most likely to optimize an elderly diabetic patient's quality of life probably will come from a strategy that incorporates existing scientific evidence, clinical judgment and counseling, and explicit involvement of an informed patient in the decision-making process ⁽²⁵⁾.

	Acknowledgments

 
This research was supported by the National Institute on Aging (Geriatric Academic Program Award 5-K12-AG-00488). This study was presented in part at the annual meeting of the Midwest Society of General Internal Medicine on September 19, 1998, in Chicago, Illinois.

	Footnotes

 
Michael Nerney is currently a medical student at the Chicago College of Osteopathic Medicine, Midwestern University, Downers Grove, Illinois.

Received July 19, 1999

Accepted December 29, 1999

	References

Top Abstract Methods Results Discussion References

 

1. Kenny SJ, Aubert RE, Geiss LS, 1995National Institutes of Health publication 95-1468.. Prevalence and incidence of non-insulin-dependent diabetes. Harris MI, Cowie CC, Stern MP, Boyko EJ, Reiber GE, Bennett PH, , ed.Diabetes in America 2nd ed. 47-67. US Government Printing Office, Washington, DC.
2. UK Prospective Diabetes Study Group1998. Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352:837-853. [Medline]
3. Diabetes Quality Improvement Project Initial Measure Set (final version). Alexandria, VA: American Diabetes Association; 1998.
4. American Diabetes Association1999. Clinical practice recommendations 1999. Diabetes Care. 22: (suppl 1) S33
5. Drass J, Kell S, Osborn M, et al. 1998. Diabetes care for Medicare beneficiaries: attitudes and behaviors of primary care physicians. Diabetes Care 21:1282-1287. [Abstract]
6. Weiner JP, Parente ST, Garnick DW, Fowles J, Lawthers AG, Palmer RH, 1995. Variation in office-based quality: a claims-based profile of care provided to Medicare patients with diabetes. JAMA 273:1503-1508. [Abstract/Free Full Text]
7. Chin MH, Zhang JX, Merrell K, 2000. Specialty differences in the care of older patients with diabetes. Med Care 38:131-140. [Medline]
8. Medicode. The International Classification of Diseases. 9th revision. Clinical Modification. 5th ed. Salt Lake City, UT: Medicode Publications; 1995.
9. Charlson ME, Pompei P, Ales KL, MacKenzie CR, 1987. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 40:373-383. [Medline]
10. Wingard DL, Barrett-Connor E, 1995National Institutes of Health publication 95-1468.. Heart disease and diabetes. Harris MI, Cowie CC, Stern MP, Boyko EJ, Reiber GE, Bennett PH, , ed.Diabetes in America 2nd ed. 429-448. US Government Printing Office, Washington, DC.
11. Lunt H, 1996. Diabetes mellitus in older patients: is tight blood glucose control warranted?. Drugs and Aging 8:401-407. [Medline]
12. Gradman TJ, Laws A, Thompson LW, Reaven GM, 1993. Verbal learning and/or memory improves with glycemic control in older subjects with non-insulin-dependent diabetes mellitus. J Am Geriatr Soc 41:1305-1312. [Medline]
13. Rayfield EJ, Ault MJ, Keusch GT, Brothers MJ, Nechemias C, Smith H, 1982. Infection and diabetes: the case for glucose control. Am J Med. 72:439-450. [Medline]
14. Morisaki N, Watanabe S, Kobayashi J, et al. 1994. Diabetic control and progression of retinopathy in elderly patients: five-year follow-up study. J Am Geriatr Soc. 42:142-145. [Medline]
15. Araki A, Ito H, Hattori A, et al. 1993. Risk factors for development of retinopathy in elderly Japanese patients with diabetes mellitus. Diabetes Care. 16:1184-1186. [Abstract]
16. Vijan S, Hofer TP, Hayward RA, 1997. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med. 127:788-795. [Abstract/Free Full Text]
17. Eastman RC, Javitt JC, Herman WH, et al. 1997. Model of complications of NIDDM. I. Model construction and assumptions. Diabetes Care. 20:725-734. [Abstract]
18. Eastman RC, Javitt JC, Herman WH, et al. 1997. Model of complications of NIDDM. II. Analysis of the health benefits and cost-effectiveness of treating NIDDM with the goal of normoglycemia. Diabetes Care. 20:735-744. [Abstract]
19. Curb JD, Pressel SL, Cutler JA, et al. 1996Published erratum appears in JAMA. 1997;277:1356.. Effect of diuretic-based antihypertensive treatment on cardiovascular disease risk in older diabetic patients with isolated systolic hypertension. Systolic Hypertension in the Elderly Program Cooperative Research Group. JAMA 276:1886-1892. [Abstract/Free Full Text]
20. UK Prospective Diabetes Study Group1998. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ 317:703-713. [Abstract/Free Full Text]
21. Yusuf A, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G, 2000. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 342:145-153. [Abstract/Free Full Text]
22. Greenfield S, Nelson EC, Zubkoff M, et al. 1992. Variations in resource utilization among medical specialties and systems of care: results from the Medical Outcomes Study. JAMA 267:1624-1630. [Abstract/Free Full Text]
23. Greenfield S, Rogers W, Mangotich M, Carney MF, Tarlov AR, 1995. Outcomes of patients with hypertension and non-insulin-dependent diabetes mellitus treated by different systems and specialties: results from the Medical Outcomes Study. JAMA 274:1436-1444. [Abstract/Free Full Text]
24. Sachs GA, 1994. Flu shots, mammograms and Alzheimer's disease: ethics of preventive medicine and dementia. Alzheimer Dis Assoc Disord 8:8-14. [Medline]
25. Chin MH, Polonsky TS, Thomas VD, Nerney MP, 2000. Developing a conceptual framework for understanding illness and attitudes in older, urban African Americans with diabetes. Diabetes Edu 26:439-449.

This article has been cited by other articles:

				F A McAlister, S R Majumdar, D T Eurich, and J A Johnson The effect of specialist care within the first year on subsequent outcomes in 24 232 adults with new-onset diabetes mellitus: population-based cohort study Qual. Saf. Health Care, February 1, 2007; 16(1): 6 - 11. [Abstract] [Full Text] [PDF]

				J. E. Morley and J. H. Flaherty Editorial It's Never Too Late: Health Promotion and Illness Prevention in Older Persons J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2002; 57(6): M338 - 342. [Full Text]

				J. E. Morley Editorial: Drugs, Aging, and the Future J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2002; 57(1): M2 - 6. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation