TITLE: Adherence to Treatment and Social Support in Patients with Non-Insulin Dependant Diabetes. Garay-Sevilla, M.E.; Nava, L.E.; Malacara, J.M.; Huerta, R.; Díaz de León, J.; Mena, A.; Fajardo, M.E. Journal of Diabetes and Its Complications. 9(2): 81-86. April-June 1995.

OBJECTIVE: To evaluate factors associated with adherence to diet and medication in patients with type 2 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Adherence to diet was associated with years since diagnosis and greater social support; adherence to medication was associated with greater social support and older age of spouse.

RECOMMENDATION: Further research should be conducted to identify psychological and social factors that influence adherence to diet and medication in patients with type 2 diabetes.

ABSTRACT: The authors studied 200 adults with type 2 diabetes recruited from diabetes support groups at two hospitals in Leon, Mexico. Only about 20 percent of patients invited actually participated in the study. These patients completed validated questionnaires detailing adherence to treatment, social support, lifestyle, family structure, family function, and knowledge of diabetes. Adherence to diet was associated with years since diagnosis (p = 0.003) and greater social support (p = 0.007); adherence to medication was associated with greater social support (p = 0.002) and older age of spouse (p = 0.016). Patients with rigid control in their families had lower adherence to medication than patients where families had laissez-faire or flexible control. Evaluation of the psychological and social factors that influence adherence to diet and medication is essential because of their important role in the management of patients with type 2 diabetes. 1 figure, 3 tables, 18 references.

TITLE: Analysis of Direct Cost of Standard Compared with Intensive Insulin Treatment of Insulin-Dependent Diabetes Mellitus and Cost of Complications. Stern, Z.; Levy, R. Acta Diabetologica. 33(1): 48-52. March 1996.

OBJECTIVE: To compare the direct costs of standard and intensive insulin treatment for type 1 diabetes; to compare these approaches when their impact on complications is considered.

CATEGORY: Secondary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: According to the authors' model, standard insulin therapy was less expensive over a 35-year period than was intensive insulin therapy, whether or not the cost of complications was considered.

RECOMMENDATION: Because of the high direct costs of intensive treatment, guidelines should be developed to better identify, select, and treat those patients for whom such therapy is warranted. Intensive treatment should be aimed particularly at patients with nephropathy.

ABSTRACT: The authors modeled annual treatment costs over a 35-year period for a hypothetical patient who had contracted type 1 diabetes at a young age. Results in the literature, including findings of the Diabetes Control and Complications Trial, were used in the model, as was standard practice in Israel, which follows the recommendations of the American Diabetes Association. Annual direct costs (in 1995 dollars) for standard and intensive insulin treatment were calculated to be $1,184 and $3,329 per patient, respectively. Over a 35-year period, total direct costs per patient were $41,000 and $116,000 for the standard and intensive treatments, respectively, a difference of $75,000. The authors found that intensified treatment lowered complication costs by $53,520 (versus standard treatment). Thus, total direct costs of standard therapy were $329,400, versus $350,980 for the intensified treatment. Assuming a discount rate of 6 percent, it was estimated that the intensive treatment entailed lower complication costs (by $20,900) than the standard treatment, but the total cost of the standard treatment was $132,900, compared with $151,900 for the intensive treatment. 3 tables, 16 references.

TITLE: Assessment of the Effect of a Comprehensive Diabetes Management Program on Hospital Admission Rates of Children with Diabetes Mellitus. Drozda, D.J.; Dawson, V.A.; Long, D.J.; Freson, L.S.; Sperling, M.A. Diabetes Educator. 16(5): 389-393. September-October 1990.

OBJECTIVE: To determine the impact of a comprehensive diabetes management program on hospitalization parameters in children with diabetes seen at a major pediatric referral center.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Care provided through the diabetes management program is expected to save $342,000 annually in direct health service costs.

RECOMMENDATION: Comparison of data from pediatric hospitals not offering comprehensive diabetes programs would help to assess the impact of such programs.

ABSTRACT: The authors evaluated the impact of a comprehensive diabetes management program initiated in July 1978 on admission of children with diabetes mellitus to Children's Hospital Medical Center in Cincinnati, Ohio. Those in the study had a primary diagnosis of type 1 diabetes; the main reason for admission was diabetic ketoacidosis, hyperglycemia, or hypoglycemia. Admission records from January 1973 through December 1987 were reviewed; comparisons were made of admission parameters for January 1973 to June 1978 (period A) and July 1978 to December 1987 (period B). The program included medical and support services and individualized educational interventions; two telephone hot lines were provided as well. Admissions for type 1 diabetes not associated with diabetic ketoacidosis or other diagnoses rose from 27 percent of all admissions in period A to 37 percent in period B (p = 0.01). The proportion of admissions for diabetes that were for diabetic ketoacidosis not associated with other diagnoses decreased from 63 percent in period A to 29 percent in period B (p = 0.0001). This positive outcome may have reflected the effect of the education program on patient self-management. Mean length of stay for admissions for diabetic ketoacidosis only decreased from 5.8 days (period A) to 4.6 days (period B) (p = 0.01), but the introduction of managed care may have encouraged early discharge. 2 figures, 2 tables, 13 references.

TITLE: Audit in General Practice by a Receptionist: A Feasibility Study. Essex, B.; Bate, J. British Medical Journal. 302(6776): 573-576. March 9, 1991.

OBJECTIVE: To assess whether a medical practice audit, including care of patients with diabetes, can be performed cost effectively by a practice's receptionist.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: The practice audit was performed by the receptionist in a cost-effective manner.

RECOMMENDATION: Audit by a practice's receptionist should be considered a cost-effective alternative to audit by the physicians themselves.

ABSTRACT: The authors assessed the feasibility and cost-effectiveness of having a medical practice audit performed by a practice receptionist. The practice was composed of 6 physicians in London, England, treating 11,500 patients. A system developed to allow the receptionist to audit medical records was evaluated over a 2-year period. Patients with diabetes were identified from disease registers; most data were derived from patient records. A total of 136 patients with diabetes were identified by the receptionist and their records reviewed. The receptionist provided the practice with a breakdown of the level of care (general practitioner only, hospital only, shared, or unknown). The receptionist notified practice physicians of any patients with diabetes under the care of a general practitioner who had not had an annual review. The receptionist spent about 4 hours weekly performing the audit as part of her general duties at a cost of £ 5.20 per hour ( £ 960 per year). Regular supervision of the receptionist the first year lasted about 30 minutes weekly; in the second year, about 30 minutes every 2 weeks. After deduction of reimbursements and tax, the cost came to £ 30 per practice physician per year. This system was extremely cost-effective compared with the costs that would be incurred were physicians to perform the audit themselves. 4 figures, 10 references.

TITLE: Bedside Blood Sugar Determinations in the Critically Ill. Newman, R.H. Heart and Lung. 17(6 Part 1): 667669. November 1988.

OBJECTIVE: To compare various glucose monitoring systems with standard laboratory testing in terms of accuracy and cost-effectiveness.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: Results for a bedside system called the Glucoscan 2000 were highly correlated with hospital laboratory results, but the bedside system took less time (mean: 2.4 minutes versus 2.6 hours) and was significantly less expensive (for 110 determinations, $55 in costs versus $990 in charges).

RECOMMENDATION: Glucose monitoring systems cannot and should not replace laboratory glucose determinations, but they can greatly reduce their frequency and supplement expensive laboratory data. A quality control regimen must be implemented for the selected glucose monitoring device to ensure its accuracy.

ABSTRACT: Intensive care unit patients may have large fluctuations in blood glucose concentration; accurate and timely glucose values must be obtained so that these fluctuations may be stabilized. The author compared three blood glucose monitoring systems (AccuCheck, Glucometer, Glucoscan 2000) with the hospital laboratory; there were no significant differences in results between the systems. In a subsequent 3month study, the Glucoscan 2000 and hospital laboratory performed 110 blood glucose determinations in 41 patients; for the first 50 blood glucose tests, mean completion time was 2.4 minutes for the Glucoscan 2000 and 2.6 hours for the hospital (routine) or 0.5 hours ("stat"). Laboratory charges for 110 laboratory determinations totaled $990, versus Glucoscan costs of $55. The author concludes that blood glucose monitoring systems can be effective for close monitoring of blood glucose in the intensive care unit or elsewhere. The purchase price of the Glucoscan 2000 was $127. 1 figure, 2 tables, 16 references.

