Several currently available diabetes mellitus

Transcription

1 Diabetes Mellitus in Managed Care: Complications and Resource Utilization CDC Diabetes in Managed Care Work Group Objectives: To implement a diabetes mellitus surveillance system designed to use administrative data and to demonstrate how it can be used by managed care organizations (MCOs) with different administrative data systems to estimate the prevalence of diabetic complications and describe utilization of services in persons with and without complications and comorbidities. Study Design, Patients, and Methods: We identified individuals with diabetes mellitus in 3 MCOs in 1993 using 4 sources of computerized data records: inpatient, pharmacy, outpatient, and laboratory. The presence of diabetes mellitus complications and cardiovascular comorbidities were determined using diagnostic and procedural codes. Use of healthcare resources by persons with and without complications and comorbidities was determined from computerized administrative data. Results: The most prevalent complication or comorbidity was cardiovascular disease (45%-53%), followed by eye disease (20%-34%), lower extremity disease (8%-20%), and renal disease (3%-6%). The presence of multiple complications was common and ranged from 14% to 34% in the 3 MCO populations. Compared with persons with none, persons with 2 or more complications or comorbidities used moderately more primary care services ( times more) and markedly more specialty care services ( times more), emergency department visits ( times more), and hospital stays ( times more). Conclusions: Diabetic complications were common and had a large impact on patients use of healthcare services. Within MCOs, administrative databases are useful tools for estimating and monitoring the prevalence of diabetic complications and the use of healthcare resources associated with these complications. (Am J Manag Care 2001;7: ) Several currently available diabetes mellitus treatments have proven effective in reducing diabetic and cardiovascular morbidity and mortality in persons with diabetes mellitus. 1-7 However, the current level of care for diabetes mellitus, ie, the delivery of effective treatments, is suboptimal As major providers of healthcare services for individuals with diabetes mellitus, 13 managed care organizations (MCOs) can potentially improve the level of care by monitoring diabetes mellitus, its complications, and use of healthcare resources associated with these complications. During the past decade, the National Committee for Quality Assurance collaborated with MCOs, health services researchers, clinical specialists, and public health experts to create the Health Plan Employer Data and Information Set 2000, a comprehensive set of quality indicators that includes a per- Members of the CDC Diabetes in Managed Care Work Group are as follows: Michael M. Engelgau, MD, and Linda S. Geiss, MA, Division of Diabetes Translation, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA; Diane L. Manninen, PhD, and Carlyn E. Orians, MA, Battelle Centers for Public Health Research and Evaluation, Seattle, WA; Edward H. Wagner, MD, Group Health Cooperative of Puget Sound, Seattle; Neal M. Friedman, MD, Kaleida Health, Buffalo, NY; Judith S. Hurley, MS, RD, and Kathryn M. Trinkaus, PhD, Lovelace Respiratory Research Institute, Southwest Center for Managed Care Research, Albuquerque, NM; and Deborah Shatin, PhD, and Krista A. Van Vorst, MS, Center for Health Care Policy and Evaluation, UnitedHealth Group, Chicago, IL. This study was funded by the Division of Diabetes Translation, Centers for Disease Control and Prevention, Atlanta, GA. Presented in part at the American Public Health Association Annual Meeting, November 15-18,1998, Washington, DC. Address correspondence to: Michael M. Engelgau, MD, Division of Diabetes Translation, Mailstop K-10, Centers for Disease Control and Prevention, 4770 Buford Highway NE, Atlanta, GA mxe1@cdc.gov. VOL. 7, NO. 5 THE AMERICAN JOURNAL OF MANAGED CARE 501

2 Table 1. Data Specifications Variable Description Specification Patient demographics Age as of July 1, 1993 Sex Days enrolled in 1993 Outpatient visits 1993 No. of visits to primary care provider Includes family practitioners, pediatricians, internal medicine providers, and doctors of osteopathy No. of visits to endocrinologist No. of visits to cardiologist No. of visits to nephrologist No. of visits to ophthalmologist/optometrist No. of visits to orthopedic surgeon No. of visits to podiatrist No. of visits to emergency department Hospital visits (inpatient, overnight) 1993 Total no. of hospital discharges Overnight stay required No. of hospital visits for stroke ICD-9 codes and No. of hospital visits for major cardiovascular disease ICD-9 codes No. of hospital visits for ischemic heart disease ICD-9 codes No. of hospital visits for lower extremity