TITLE: Cardiovascular Morbidity and Mortality in Type 2 Diabetic Patients: A 22 Year Historical Cohort Study in Dutch General Practice. de Grauw, W.J.C.; van de Lisdonk, E.H.; van den Hoogen, H.J.M.; van Weel, C. Diabetic Medicine. 12(2): 117122. February 1995.

OBJECTIVE: To assess the impact of cardiovascular morbidity and mortality on patients in general practice with type 2 diabetes over a 22year period.

CATEGORY: Secondary intervention.

Type of Study: Historical cohort.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Patients in general practice with type 2 diabetes were at higher risk of cardiovascular morbidity, mortality from any cause, and cardiovascular mortality than were those without diabetes.

RECOMMENDATION: None.

ABSTRACT: An historical cohort study was performed to measure cardiovascular morbidity and mortality in patients with type 2 diabetes; data were collected from 1967 to 1989 in four Dutch general practices involved in the Continuous Morbidity Registration Nijmegen. A total of 265 patients with type 2 diabetes (112 men, 153 women) who met World Health Organization (WHO) criteria were included in the cohort. The remaining 162 patients who were registered during the study period as diagnosed with diabetes were not included because they did not meet WHO criteria or because of other reasons. At diagnosis, 60 percent of the study cohort were aged 65 or under; mean follow-up was 6.8 years. Compared with a matched control group of persons without diabetes, those with type 2 diabetes had higher cardiovascular morbidity (risk ratio, 1.76; 95 percent confidence interval, 1.34 to 2.30). The risk of mortality was also higher in patients with diabetes than in controls (relative risk, 1.54; 95 percent confidence interval, 1.07 to 2.23). Relative risk of cardiovascular mortality was 2.05 in patients with diabetes (95 percent confidence interval, 1.24 to 3.37). Mortality after 10 years for patients with type 2 diabetes was 36 percent, versus 20 percent for the control group (p < 0.01). Cumulative survival rates in the group aged 65 to 74 years were significantly lower (p < 0.01) in patients with type 2 diabetes than in controls. 2 figures, 3 tables, 32 references.

TITLE: Care of the Diabetic Child in the Community. Farquhar, J.W.; Campbell, M.L. British Medical Journal. 281(6254): 1534-1537. December 6, 1980.

OBJECTIVE: To describe community care for children with newly diagnosed type 1 diabetes, and to describe in-hospital care for such children.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Descriptive analysis.
Perspective: Health care system.

CONCLUSION: Admitting children with newly diagnosed diabetes to the hospital, then providing them with a home care team, offers numerous advantages for the children and their parents.

RECOMMENDATION: None.

ABSTRACT: The authors describe experience with children newly diagnosed with diabetes who have been admitted to the Royal Hospital for Sick Children in Edinburgh, Scotland. Admitting to the hospital a child newly diagnosed with diabetes along with one or both parents allows parents to develop trust in the staff, to learn more effectively about their child, and to meet the home care sisters. Staff can obtain an indication of parental ability and stability, and an insulin reaction can be induced safely so that the parent can watch the child's reaction to hypoglycemia. The child is normally discharged within 1 week. Having a home care team enabled the hospital to have an average length of stay of about 4.5 days in 1976, versus the national average of 8 days. The home care team permits reduced readmissions, relieves maternal and school teacher anxiety, and decreases school absences. The home care team continually assesses the home — its organization, finances, presence of alcoholism, stability of the parents and marriage, relationships of parents and children and relationships between children, how procedures are conducted, and the extent to which parents seek further education toward controlling diabetes. The home care team visits the school and briefs staff and the school nurse on symptoms and treatment of hypoglycemia and the needs of the child. The team provides continuing care through phone consultations, visits the home when the children are in school, and obtains spot urine specimens or blood checks if there is a question of diabetic control. To avoid disharmony between the hospital clinic and the primary care and community health services, each important contact with the home care team or the unit staff should be recorded. 4 references.

TITLE: A Comparison of Accuracy and Estimated Cost of Methods for Home Blood Glucose Monitoring. Shapiro, B.; Savage, P.J.; Lomatch, D.; Gniadek, T.; Forbes, R.; Mitchell, R.; Hein, K.; Starr, R.; Nutter, M.; Scherdt, B. Diabetes Care. 4(3): 396-403. May-June 1981.

OBJECTIVE: To compare the accuracy, convenience, and cost of various methods of home blood glucose monitoring.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: All the capillary blood glucose methods correlated well with laboratory blood glucose measurements and all were more accurate than urine glucose measurements. The Chemstrip bG was the least expensive and easiest method to measure blood glucose concentrations at home.

RECOMMENDATION: None.

ABSTRACT: The authors compared accuracy, convenience, and cost of four products for home blood glucose monitoring. Sixteen patients used Eyetone, Dextrometer, StatTek (all of which used a reflectance meter and a reagent strip), and Chemstrip bG (reagent strip only) twice daily within 1 minute of collection of venous blood for laboratory blood glucose measurement. Urine glucose was measured from collections made before and after blood collection. For laboratory blood glucose concentrations of 48 to 464 mg/dL and 48 to 250 mg/dL, correlation coefficients for Eyetone, Dextrometer, and StatTek reflectance meter readings (by a physician or nurse) and for Chemstrip bG readings by the patient, a nurse, and a physician ranged from 0.90 to 0.95 and from 0.85 to 0.92, respectively (p < 0.0001). All home blood glucose monitoring methods except StatTek underestimated blood glucose concentrations between 48 and 250 mg/dL. The correlation coefficient for urine glucose concentrations was 0.74. When Chemstrip bG was reread at intervals over 14 days, glucose readings decreased by 6.8 to 10.1 percent (not clinically significant) in the first 2 days; no further decrease was noted. The coefficient of variation (CV) for repeated readings of the three reflectance meters was 8.9 to 12.0 percent for nurses and 6.7 to 9.0 percent for physicians; the CV for Chemstrip bG was 17.8 and 11.6 percent, respectively. The costs (1980 dollars) per glucose determination (excluding cost of reflectance meter) for Dextrometer, Eyetone, StatTek, and Chemstrip were $0.71 to $1.32, $0.51 to $0.71, $0.72 to $1.01, and $0.53, respectively, and reflectance meters for the first three cost between $339.95 and $395.00. The Chemstrip bG method is easy to use and less expensive than the other systems, and readings can be verified by a physician up to 2 weeks later. 3 figures, 5 tables, 26 references.

TITLE: Comparison of Different Models of Diabetes Care on Compliance with Self-Monitoring of Blood Glucose by Memory Glucometer. Hoskins, P.L.; Alford, J.B.; Handelsman, D.J.; Yue, D.K.; Turtle, J.R. Diabetes Care. 11(9): 719-724. October 1988.

OBJECTIVE: To evaluate the accuracy of patient-generated records of blood glucose concentrations, and to determine whether the model of diabetes care influences the validity of these patient records.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: There was no difference in validity of reported blood glucose results based on instruction technique, but private health insurance status was associated with greater validity of patient records.

RECOMMENDATION: Health professionals should not assume that patients will follow treatment instructions because they have been fully counseled.

ABSTRACT: The authors evaluated 34 patients with diabetes on their self-monitoring of blood glucose. Participants had been referred to the diabetes center of Royal Prince Alfred Hospital in Camperdown, Australia; 17 had gestational diabetes; 13 had type 2 diabetes; and 4 had type 1 diabetes. Participants were randomized to one of three models of instruction: mutual decision making (emphasized team approach; patient could adjust treatment based on results), didactic (nurse educator would make decisions based on results), and authoritarian (physician would make decisions based on results). Participants performed blood glucose estimations four times daily over 2 weeks using a blood glucose monitor with memory but were not informed of the memory capacity. Blood glucose results recorded in patient logbooks were compared with monitor records. No difference was found in the validity of patients' records by instruction group. The overall proportion of correctly recorded estimations was 86 percent; 70 percent of examinations were actually performed as requested and correctly recorded, and another 16 percent were correctly recorded as not done. Conversely, 30 percent of scheduled examinations were not performed properly or were recorded incorrectly. Compared with patients with type 2 diabetes, patients with gestational diabetes recorded a lower percentage of blood glucose estimations accurately. Private health insurance was one of two predictors of greater validity of blood glucose records (the other was perceived high intelligence). Although there were no direct financial disincentives for compliance with testing, as patients were not required to copay, health insurance status may represent a surrogate socioeconomic indicator. Some patients may have experienced financial or social pressures that act as a barrier to successful completion of diabetes self-care activities. 6 tables, 8 references.

TITLE: Continuous Quality Improvement Can Improve Glycemic Control for HMO Patients with Diabetes. O'Connor, P.J.; Rush, W.A.; Peterson, J.; Morben, P.; Cherney, L.; Keogh, C.; Lasch, S. Archives of Family Medicine. 5(9): 502-506. October 1996.

OBJECTIVE: To evaluate the impact of a continuous quality improvement (CQI) intervention on glycemic control of patients with diabetes mellitus attending a primary care clinic.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Use of a CQI intervention can improve glycemic control without increasing utilization or charges.

RECOMMENDATION: Health care providers should consider implementing CQI to improve adult diabetes care; several of the strategies developed by the CQI team may be useful for other primary care settings.

ABSTRACT: The authors analyzed the effect of implementing a CQI plan on glycemic control in patients with diabetes. Patients were followed for 18 months at two clinics of a Midwest staff model health maintenance organization. The intervention clinic had 4,100 patients; the comparison clinic, 4,700 patients. The staff at the intervention clinic developed a computerized patient audit system and held monthly 1-hour CQI meetings. The patient audit and review of the process of care led to a modified system using computerized printouts, targeting certain patients for special attention, and authorizing more aggressive outreach by resource nurses. Patients were selected for readiness to undertake behavioral change and given appropriate levels of support. Success of management was based on change in glycosylated hemoglobin (HbA_1C) levels. After intervention, in the CQI and comparison clinics, respectively, (HbA_1C levels were acceptable, # 8 percent) in 51 versus 40 percent of patients, fair (8 to 10 percent) in 37 versus 33 percent, and poor ($ 10 percent) in 12 percent versus 27 percent (p = .008). Intervention clinic patients had significant improvement in HbA_1C levels relative to patients at the comparison clinic (p = .01). Mean outpatient visits at the intervention and the comparison clinics were 7.86 and 7.40 prior to and 9.08 and 8.96 after initiation of CQI, respectively. Mean charges for outpatient services were $991 and $958 prior to and $1,218 and $1,281 after initiation of CQI, respectively. 1 table, 25 references.

TITLE: A Cost-Benefit Analysis of Subcutaneous Insulin Pump Infusion Therapy for Insulin-Dependent Diabetes Mellitus. Turkelson, C.M.; Coates, V. Abstract 195. Annual Meeting of the International Society for Technology Assessment in Health Care. 11. 1995.

OBJECTIVE: To compare the costs of subcutaneous insulin pump infusion therapy and conventional therapy for type 1 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Health care system.

CONCLUSION: Insulin pump therapy may be more cost-beneficial than conventional therapy, but it may take 9 years of treatment for this to be true. How many patients will remain on pump therapy for this long is not known.

RECOMMENDATION: The conclusions of this study and other cost analyses of insulin pump therapy must be treated cautiously.

ABSTRACT: The authors used Markov analysis to compare the costs of insulin pump and conventional therapy. Separate models were used to determine treatment-specific costs associated with diabetic nephropathy, retinopathy, and amputations. When just retinopathy and amputations were considered, insulin pump therapy did not provide as great a cost-benefit as conventional therapy. However, the savings associated with the third complication (diabetic nephropathy) were large enough for a greater cost-benefit to be realized for insulin pump therapy after 9 years of treatment when all three complications were considered simultaneously. The authors note that the proportion of patients who should be given this type of therapy (or any form of intensive therapy) is unknown, as is the effectiveness of this therapy under regular clinical practice conditions.

TITLE: Cost-Effectiveness Analyses of the Conversion of Patients with Non-Insulin-Dependent Diabetes Mellitus from Glipizide to Glyburide and of the Accompanying Pharmacy Follow-up Clinic. Law, A.V.; Pathak, D.S.; Segraves, A.M.; Weinstein, C.R.; Arneson, W.H. Clinical Therapeutics. 17(5): 977987. September/October 1995.

OBJECTIVE: To retrospectively evaluate the cost effectiveness of converting from glipizide therapy to glyburide therapy and of using a pharmacy follow-up clinic.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: The conversion from glipizide to glyburide was cost effective, as was the follow-up clinic. For every 1 percent of patients in the follow-up group who achieved good glycemic control, the Department of Veterans Affairs would spend only $1.01 more per patient on costs related to the follow-up clinic.

RECOMMENDATION: Due to its cost-effectiveness, the conversion from glipizide to glyburide is recommended in practice settings similar to that of the Department of Veterans Affairs. Follow-up clinics are justifiable because they are cost effective and possibly increase patient compliance with the medication regimen or with nondrug factors (e.g., diet, exercise).

ABSTRACT: During a 6month period in 1993, the Department of Veterans Affairs Outpatient Clinic in Columbus, Ohio, converted patients with type 2 diabetes from glipizide to glyburide therapy as a way to reduce costs. For this retrospective study, patients were divided into groups: I included all 730 patients while they were being treated with glipizide; II, 408 patients who converted to glyburide and returned to the follow-up clinic; III, 244 patients who converted to glyburide but did not return to the follow-up clinic (records were missing for the other 78 patients). Cost data were gathered by examining the direct costs (in 1994 dollars) of the conversion and the follow-up clinic. Costs common to all three groups were not included in the analysis. The effectiveness measure was defined as the percentage of patients whose glycemic control was rated as good (i.e., fasting blood glucose 200 mg/dL or less). The authors found that the average mean daily dose for glipizide was 1.82 to 1.85 times that for the glyburide dose. Based on a daily dose of two 10mg tablets of glipizide (Group I) and two 5mg tablets of glyburide (Groups II and III), the cost savings to the Department of Veterans Affairs for the first year after conversion were $104,974.12 for Group II and $132,132.12 for Group III. Although Group III achieved higher cost savings, Group II had a higher percentage of patients who achieved good glycemic control. 5 tables, 14 references.

TITLE: Cost-Effectiveness Study of a Lipid-Lowering Therapy in Hyperlipoproteinaemia Type IIb and Type IV (Frederickson). Bergemann, R.; Brandt, A.; Siegrist, W. PharmacoEconomics. 3(Supplement 2): 131-139. February 1993.

OBJECTIVE: To conduct a cost-effectiveness analysis of four drugs (acipimox, bezafibrate, fenofibrate, and gemfibrozil) for treating patients with hyperlipoproteinemia types IIb and type IV (according to Frederickson), with and without type 2 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: When hospitalization costs for treating gallstones are taken into account, acipimox is a more cost-effective therapy than the other three drugs.

RECOMMENDATION: None.

ABSTRACT: In this Swiss study, a computer model was developed of hyperlipoproteinemia types IIb and IV and their treatment. The study was based on up to 7 years of therapy; in 10.5 percent to 11.8 percent of patients, according to the model, the therapy was interrupted because of adverse events or inadequate therapeutic effect. Only direct costs were included (for drugs, physician visits, and hospitalizations for coronary vascular disease, coronary artery bypass graft surgery, peripheral arterial occlusive disease, and gallstone operations). Calculations were based on computer-aided medical decision analysis and the Monte Carlo method; the discount rate was 5 percent per year. Results from various epidemiologic and clinical trials were used to compare long-term outcomes. Efficacy studies with long-term endpoints are available for the three fibrate drugs; the authors used various studies to derive estimates for the effects of acipimox. Yearly costs of medication for the four drugs ranged from 434 to 446 Swiss francs. Annual treatment costs with the three fibrate drugs were about 100 Swiss francs more than with acipimox for both kinds of hyperlipoproteinemia, with or without type 2 diabetes. The risk of gallstones with the fibrate drugs was more than three times that for acipimox; this difference, according to the authors, explained the marked differences in the cost of therapy. The authors state that estimated savings with acipimox were conservative because only the cost of the hospital stay was included; additional expenses for diagnosing and treating gallstones as well as for lost productivity were not included. They note that the dosages (mg/day) in the study were specific to Switzerland: 500 for acipimox; 400, bezafibrate; 300, fenofibrate; and 900, gemfibrozil. 1 figure, 7 tables, 1 appendix, 38 references.