amputation ICD-9 procedure code 84.1; CPT-4 codes 28810, 28820, 28825, 28800, 28805, , , 27886, , 27596, 27598, 27290, and No. of hospital visits for kidney transplantations ICD-9 procedure code 55.6; CPT-4 codes and Laboratory tests or procedures performed, outpatient only 1993 No. of glycosylated hemoglobin tests CPT-4 code Dilated eye examination CPT-4 codes 92002, 92004, 92012, 92014, 92018, 92019, 92225, 92226, 92235, and Prescription drug information 1993 Prescription for an oral glycemic medication Tolbutamide, chlorpropamide, acetohexamide, tolazamide, glyburide, glipizide, or metformin Prescription for insulin Regular insulin, NPH insulin, lente insulin, ultralente insulin, or protamine zinc insulin Prescription for human insulin To identify pump use Prescription for an angiotensin-converting enzyme inhibitor Quinapril hydrochloride, ramipril, captopril, benazepril hydrochloride, fosinopril sodium, lisinopril, or enalapril maleate Patient complication or comorbidity 1993 Any major cardiovascular disease ICD-9 codes Ischemic heart disease ICD-9 codes Cerebrovascular disease ICD-9 codes Hypertensive disease ICD-9 codes Any renal disease ICD-9 codes 250.4, 581, 583, 791.0, 585, 586, V42.0, V45.1, and V56 End-stage renal disease ICD-9 codes 585, 586, V42.0, V45.1, and V56 Any eye disease ICD-9 codes 250.5, 361, 365, 366, 369, 379.2, 362.0, and Retinopathy ICD-9 codes and Any lower extremity disease Any of the following: amputation, peripheral vascular disease, ulcer, musculoskeletal deformity, or neuropathy Amputation If lower extremity amputation is done in the hospital or as an outpatient Peripheral vascular disease ICD-9 codes 250.7, 440.2, 442.3, , and Ulcer, inflammation, or infection ICD-9 codes 454, 707.1, , 681.1, , , , , , , , , and Musculoskeletal deformity ICD-9 codes 731.8, 735, and Neuropathy ICD-9 codes 337.1, 357.2, 355, 358.1, 713.5, and ICD-9 = International Classification of Diseases, Ninth Edition; CPT-4 = Current Procedural Terminology 4; NPH = neutral protamine 502 THE AMERICAN JOURNAL OF MANAGED CARE MAY 2001

3 ... Diabetes Complications and Resource Utilization... formance measurement set for diabetes mellitus. 14 In addition, the National Committee for Quality Assurance has called for development of integrated health information networks that can track clinical performance. Such information networks contain demographic data and information about preventive services, referrals and consultations, and services utilization. 15 These systems have proven to be a valid means of identifying, tracking, and determining the cost of various comorbidities Thus, monitoring this information is important for planning and targeting interventions to reduce morbidity and the cost of serious diabetic complications. To date, few studies have described the current level of diabetes mellitus related complications in MCO diabetic patients and the use of healthcare services by persons with these complications. In this study, we use a diabetes mellitus surveillance health information system designed to monitor diabetes mellitus using administrative data and demonstrate how 3 MCOs with different administrative data systems used it to estimate the prevalence of diabetic complications and describe utilization of services in persons with and without complications and comorbidities.... RESEARCH DESIGN AND METHODS... We used a diabetes mellitus surveillance system described previously by investigators at the Centers for Disease Control and Prevention, Atlanta, GA; Battelle Centers for Public Health Research and Evaluation, Seattle, WA; and 3 MCOs: Group Health Cooperative of Puget Sound, Seattle; Lovelace Health Systems, Albuquerque, NM; and UnitedHealth Group affiliated health plan, Chicago, IL. 12 The surveillance system identified patients with diabetes mellitus from each of the 3 MCOs in 1993 using 4 sources of computerized data records: inpatient (hospitalization), pharmacy, outpatient, and laboratory. To be eligible for inclusion, patients were required to meet 1 or more of the following criteria: 1 or more overnight hospital visits with a primary or secondary diagnosis of diabetes mellitus (International Classification of Diseases, Ninth Edition, code 250), 1 or more prescriptions for insulin (regular insulin, NPH insulin, lente insulin, ultralente insulin, or protamine zinc insulin) or an oral hypoglycemic medication (tolbutamide, chlorpropamide, acetohexamide, tolazamide, glyburide, glipizide, or metformin), 2 or more outpatient visits with a diagnosis of diabetes mellitus (International Classification of Diseases, Ninth Edition, code 250), or 2 or more glycosylated hemoglobin laboratory tests (Current Procedural Terminology code 83036). Patients were excluded if they had a diagnosis indicating a pregnancy at any time during the year. This was intended to exclude individuals with gestational diabetes