TITLE: The Costs and Effects of Two Different Lipid Intervention Programmes in Primary Health Care. Tomson, Y.; Johannesson, M.; Åberg, H. Journal of Internal Medicine. 237(1): 1317. January 1995.

OBJECTIVE: To compare the costs and effects of two different intervention strategies for the nonpharmacological treatment of hypercholesterolemia.

CATEGORY: Secondary intervention.

Type of Study: Other prospective trial.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: There was no difference in the effectiveness of the two intervention strategies.

RECOMMENDATION: Of the two strategies, the low-intensity program is preferable from a Cost-effectiveness viewpoint.

ABSTRACT: All persons aged 25 to 54 years who resided in the catchment area of a health center south of Stockholm were invited to a health check; 2,338 were screened for hypercholesterolemia. From a group of 372 persons found to have this disorder, the authors selected men and women with a serum cholesterol in the range of 7.07.8 mmol L-1. Those with hypertension, diabetes mellitus, and ischemic heart disease were excluded. The 92 persons who met all criteria were randomized to low-intensity (n = 45) or moderate-intensity (n = 47) intervention programs. The low-intensity program focused on diet modification; participants were given brochures. The moderate-intensity program, which followed Swedish national guidelines, also focused on diet modification but included more active involvement with the participant's general practitioner and with a dietitian. Both groups were followed for a year and compared with respect to serum cholesterol concentration and treatment costs in Swedish crowns (SEK), using 1993 prices. (Apart from the health care costs, the treatment costs for the patient included the time cost for the visits and travel costs to the health care center.) Both intervention strategies resulted in small decreases (3.3 percent for low intensity, 3.7 percent for moderate intensity) in total cholesterol, with no significant difference between the groups. The total cost per patient in the low-intensity group was SEK 753; in the medium-intensity group, SEK 3,614. The authors estimate that if only 30 percent of the population aged 25 to 54 years in Stockholm county with hypercholesterolemia were discovered by the primary health care system and followed the dietary advice, the net savings with the low-intensity model (versus the moderate-intensity model) would be SEK 93 million. 5 tables, 10 references.

TITLE: The Development of Community Orientated Recommendations for Diabetes Care in South Auckland. Wilson, P.; Simmons, D. New Zealand Medical Journal. 107(989): 456-459. November 9, 1994.

OBJECTIVE: To describe the development of recommendations for preventing and controlling diabetes and its effects in South Auckland, New Zealand, a community in which diabetes is a major public health priority.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Historical review.
Perspective: Societal.

CONCLUSION: The provision of scientifically derived data, followed by discussion of the data with community representatives and its subsequent dissemination throughout the community, has provided an opportunity to begin controlling diabetes in the South Auckland area.

RECOMMENDATION: Numerous recommendations are set forth in the body of the report.

ABSTRACT: South Auckland, New Zealand, a community with considerable socioeconomic deprivation, has many residents of Maori and Pacific Islander descent, two groups with a disproportionate risk for diabetes. In 1990-1991 the South Auckland Diabetes Survey canvassed all general practitioners and practice nurses in that district on diabetes management and barriers to care. Community experience was obtained from existing diabetes focus groups. Patients and selected general practitioners were interviewed about barriers to care. Additional interviews were conducted with physicians, dietitians, and others involved in diabetes care, and a 54-page preliminary information document was prepared. A 12-member planning group found the document to be too long and technical; a 4-page summary was prepared and distributed to a variety of groups. The next draft followed 12 meetings and extensive community consultation. Sixty-eight recommendations were eventually made on diabetes care, including altering the environment, coordination and standardization, access, cost reduction, diabetes support groups, foot and eye care, diabetes in pregnancy and impaired glucose tolerance, and screening. A cost-outcome evaluation of the plan suggested significant savings in direct health expenditure, and full funding was subsequently recommended. 1 table, 1 figure, 22 references.

TITLE: Diabetes in Urban African Americans: III. Management of Type II Diabetes in a Municipal Hospital Setting. Ziemer, D.C.; Goldschmid, M.G.; Musey, V.C.; Domin, W.S.; Thule, P.M.; Gallina, D.L.; Phillips, L.S. American Journal of Medicine. 101(1): 25-33. July 1996.

OBJECTIVE: To prospectively study the effectiveness of multidisciplinary, nonphysician, nonpharmacologic management in an urban outpatient setting of a largely African American group of patients with type 2 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Nonpharmacologic, nonphysician management of type 2 diabetes is effective in a socioeconomically disadvantaged population.

RECOMMENDATION: Further study is needed to identify risk factors for patient loss to follow-up.

ABSTRACT: The authors prospectively evaluated care provided to patients with type 2 diabetes in a municipal hospital clinic in Atlanta, Georgia; most study patients were African Americans and socioeconomically disadvantaged. At an initial visit, the diagnosis was confirmed, family history obtained, patient education initiated, and a nurse practitioner or registered nurse assigned for follow-up; obese patients were instructed in a hypocaloric diet. Drug therapy was initiated, adjusted, or eliminated based on fasting plasma glucose levels at each visit. Data were analyzed for 325 patients who returned for follow-up visits at 2, 4, 6, and 12 months. Patients were more likely to return if diagnosed less than 1 year previously, over age 55 years, or female; 80 percent missed at least 1 visit. Compared with initial levels, both fasting plasma glucose and glycosylated hemoglobin improved significantly in the first 2 months (p < .001); levels were somewhat higher at 12 months. Initially, 36 percent of obese patients were managed by diet only, 24 percent by sulfonylureas, and 40 percent by insulin. After 2 months, 70 percent were managed by diet alone; at 6 months, 61 percent; and at 12 months, 55 percent. Lean patients had little change in management over the 12 months. Neither weight gain nor loss correlated with metabolic control at 12 months. Keeping appointments correlated with better metabolic control. Average cost per patient visit for physician and staff salaries, rent, utilities, and administrative overhead was $59 (1994 costs). The program was effective for a socioeconomically disadvantaged population, but further study is needed to identify risk factors for patient loss to follow-up. 7 figures, 1 table, 37 references.

TITLE: Effect of Value-Added Utilities on Prescription Refill Compliance and Medicaid Health Care Expenditures: A Study of Patients with Non-Insulin Dependant Diabetes Mellitus. Skaer, T.L.; Sclar, D.A.; Markowski, D.J.; Won, J.K. Journal of Clinical Pharmacy and Therapeutics. 18(4): 295-299. August 1993.

OBJECTIVE: To determine the effect of mailed prescription-refill reminders, specialized packaging, or a combination of both interventions on prescription refill compliance and health service utilization among patients with type 2 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Randomized trial.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Mailed prescription refill reminders, specialized packaging, or a combination of both significantly increased patient refill compliance. Patients receiving both interventions also had decreased expenses for physician, laboratory, and hospital services.

RECOMMENDATION: Pharmacy-based value-added utilities should be used more under both public and private health insurance programs.

ABSTRACT: From 10 to 30 percent of patients with type 2 diabetes withdraw from prescribed treatment within 1 year of diagnosis; of the remainder, about 20 percent do not take enough prescribed medication to maintain their blood glucose at satisfactory concentrations. A randomized trial was undertaken to determine the effects of valued-added pharmacy services on the compliance of type 2 patients with prescription refills and their use of health services. The 1-year study included 258 Medicaid beneficiaries with previously untreated type 2 diabetes. Patients lived in South Carolina, were less than 65 years old, had prescriptions to administer 5 mg glyburide twice daily, and were not to have received an alternative sulfonylurea or used insulin after receipt of the initial glyburide prescription. The patients were assigned to one of four groups: a control group (standard pharmacy care), standard pharmacy care plus mailing of a reminder 10 days before the refill date, standard pharmacy care plus unit-of-use packaging, and standard pharmacy care plus both mailed refill reminders and unit-of-use packaging. The medication possession ratio (MPR) was used to compare groups; the MPR is the days' supply of medication obtained by the patient relative to the days in the study period (optimal = 1.00). Dispensing of drugs was authorized in 30-day supplies. Analysis of data after 1 year revealed that patients receiving mailed prescription refill reminders, unit-of-use packaging, or a combination of both achieved a significant (p < 0.05) increase in the MPR for sulfonylurea therapy relative to the controls. The MPR for those receiving both interventions was significantly higher than for the groups receiving one intervention. Multivariate regression analysis revealed that patients receiving both interventions had significant (p < 0.05) reductions in expenditures for physician, laboratory, and hospital services relative to the control group. 2 tables, 22 references.

TITLE: Efficacy of Insulin and Sulfonylurea Combination Therapy in Type II Diabetes. A Meta-Analysis of the Randomized Placebo-Controlled Trials. Johnson, J.L.; Wolf, S.L.; Kabadi, U.M. Archives of Internal Medicine. 156(3): 259-264. February 12, 1996.

OBJECTIVE: To evaluate the efficacy of combination therapy with insulin and sulfonylurea in type 2 diabetes by performing a meta-analysis of studies that met strict criteria.

CATEGORY: Secondary intervention.

Type of Study: Formal meta-analysis of randomized clinical trials.
Methodology: Data analysis.
Perspective: Health care system

CONCLUSION: Combination therapy improves glycemic control while lowering daily insulin dosages and without adversely affecting body weight. In developing countries, combination therapy is far less expensive than insulin monotherapy.

RECOMMENDATION: None.

ABSTRACT: Forty-three citations were obtained from a search of the MEDLINE database from January 1980 through March 1982; 16 studies with a total population of 351 participants were analyzed — these studies met strict criteria for study design and outcome measures. The meta-analysis showed that the combination of insulin and sulfonylurea improves metabolic control for periods of at least 16 weeks. This improved metabolic control is achieved despite a decrease in daily insulin dose and does not result in a significant change in body weight. Combination therapy appears to be more expensive than mono-therapy with insulin, as the total monthly cost for combination therapy is about $60 to $85, versus $20 for two daily injections of a total dose of 100 U of intermediate-acting insulin. However, for patients using insulin monotherapy, the costs of insulin administration are higher, more frequent glucose monitoring may increase expense, and more frequent office and emergency department visits may be needed. In addition, money could be saved by substituting a first-generation sulfonylurea such as tolazamide for a second-generation drug. However, second-generation sulfonylureas are dropping in price. In developing countries, the cost of sulfonylureas is minuscule compared with insulin. 7 tables, 61 references.

TITLE: Evaluation of a Practice-Based Programme of Health Checks: Financial Cost and Success at Risk Detection. Sacks, G.; Marsden, R. Journal of the Royal College of General Practitioners. 39(326): 369-372. September 1989.

OBJECTIVE: To evaluate the efficacy and cost of a screening program for cardiovascular risk factors among patients in a large group practice in Oxford, England.

CATEGORY: Secondary intervention.

Type of Study: Patient screening.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: The screening program increased recording of risk factors but its cost was not fully met by reimbursement for salaries or services.

RECOMMENDATION: To avoid the financial disincentive to general practitioners to provide such preventive medicine programs, salaries for program staff should be completely reimbursed by the government.

ABSTRACT: The authors evaluated the efficacy and cost of a screening program to identify patients at risk for cardiovascular disease in a large suburban group practice in Oxford, England. Costs for the unfunded program were to be recouped through fees generated by referrals for tetanus and polio shots and cervical cytology exams. Of 1,470 patients invited to participate in the nurse-run health checks, 1,382 (94.0 percent) participated. In this group, 66 cases of previously undiagnosed hypertension and 11 cases of previously undiagnosed diabetes were identified. Just over one-third (34.9 percent) of patients were overweight and 27.0 percent smoked. Total cost of the program was £ 11,785, of which £ 11,510 was for staff salaries. Estimated service fees generated totaled only £ 1,167, which was adjusted to £ 1,342 to reflect patients who had died or left the area. With £ 8,057 of income added to this amount from 70 percent reimbursement of salaries, the screening program operated at a net cost to the practice of £ 2,386, or £ 1.73 per patient. This outcome casts doubt on the results of other studies that have suggested that the costs of government-recommended prevention programs can be fully recouped through item-of-service payments. 1 figure, 2 tables, 14 references.

TITLE: The Feasibility of a Potentially "Ideal" System of Integrated Diabetes Care and Education Based on a Day Centre. Day, J.L.; Johnson, P.; Rayman, G.; Walker, R. Diabetic Medicine. 5(1): 70-75. January 1988.

OBJECTIVE: To evaluate the feasibility of a new system of clinical and educational care designed to replace traditional diabetes clinics.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: The new integrated diabetes care system improved quality of care of patients without incurring major costs.

RECOMMENDATION: None.

ABSTRACT: The authors evaluated a new, integrated diabetes care and education program in Ipswich, England, at a new day facility. Staff from a traditional hospital clinic were reorganized and the role of the specialist nurse changed to providing primary counseling (physicians also provided care in the new center). Clinic hours were revised and a new appointment system created. After 1 year, the new program was found to have provided a pleasant environment for diabetes care and education, largely eliminated excessive waiting time, established continuity in consultation, and made available a more comprehensive and systematized educational program. Mean glycosylated hemoglobin levels of patients did not differ significantly by whether they saw the physician or specialist nurse. These results were achieved without major additional expenses: building maintenance, £ 3,000; nursing staff, £ 2,275; phlebotomist services, £ 300; and receptionist services, £ 2,476. Most costs were within original financial targets. In addition, the British Diabetic Association provided a 3-year support grant of £ 17,129. Capital costs of the new building and equipment (total of £ 88,000) were considered small when spread over the potential lifetime of the building. 6 tables, 12 references.

TITLE: Fructosamine or Glycated Haemoglobin as a Measure of Diabetic Control? Allgrove, J.; Cockrill, B.L. Archives of Disease in Childhood. 63(4): 418-422. April 1988.

OBJECTIVE: To determine whether measuring fructosamine concentration is a suitable alternative or useful adjunct to the measurement of glycated hemoglobin A1 (HbA_1C) in the routine management of children with diabetes.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: In children with diabetes, fructosamine values were well correlated with HbA_1C levels, but only when values outside the normal range were not excluded. Measuring fructosamine is not a direct substitute for assaying HbA_1C.

RECOMMENDATION: Use of fructosamine concentration for routine diabetes control monitoring, with HbA_1C assayed under select conditions, should be considered a cost-saving alternative to routine HbA_1C monitoring.

ABSTRACT: The authors evaluated the potential of measuring fructosamine concentration as an alternative or adjunct to HbA_1C measurement in routinely managed children with diabetes. HbA_1C, fructosamine, and total proteins were assayed in 61 children with diabetes and 30 normal children. Linear regression analysis showed a highly significant correlation in children with diabetes between HbA_1C and fructosamine; however, when values outside the normal range were excluded, the significant relation- ship disappeared. The same results were obtained when HbA_1C values were plotted against the fructosamine:protein ratio. In three newly diagnosed patients whose disease was being brought under control, fructosamine concentrations were lower than expected; in one deteriorating patient, values were higher than expected. The authors conclude that fructosamine is not a direct substitute for HbA_1C. The routine addition of a fructosamine test to a HbA_1C measurement would increase overall costs of assessing diabetic control (by 10 to 15 percent if the reagents are made up in the laboratory) and cannot be justified. Conversely, substituting fructosamine for HbA_1C monitoring would provide only small savings in a pediatric clinic and cannot be justified on clinical grounds. Using fructosamine for most routine monitoring, with HbA_1C assayed on selected occasions, would reduce overall costs of assessing diabetic control while retaining the flexibility of having both assays available. 2 tables, 4 figures, 12 references.

TITLE: Importance of Outpatient Supervision in the Prognosis of Juvenile Diabetes Mellitus: A Cost/Benefit Analysis. Deckert, T.; Poulsen, J.E.; Larsen, M. Diabetes Care. 1(5): 281284. September/October 1978.