mellitus but also resulted in the exclusion of type 1 and type 2 diabetes mellitus patients who experienced a pregnancy in Five categories of variables were collected from existing administrative data for each patient: demographics, overnight hospital stays, outpatient visits, selected laboratory tests and procedures, and prescription drugs. For outpatient visits, data were collected only on visits to particular kinds of specialists (those most likely to be used by patients with diabetes mellitus): primary care physicians, endocrinologists, cardiologists, nephrologists, ophthalmologists/optometrists, orthopedic surgeons, and podiatrists. Codes used to define each broad complication and its disease subsets are shown in Table 1. Cardiovascular disease (including hypertension, ischemic heart disease [IHD], and cerebrovascular disease), renal disease (including proteinuria, renal failure, end-stage renal disease, and kidney replacement), eye disease (including retinopathy, detached retina, vitreous hemorrhage, glaucoma, cataract, and blindness), or lower extremity disease (LED) (including neuropathy, peripheral vascular disease, musculoskeletal deformity, ulcer, and amputation) were considered present if a diagnostic or procedural code corresponding to the condition was recorded at least once during a hospital stay or outpatient visit. Analysis We determined the prevalence of complications and the rates of service utilization associated with these complications. We also examined the prevalence of additional complications given the presence of 1 complication (eg, we examined the prevalence of eye disease, renal disease, and LED among persons with cardiovascular disease). We also calculated the number of major complications per person and then examined the relation between number of complications and services utilization. Prevalence and utilization data were adjusted by age using 1980 estimates of the US population with diabetes mellitus to facilitate comparisons with other existing estimates. VOL. 7, NO. 5 THE AMERICAN JOURNAL OF MANAGED CARE 503

4 ... RESULTS... In all, there were 16,363 patients with diabetes mellitus at 3 MCOs that provided healthcare services in different regions of the United States (northwest, southwest, and southeast). These organizations are identified as MCO1, MCO2, and MCO3. There were racial and ethnic differences among these 3 populations, but racial/ethnic data were not included in the MCO administrative records and therefore were not analyzed. MCO2 had a younger diabetic population (mean age, 48 years; 7% 65 years) than MCO1 and MCO3 (mean ages, 60 and 58 years; 43% and 39% 65 years, respectively). Male patients ranged from 50% of the enrollment in MCO1 to 56% in MCO2. Twenty-six percent of patients in MCO1 and MCO2 and 28% in MCO3 were taking insulin; the remainder took oral agents (MCO1, 55%; MCO2, 53%; and MCO3, 46%) or no hypoglycemic medication (MCO1, 19%; MCO2, 21%; and MCO3, 26%). Complications and Comorbidities Cardiovascular disease was the most common of the 4 major complications examined (Table 2). Approximately half of the patients with diabetes mellitus (45%-53%, adjusted for age) had some form of cardiovascular disease. Hypertension was the Table 2. Age-Adjusted Prevalence of Diabetic Complications and Cardiovascular Comorbidities per 100 Patients With Diabetes Mellitus, 1993 MCO = managed care organization. *Total includes diseases or conditions that are not listed. See Table 1. dominant form of cardiovascular disease, accounting for approximately two thirds of all cardiovascular disease. Eye disease was the second most common complication, with an age-adjusted prevalence range of 20% to 34%. Almost half of all persons with eye disease had diabetic retinopathy. The age-adjusted prevalence of LED varied 2- to 3-fold across the 3 MCOs, ranging from a low of 8% to a high of 20%. Renal disease was the least common of the complications examined, with an age-adjusted prevalence range of 3% to 6%. End-stage renal disease was present in 2% to 3% of patients. We determined the frequency of any complication (ie, any cardiovascular disease, eye disease, LED, or renal disease) and found that the majority of patients had at least 1 complication (MCO1, 62%; MCO2, 54%; and MCO3, 68%). Having 2 or more complications also was fairly common, with the frequency ranging from 14% in MCO2 to 24% in MCO1 to 34% in MCO3. Among those who had at least 1 diabetic complication, cardiovascular complications were common across all the MCOs and varied by specific complication. Among those with renal disease as 1 of their complications, hypertension was common (56%-71%), followed by IHD (24%-33%) and cerebrovascular disease (5%- 9%). For those with LED or eye disease as 1 of Patients, % Complication MCO1 MCO2 MCO3 Major cardiovascular diseases Total* Ischemic heart disease Cerebrovascular disease Hypertension Eye disease Total* Diabetic retinopathy Lower extremity disease Total* Peripheral vascular disease Neuropathy Renal disease Total* End-stage renal disease their complications, a similar pattern but with lower frequencies of cardiovascular complications or comorbidities was noted. Among those with LED, 38% to 45% had hypertension, 20% to 25% had IHD, and 5% to 12% had cerebrovascular disease; among those with eye disease, 36% to 48% had hypertension, 15% to 24% had IHD, and 5% to 7% had cerebrovascular disease. In all MCOs, patients with renal disease tended to be more likely than those with LED or eye disease to have any of the cardiovascular comorbidities. 504 THE AMERICAN JOURNAL OF MANAGED CARE MAY 2001