OBJECTIVE: To investigate whether outpatient follow-up visits to a subspecialized clinic have a favorable effect on survival of patients with type 1 diabetes, and to evaluate whether the cost of this effort is in reasonable proportion to the benefit obtained.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost-benefit analysis.
Perspective: Health care system.

CONCLUSION: A clear relationship was found between the frequency of outpatient visits to a subspecialized clinic and the duration of survival with diabetes. Outpatient supervision in a subspecialized clinic involves relatively little cost compared with the benefit derived.

RECOMMENDATION: None.

ABSTRACT: The study consisted of 180 insulin-treated patients with diabetes (89 females, 91 males) who had been referred for admission to the Steno Memorial Hospital in Denmark before their 15th year of diabetes. Sixteen patients could not be traced after 40 years of diabetes; the others were followed until death or until 40 years of diabetes. Subsequent to the first hospital admission, 77 patients never attended follow-up in the outpatient clinic; the others attended 1 to 145 times in the course of their first 20 years of diabetes. (Only outpatient visits that took place within these first 20 years were included in the authors' analysis.) The aim of the outpatient supervision was to keep the patient symptom free and socially well adapted, to avoid long periods of poor metabolic regulation, to prevent insulin reactions, and to adapt the treatment to intercurrent diseases. The authors found a clear correlation between the number of outpatient visits and the duration of survival with diabetes: patients seen more than 20 times in the outpatient clinic in the course of their first 20 years of diabetes had a longer survival with diabetes (mean, 38.5 years) than patients who did not have outpatient follow-up (mean, 26.6 years); this difference was highly significant (p = 0.0001). Those in the first group averaged 4.4 outpatient follow-up visits annually over a 12.5-year period (first visit was at an average of 7.5 years of follow-up). If regular follow-up for these patients produced 11.9 additional years of work, total benefit from years gained working (1976 earnings of an unskilled laborer were used) and hospital admissions eliminated was estimated to be $100,656. In contrast, cost per patient (1976 prices) of running the clinic for 40 years was estimated at $10,468. 1 figure, 1 table, 2 references.

TITLE: Inpatient or Outpatient Initiation of Insulin Therapy: Experience and Cost Effective Analysis in a Suboptimal Clinical Setting. Mengistu, M.; Lungi, Y.; Mamo, F. Tropical and Geographical Medicine. 43(1-2): 180-183. January-April 1991.

OBJECTIVE: To assess the safety, feasibility, and cost-effectiveness of initiating insulin therapy in an outpatient setting in an area where day-care centers for persons with diabetes do not exist.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost-effectiveness analysis.
Perspective: Health care system.

CONCLUSION: Initiation of insulin therapy in the outpatient setting was safe, feasible, and cost effective even though the setting was suboptimal.

RECOMMENDATION: None.

ABSTRACT: The authors report initiation of insulin therapy in a suboptimal clinical setting in Ethiopia for patients with newly diagnosed diabetes mellitus. Between 1985 and 1989, 53 patients with diabetes and no impending or established ketoacidosis began insulin therapy as outpatients. Treatment began on a Monday to allow sufficient working days for patient education. Patients returned to the outpatient clinic for 3 to 4 weekly visits, then were followed every 2 to 3 months. Outpatient costs included clinic staff salaries and taxi fare; examining room rent was excluded as minimal. Results were compared with 51 retrospectively chosen inpatients hospitalized for insulin initiation because of distance from the clinic or existing practice. Inpatient costs included medical and nonmedical staff and the declarable cost of beds, which included food and other utilities. Costs for third-class (n = 46), second-class (n = 4), and first-class (n = 1) beds were 2.2, 12.0, and 30.0 Birr, respectively (2.07 Birr = $1 U.S.). Mean blood glucose levels before and after treatment were 357.9 and 197.6 mg/dl, respectively, for inpatients and 325.4 and 205.8 mg/dl, respectively, for outpatients. Mean weight before and after treatment was 44.8 and 49.1 kg, respectively, for inpatients and 53.3 and 56.4 kg, respectively, for outpatients. Mean insulin dose for inpatients and outpatients was 29.4 " 14.7 units and 28.7 " 9.2 units, respectively. Mean duration of hospitalization for inpatients was 24.3 " 9.2 days. Total cost per outpatient visit was 6.8 Birr; total cost per inpatient day was 24.3 Birr. Total outpatient and inpatient costs were 1,346.2 and 30,115.0 Birr, respectively. The savings realized with outpatient care equaled $14,000 (U.S. dollars). Outpatient initiation of insulin therapy is safe, acceptable, and cost effective. 3 tables, 10 references.

TITLE: Insulin Injection Technique Can Be Taught without Hospitalization. Lester, F.T.; Demissie, Y.; Negash, A. Ethiopian Medical Journal. 28(4): 191-195. October 1990.

OBJECTIVE: To assess the feasibility of outpatient instruction in insulin injection technique for patients with diabetes.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Outpatient instruction in insulin injection technique is safe and feasible.

RECOMMENDATION: None.

ABSTRACT: The authors assessed the feasibility of outpatient instruction in insulin injection technique as an alternative to the traditional Ethiopian practice of inpatient instruction for patients with diabetes. A total of 144 patients were instructed in insulin injection technique over a 1-year period at an Addis Ababa hospital; 85 of these patients were instructed while hospitalized and 59 were instructed as outpatients. Patients were instructed in a single group by two nurses; sessions were held 6 days a week. Outpatients needed an average of 4.7 days to learn the technique; an additional 3 to 4 weeks of frequent monitoring was needed to attain control. No severe hypoglycemic reactions were seen among the outpatients. Instruction was unsuccessful for two very symptomatic outpatients. Six to 16 months after the instruction period, 62.7 percent of the outpatients but only 50.6 percent of the inpatients were still attending the clinic. Hospitalization is disruptive and costly and diabetic control achieved there is artificial, i.e., control has not been achieved in the patient's natural environment. 3 tables, 6 references.

TITLE: Insulin-Glyburide Combination Therapy for Non-Insulin-Dependent Diabetes Mellitus: A Long-Term Double-Blind, Placebo-Controlled Trial. Casner, P.R. Clinical Pharmacology and Therapeutics. 44(5): 594-603. November 1988.

OBJECTIVE: To determine whether the combination of a sulfonylurea agent and insulin can improve glycemic control in patients with type 2 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Randomized clinical trial.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: The combination of a sulfonylurea agent and insulin did not improve glycemic control during the study year, and the extra cost and inconvenience of combination therapy would not make it worthwhile except in a few select patients.

RECOMMENDATION: Providing combination therapy in a pulse form instead of continually may be more effective and may need to be studied; insulin clamp studies are required to address insulin sensitivity.

ABSTRACT: Sixty-four patients with insulin-treated type 2 diabetes were treated with a combination of insulin and oral sulfonylurea therapy or insulin and a placebo in a double-blind controlled trial conducted over 12 months. There were no significant differences between the two groups of patients except for age; the control group was 4.2 years older. In the treatment group, fasting blood glucose concentrations were significantly below those for the control group only at 3 and 4 months. Insulin dose for the treatment group was significantly below the dose for the control group at month 8 only. In comparisons of the two groups, glycohemoglobin values were significantly lower for the control group at months 3 and 6 only; C-peptide levels were significantly higher in the treatment group at 3 and 9 months only. At each visit, placebo or glyburide was titrated to higher doses, with both groups attaining maximal dosage (20 mg/day). Oral glucose tolerance test results did not significantly differ between groups except at 3 months, when the control group's were higher. The treatment group was divided into responders (a fasting blood glucose concentration # 140mg/dL on at least 3 visits and a 25 percent reduction in total daily insulin requirements sometime during the study) and nonresponders. Forty-two percent were responders; 29 percent of these responders received a > 50 percent reduction of insulin dose at some time during the study. By the end of the study, the daily insulin dose for responders was about 50 percent lower than that of the control group and non-responders, suggesting that combination therapy may improve insulin sensitivity and enhance insulin secretion. Patients in the treatment group experienced no serious side effects from the combination of insulin and glyburide. Combination therapy may increase the cost of treatment by nearly 50 percent, and its effect appears to be transitory in most patients. Patients most likely to benefit are those who show an increase in C-peptide levels when taking oral hypoglycemics. 6 figures, 2 tables, 41 references.

TITLE: Less Expensive, Reliable Blood Glucose Self-Monitoring. Spraul, M.; Sonnenberg, G.E.; Berger, M. Diabetes Care. 10(3): 357-359. May-June 1987.

OBJECTIVE: To determine whether splitting test strips for capillary blood glucose self-monitoring will permit reliable measurement of blood glucose levels.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Cutting capillary blood glucose test strips (Chemstrips bG) in half had no effect on the reliability of self-assessment of blood glucose levels and halved the cost of the test.

RECOMMENDATION: All patients should be advised to split their test strips (Chemstrips bG) to reduce the costs of blood glucose self-monitoring.

ABSTRACT: The authors tested the reliability of using split test strips to self-assess blood glucose levels by experimenting with Haemoglukotest 20-800 (Chemstrips bG) test strips. Ten participants (patients and laboratory personnel) who were trained in reading test strips but had never used split test strips were each asked to test blood samples using 10 full-size strips, 10 strips halved mechanically with a splitting device, and 10 strips halved manually with scissors. Results were compared with plasma glucose concentrations measured in a Beckman glucose analyzer. The correlation coefficients (r) values for uncorrected plasma glucose concentrations were 0.95 for full-thickness strips, 0.966 for mechanically split strips, and 0.966 for manually split strips, respectively. The mean absolute deviation from the reference value and standard error (in mg/dl) was 17 " 2 for full strips, 17 " 1 for mechanically split strips, and 17 " 1 for manually split strips. There was no significant difference between the three self tests and the laboratory analyzer by analysis of variance (p = .98), variance between participants (p = .12), or interrelationship between participants' readings and the three methods (p = .61). Test strips reliably monitor blood glucose concentrations, and halving the strips with scissors cuts the cost of monitoring in half without reducing reliability. Patients should be advised to use split test strips to save money. 1 figure, 1 table, 5 references.

TITLE: Lifetime Benefits and Costs of Intensive Therapy as Practiced in the Diabetes Control and Complications Trial. The Diabetes Control and Complications Trial Research Group. Journal of the American Medical Association (JAMA). 276(17): 14091415. November 6, 1996.

OBJECTIVE: To examine the costs and benefits of intensive treatment of type 1 diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Health care system.

CONCLUSION: Intensive therapy for those meeting eligibility criteria would improve quality and length of life and represents a good monetary value.

RECOMMENDATION: None.

ABSTRACT: The authors developed a Monte Carlo simulation model to estimate the lifetime benefits and costs of conventional and intensive insulin therapy for patients with type 1 diabetes. The intensive therapy regimen was designed to achieve blood glucose levels as close to the nondiabetic range as possible with three or more daily insulin injections or with treatment with an insulin pump, self-monitoring of blood glucose levels four times daily, and monthly outpatient visits with a multidisciplinary team. Conventional therapy consisted of one or two injections each day, daily self-monitoring, and quarterly visits. Data were collected as part of the Diabetes Control and Complications Trial (DCCT) and supplemented with data from other clinical trials and epidemiologic studies. The authors assumed that approximately 17 percent of the U.S. population with type 1 diabetes (120,000 persons) would qualify for enrollment in the DCCT. They included all direct medical costs for the Cost-effectiveness analysis but did not include direct nonmedical costs (e.g., transportation, lodging) or potential production losses arising from disease related absence from work, long-term disability, or premature death. The costs of conventional and intensive therapy were based on actual resources used in the DCCT. The Monte Carlo model simulated the course of the patient's disease over his or her expected lifetime. The authors found that implementing intensive rather than conventional therapy among the 120,000 persons meeting DCCT criteria would gain 920,000 years of sight, 691,000 years free from end stage renal disease, 678,000 years free from lower extremity amputation, and 611,000 years of life, at an additional cost of $4 billion (1994 U.S. dollars with future costs and effects discounted 3 percent) over the lifetime of the group. The incremental cost per year of life gained was $28,661. The incremental cost per quality-adjusted life-year gained with intensive treatment was $19,987. 2 figures, 4 tables, 49 references.

TITLE: Management of Patients with Diabetes by Nurses with Support of Subspecialists. Peters, A.L.; Davidson, M.; Ossorio, R.C. HMO Practice. 9(1): 8-13. March 1995.

OBJECTIVE: To describe a program for managing patients with diabetes via diabetes nurses within a health maintenance organization (HMO) system, and to report on the cost-effectiveness of the program.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: Patients achieved significant decreases in glycated hemoglobin levels; the program appears to have achieved substantial cost savings by avoiding acute hospitalizations for diabetes in 244 patients.

RECOMMENDATION: None.

ABSTRACT: Staff at the Cedars-Sinai Medical Center in Los Angeles developed a program to manage diabetes patients called the Comprehensive Diabetes Care Service. Diabetes nurse specialists, supervised by diabetologists, provide diabetes and lipid management per protocols. Staffing ratios are one nurse and one staff assistant for each 250 patients with diabetes and one diabetologist for each 1,000 patients. An independent practice association at the hospital with approximately 150 physicians and a capitation contract with 12 HMOs (as of 1986) was persuaded to participate in the diabetes program. The program began in February 1987 with a capitated population of approximately 8,000; in July 1994, that population was approximately 65,000. Actual patients with diabetes followed in the program numbered 236 in January 1990 and 754 in December 1993. A diabetologist sees the patient in conjunction with the nurse initially and annually. The nurse sees the patient quarterly and follows a defined format at each visit. Patients are referred to a dietitian at entry and are seen by an ophthalmologist at their first visit. A computer system was designed to improve patient management. The authors found that glycated hemoglobin levels improved significantly (p < 0.01) in a subset of all patients and a subset of compliant patients but not in noncompliant patients. The authors estimated that over a 4-year period (1990 to 1993), 244 patients who would probably have been hospitalized by other physicians at the hospital were not because of the intensive outpatient follow-up available. The authors estimated the potential net cost savings to be $568,721 over the 4 years. 1 figure, 2 tables, 9 references.

TITLE: Nursing Case Management: An Innovative Model of Care for Hospitalized Patients with Diabetes. Edelstein, E.L.; Cesta, T.G. Diabetes Educator. 19(6): 517-521. November-December 1993.

OBJECTIVE: To evaluate the effect of using a diabetes team with a case management approach to quality of care and length of hospital stay.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Length of hospital stay was reduced for patients with a primary and secondary diagnosis of diabetes whose care was managed by the team. Blood glucose control before discharge was at least fair in more than 98 percent of team patients before discharge.

RECOMMENDATION: None.

ABSTRACT: The authors reported on 160 patients admitted to Beth Israel Medical Center in New York City who did not have private physicians. Eighty-five patients had a principal diagnosis of diabetes; 49, a secondary diagnosis of diabetes; and 26 were antepartum patients admitted for blood glucose control. A team consisting of a case manager, diabetes nurse educator, diabetologist, and nutritionist managed the care of these patients. Consultations per patient totaled 7.5; average length of stay of the patients with a principal diagnosis of diabetes was 4.14 days, 58 percent lower than the hospital's 1990 average for such patients. For patients with a secondary diagnosis of diabetes, the average length of stay was 14 percent below the 1990 Beth Israel average. The study readmission rate was 7.33 percent within 30 days of discharge, but readmissions were unrelated to diabetes. The team patients had excellent (28.7 percent), good (26.9 percent), or fair (42.7 percent) blood glucose control prior to discharge. 3 figures, 14 references.

TITLE: Pointers to Preventing Hyperglycaemic Emergencies in Soweto. Buch, E.; Irwig, L.M.; Huddle, K.R.; Krige, L.P.; Krut, L.H.; Kuyl, J.M. South African Medical Journal. 64(18): 705-709. October 22, 1983.

OBJECTIVE: To analyze emergency hospital admissions for hyperglycemia in Soweto, South Africa, including associated costs; to assess the use of health services by emergency patients before and after hospital admission.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: Health service costs for admissions for hyperglycemic emergencies in Soweto were estimated as at least 462,000 rand (South African currency) in 1981; many admissions were preventable, and patient use of health resources and self-monitoring prior to and after admission were inadequate.