5 ... Diabetes Complications and Resource Utilization... Among persons with at least 1 cardiovascular comorbidity, diabetic complications were also common (Figure 1). Eye disease, present in approximately 23% to 43% of patients, was most common, followed by LED (8%-34%). Renal disease was least common (5%-16%). Across MCOs, regardless of which cardiovascular comorbidity was present, the range and pattern of the prevalence of diabetic complications were similar. Healthcare Utilization Service utilization rates were examined separately for patients with no, 1, and multiple complications or comorbidities. Patients with no complications had 0.09 to 0.12 hospital stays, 0.10 to 0.22 emergency department visits, 0.98 to 1.57 specialist visits, and 2.93 to 4.70 primary care visits during Utilization rates increased dramatically as the number of complications and comorbidities increased (Table 3). Compared with persons with none, persons with 2 or more complications or comorbidities had markedly increased hospital stays ( times more), emergency department visits ( times more), and use of specialty care ( times more) but only modestly increased use of primary care services ( times more). Regarding preventive care practices, persons with 2 or more complications tended to undergo more glycosylated hemoglobin tests and eye examinations. A fairly substantial proportion of persons with no or 1 complication or comorbidity received 2 or more glycosylated hemoglobin tests during the year (no complications, 22%-48%; 1 complication, 18%- 57%), and this increased to 25% to 61% in persons with 2 or more complications or comorbidities. However, the proportion who received eye examinations increased more dramatically with the number of complications, from 16% to 30% in persons with no complications or comorbidities to 56% to 71% in persons with 2 or more complications or comorbidities (Figure 2). Medication use varied by the number of complications present. Insulin use had a similar range in persons with no or 1 complication or comorbidity (24%-34%) but increased in those with 2 or more (40%-45%). Angiotensin-converting enzyme inhibitor use increased from 5% to 9% in those with no complications or comorbidities to 25% to 33% in those with 1 and 33% to 38% in those with 2 or more. We also examined utilization of services among persons with specific complications and comorbidities and found substantial variation (Table 4). Persons with renal disease or LED tended to have high rates of service utilization and, on average, more hospital stays, emergency department visits, and specialist visits than persons with cardiovascular disease (ie, hypertension, IHD, or cerebrovascular disease) or eye disease. However, persons with renal disease and those with LED had only modestly increased numbers of primary care visits ( times that of persons with no complications or comorbidities), and these rates were similar to those of persons with other complications. Patients with renal disease averaged 0.79 to 1.61 hospital stays, 1.05 to 1.30 emergency department visits, and 8.17 Figure 1. Age-Adjusted Percentage of Patients With at Least 1 Cardiovascular Comorbidity Who Had Additional Diabetic Complications Patients, % MCO1 MCO2 MCO3 Hypertension IHD Cerebro Cardiovascular Comorbidity Renal Disease Lower Extremity Disease Eye Disease MCO = managed care organization; IHD = ischemic heart disease; Cerebro = cerebrovascular disease. VOL. 7, NO. 5 THE AMERICAN JOURNAL OF MANAGED CARE 505

6 to specialist visits. These rates were 7 to 20 times, 6 to 11 times, and 8 to 11 times higher, respectively, than those in persons with no complications. For patients with LED, rates were 4 to 13 times, 3 to 7 times, and 5 to 7 times higher, respectively, than those in persons with no complications.... DISCUSSION... Using currently available computerized data maintained by 3 MCOs, we determined the prevalence of several complications and cardiovascular comorbidities in diabetic populations and examined patterns in the use of healthcare services among patients with these complications. Most patients had at least 1 of the 4 major complications examined (in order of prevalence): cardiovascular disease, eye disease, LED, and renal disease. Our estimates of cardiovascular disease prevalence in the 3 MCO populations are similar to those presented in a recent study 19 of the impact of cardiovascular disease on healthcare utilization in the diabetic population of an MCO. In that study, researchers reviewed medical charts and found that 57.9% of patients with diabetes mellitus have some type of cardiovascular disease, 4.2% have peripheral vascular disease, and 4.9% have cerebrovascular disease. In our study, cardiovascular disease prevalence ranged from 45% to 53%, peripheral vascular disease from 3% to 8%, and cerebrovascular disease from 3% to 6%. Our estimates for hypertension (31%- 42%) were somewhat lower than those in the other study (47%). Because hypertension was determined by International Classification of Diseases, Ninth Edition, codes, the diagnostic criteria may have varied across these MCOs, and this might account for the lower prevalence. We also found that the presence of 1 or more diabetic complications or cardiovascular comorbidities had a great impact on the level and type of resource utilization. With 1 complication, rates for hospital, specialist, and emergency department services all increased substantially and for primary care use increased modestly; when multiple complications were present, the rates increased even more. Use of medical services differed by type of complication present. Patients with renal disease had the highest rates of using hospital and emergency department services (6-20 times higher than patients having no complications) followed by patients with LED. Table 3. Age-Adjusted Mean Number per Person With Diabetes Mellitus and Relative Rate of Heathcare Service Use, by Type of Service and Number of Complications, Complication 2 or More Complications Type of Service MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 Hospital discharges* Age-adjusted mean Relative rate Emergency department visits Age-adjusted mean Relative rate Visits to specialists Age-adjusted mean Relative rate Visits to primary care Age-adjusted mean Relative rate MCO = managed care organization. *Overnight stay required. Relative to no complications. Restricted to the following specialists: endocrinologists, cardiologists, nephrologists, ophthalmologists/optometrists, orthopedic surgeons, and podiatrists. 506 THE AMERICAN JOURNAL OF MANAGED CARE MAY 2001

7 ... Diabetes Complications and Resource Utilization... Using administrative databases to estimate the prevalence of complications has limitations. By relying on diagnosis and procedural codes assigned during an encounter to measure complications, patients with complications that were not treated or patients who did not visit a provider for those conditions would be missed, resulting in a lower estimate than the true prevalence of that complication. On the other hand, screening tests used to rule out the presence of a condition might have resulted in the use of a diagnosis code and thus inflated the prevalence rates. However, validation studies 16 of administrative data to identify complications have found reasonable performance. One must also consider that using diagnostic codes and administrative data may likely detect patients who are more Figure 2. Age-Adjusted Percentage of Patients Who Received a Dilated Eye Examination by Number of Cardiovascular or Diabetic Complications or Comorbidities in the Managed Care Organization (MCO) Populations Patients, % MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 0 1 No. of Complications 2+ Table 4. Age-Adjusted Mean Number per Person With Diabetes Mellitus and Relative Rate of Healthcare Service Use, by Type of Service and Type of Complication,* 1993 Major Cardiovascular Diseases Renal Disease Eye Disease Lower Extremity Disease Type of Service MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 MCO1 MCO2 MCO3 Hospital discharges Age-adjusted mean Relative rate Emergency department visits Age-adjusted mean Relative rate Visits to specialists Age-adjusted mean Relative rate Visits to primary care Age-adjusted mean Relative rate MCO = managed care organization. *These categories are not mutually exclusive. Patients with a particular complication may (and usually do) have other complications. Overnight stay required. Relative to no complications. Restricted to the following specialists: endocrinologists, cardiologists, nephrologists, ophthalmologists/optometrists, orthopedic surgeons, and podiatrists. VOL. 7, NO. 5 THE AMERICAN JOURNAL OF MANAGED CARE 507

8 ill because they have had more encounters with the healthcare system. Another limitation is that because patients did not have to be enrolled in the MCO for the entire year to be included in the estimates (11% were enrolled for less than the entire year), true prevalence and utilization rates may have been slightly underestimated. A final issue to consider is that the prevalence of each complication was not uniform across the various plans, which can be attributed, at least in part, to differences in the severity of diabetes mellitus (duration of diabetes mellitus, risk factors, and comorbidities) in each plan s diabetic population. Although we adjusted for age differences between health plans, we were not able to adjust for risk factors related to the development of complications (such as glycemic control). 1,2 This issue must be considered when using health services research to determine the effectiveness of diabetes mellitus interventions. Although severity index measures exist for diabetes mellitus and other conditions, 20,21 they are not routinely available in administrative and other MCO data. In conclusion, many MCOs are or will be developing disease management programs and initiatives to improve the quality of care for patients with diabetes mellitus. Evaluating the health status and service utilization of these patients will be a critical first step for projecting future needs and resources for these efforts. As demonstrated in this study, administrative data can be used to determine and monitor the prevalence of diabetic complications and resource utilization during specific diabetes mellitus care initiatives, such as those recommended by the National Committee for Quality Assurance. Identification of these patterns will be useful to policy makers, healthcare providers, and quality improvement managers in developing research initiatives, new interventions, healthcare plans, and strategies for reducing the complications of diabetes mellitus and their associated human and economic burden.... REFERENCES The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulindependent diabetes mellitus. N Engl J Med 1993;329: The UK Prospective Diabetes Study (UKPDS) Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ 1998;317: The UK Prospective Diabetes Study (UKPDS) Group. 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 1998;352: Early Treatment Diabetic Retinopathy Study (ETDRS) Research Group. Early photocoagulation for diabetic retinopathy: ETDRS report number 9. Ophthalmology 1991;98: ETDRS Investigators. Aspirin effects on mortality and morbidity in patients with diabetes mellitus: Early Treatment Diabetic Retinopathy Study report 14. JAMA 1992;268: Ravid M, Lang R, Rachmani R, Lishner M. Long-term renoprotective effect of angiotensin-converting enzyme inhibition in non insulin-dependent diabetes mellitus: A 7-year follow-up study. Arch Intern Med 1996;156: Pyorala K, Pederson TR, Kjekshus J, Faergeman O, Olsson AG, Thorgeirsson G. Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease. Diabetes Care 1997;210: Hiss RG, Anderson RM, Hess GE, Stepien CJ, Davis WK. Community diabetes care: A 10-year perspective. Diabetes Care 1994;17: Peters AL, Legorreta AP, Ossorio RC, Davidson MB. Quality of outpatient care provided to diabetic patients: A health maintenance organization experience. Diabetes Care 1996;19: Beckles GL, Engelgau MM, Narayan KM, Herman WH, Aubert RE, Williamson DF. Population-based assessment of the level of care among adults with diabetes in the United States. Diabetes Care 1998;21: Brechner RJ, Cowie CC, Howie LJ, Herman WH, Will JC, Harris MI. Ophthalmic examination among adults with diagnosed diabetes mellitus. JAMA 1993;270: Engelgau MM, Geiss LS, Manninen DL, et al. Use of services by diabetes patients in managed care organizations: Development of a diabetes surveillance system. Diabetes Care 1998;21: Quickel KE Jr. Diabetes in a managed care system. Ann Intern Med 1996;124: National Committee for Quality Assurance Web site. Available at: http//: Accessed February 15, Schneider EC, Riehl V, Courte-Wienecke S, Eddy DM, Sennett C. Enhancing performance measurement: NCQA s road map for a health information framework. JAMA 1999;282: Newton KM, Wagner EH, Ramsey SD, et al. The use of automated data to identify complications and comorbidities of diabetes: A validation study. J Clin Epidemiol 1999;52: Ramsey SD, Newton K, Blough D, McCulloch DK, Reiber GE, Wagner EH. Incidence, outcomes and cost of foot ulcers in patients with diabetes. Diabetes Care 1999;22: Ramsey SD, Newton K, Blough D, McCulloch DK, Sandhu N, Wagner EH. Patient-level estimates of the cost of complications in diabetes in a managed-care population. PharmacoEcon 1999;16: Glauber H, Brown J. Impact of cardiovascular disease on health care utilization in a defined diabetic population. J Clin Epidemiol 1994;47: Greenfield S, Sullivan L, Dukes KA, Silliman R, Dagostino R, Kaplan SH. Development and testing of a new measurement of case mix for use in office practice. Med Care 1995;33:AS27-AS Berlowitz DR, Rosen AK, Moskowitz MA. Ambulatory care case-mix measures. J Gen Intern Med 1995;10: THE AMERICAN JOURNAL OF MANAGED CARE MAY 2001