RECOMMENDATION: Ambulatory services for patients with diabetes should be developed and upgraded in Soweto, with the following specific considerations: improved patient education, allocation of special staff for hyperglycemic emergencies, establishment of a registry for such emergencies, improved records systems, and application of compliance-improving strategies.

ABSTRACT: The authors reviewed the incidence of admissions for hyperglycemic emergencies during a 2-month period in 1981 at Baragwanath Hospital in Soweto, the associated costs, and the use of health services by patients with diabetes prior to and following admission. Problems related to providing diabetes care are compounded in Soweto by poor socioeconomic conditions. During the study period, 60 patients were admitted for hyperglycemic emergencies; 15 (25 percent) died in the hospital; just 3 of the deaths may have been unrelated to the emergency. Of 56 patients interviewed, 7 percent were previously undiagnosed with diabetes. Average hospital stay was 20 days, and the estimated cost of treating the emergency admissions during the study period was 77,000 rand. Of the 49 patients known to have diabetes prior to admission, only 19 (39 percent) had attended any health services 3 or more times during the 3 months before admission, and 25 percent had not attended a health service at all. (Three attendances would be the minimum to receive adequate medication.) Of the 36 patients with whom follow-up interviews were conducted 3 months after discharge, 12 (33 percent) had been readmitted. Extrapolating from the study sample, the cost of hyperglycemic emergencies to Soweto's health services was estimated as at least 462,000 rand for 1981. 5 figures, 3 tables, 7 references.

TITLE: Precision and Costs of Techniques for Self-Monitoring of Serum Glucose Levels. Chiasson, J.L.; Morrisset, R.; Hamet, P. Canadian Medical Association Journal. 130(1): 38-43. January 1984.

OBJECTIVE: To evaluate the accuracy and cost of blood glucose monitoring techniques involving a glucose oxidase reagent strip with or without a reflectance meter.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: The techniques evaluated, particularly those involving reflectance meters, compared favorably with standard laboratory methods, were easy to perform, and were less expensive.

RECOMMENDATION: Blood glucose monitoring using a glucose oxidase reagent strip and reflectance meter should be considered a feasible, cost-effective alternative to standard laboratory testing.

ABSTRACT: The authors evaluated the accuracy and cost of seven blood glucose monitoring techniques available for patient self-monitoring; they also examined the correlation between serum glucose concentrations obtained. The blood glucose techniques, which involved a glucose oxidase reagent strip with or without a reflectance meter, were evaluated in more than 100 patients with or without diabetes. Two of the techniques were tested by members of a rotating staff of hospital nurses, all seven by a nurse specifically trained for the task, and four of the seven by the patients themselves. Results were compared with reference values determined by the hexokinase method in a hospital biochemistry laboratory using the Simultaneous Multiple Analysis, Computerized (SMAC) or Autoanalyzer II (AAII) (Technicon Instruments Corporation, Tarrytown, New York) assays. Costs were evaluated via a survey of drugstore prices. Techniques involving a reflectance meter showed excellent correlation with the laboratory reference method (r2 = 0.85 to 0.96) for all three types of testers. Correlation with the reference standard was poorer for techniques not involving a reflectance meter (r2 = 0.69 to 0.90); correlation improved when the techniques were performed by a specially trained nurse. Reflectance meters ranged in price (Canadian dollars) from $290 to $565. Costs for bottles of 25 reagent strips ranged from $13 to $19.80. Costs for each determination using the self-monitoring techniques ranged from $0.46 to $0.99; these values were up to three times less expensive than the laboratory methods ($1.86 for the SMAC and $1.04 for the AAII). These blood glucose monitoring techniques, particularly those using a reflectance meter, were reliable, easily performed by patients, and less expensive per determination when compared with laboratory testing. In contrast, the correlation between urine and glucose levels was poor (r2 = 0.21). 4 figures, 12 tables, 19 references.

TITLE: Randomized Controlled Multicentre Evaluation of an Education Programme for Insulin-Treated Diabetic Patients: Effects on Metabolic Control, Quality of Life, and Costs of Therapy. de Weerdt, I.; Visser, A.P.; Kok, G.J.; de Weerdt, O.; van der Veen, E.A. Diabetic Medicine. 8(4): 338-345. May 1991.

OBJECTIVE: To assess the effect of an education program for insulin-treated patients with diabetes on their metabolic control, quality of life, and costs of therapy.

CATEGORY: Secondary intervention.

Type of Study: Randomized clinical trial.
Methodology: Cost analysis.
Perspective: Health care system.

CONCLUSION: The education program had no effect on metabolic control, quality of life, and costs of therapy, although the patients significantly increased their level of self-monitoring blood glucose and diabetes knowledge. The authors suggest various reasons for the failure to find effects.

RECOMMENDATION: None.

ABSTRACT: This study was conducted in 15 randomly selected hospitals in The Netherlands. Control patients were taken from 5 hospitals and intervention patients (experimental) from the other 10. In five of the experimental hospitals the program was led by a health care worker; in the other five, by a fellow patient. A total of 625 patients were asked to participate, 558 (89 percent) of whom agreed. Each group had 40 to 50 patients. The patient characteristics of the groups (age, occupational status, duration of diabetes, duration of insulin use, and number of daily injections) did not differ significantly. Patients in the experimental group were evaluated immediately before and after the education program and at 1 and 6 months following the program's conclusion. Patients in the control group were assessed twice (at 6- to 7-month intervals). At each evaluation, participants completed a questionnaire and a blood sample was taken. The effect of the program on metabolic control (glycosylated hemoglobin, serum fructosamine concentration, number and severity of hypoglycemic reactions, and clinical signs of poor control), quality of life (Affect-Balance Scale and subjective social indicators), and costs of therapy (hospitalizations, doctor visits, medication, self-testing, sick days, and cost of education program) was assessed. No significant effect from the education program on any one of these variables could be found, although there was significant improvement in self-care and diabetes knowledge. This lack of effect could be due to the quality of the education program, short follow-up of the study, the lack of integration of the education program in standard therapy, the lack of follow-up of the education given, or the lack of concurrent changes in diabetes therapy. 4 tables, 30 references.

TITLE: A Risk-Benefit Assessment of Conventional versus Intensive Insulin Therapy. Reichard, P. Drug Safety. 10(3): 196202. March 1994.

OBJECTIVE: To assess the benefits of intensive insulin therapy for type 1 diabetes.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Review of studies.
Perspective: Health care system.

CONCLUSION: Intensified insulin therapy leads to a slowing down or halting of microvascular complications in patients with type 1 diabetes, including those with diabetes of short duration and those with longer term diabetes and nonproliferative retinopathy.

RECOMMENDATION: An intensified treatment program is generally indicated for the majority of adults with type 1 diabetes; such treatment should start before the complications become too advanced.

ABSTRACT: The author explores whether intensified insulin treatment to lower blood glucose concentrations is effective in reducing diabetes complications by addressing five questions: (1) Is there an association between blood glucose concentrations and complications? (2) Is the association causal? (3) Are there any adverse effects of intensified treatment? (4) Is intensified treatment acceptable? and (5) How great could the gains be with general acceptance and execution of the treatment program? By reviewing various studies, most notably the Stockholm Diabetes Intervention Study and the Diabetes Control and Complications Trial, the author finds the following: Lower blood glucose concentrations are associated with reduction of complications affecting the eyes, kidneys, and nerves. Prospective, randomized studies are needed to answer the question of a causal association. The chief adverse effects of intensified insulin treatment are weight gain and an increased frequency of serious hypoglycemic episodes. Intensified treatment is acceptable to patients. The gains are very large, not only for the individual patient, but also for society as a whole. The author speculates that if 400,000 patients were put on an intensive treatment program, 64,000 who would develop nephropathy if conventionally treated would not do so, and 100,000 would not need laser treatment for retinopathy who otherwise would. Implementing an intensified treatment program for type 1 diabetes would bring a significant reduction of serious complications and a gain in terms of patient discomfort and cost. 2 figures, 32 references.

TITLE: Utility of Serum Fructosamine as a Measure of Glycemia in Young and Old Diabetic and Non-Diabetic Subjects. Negoro, H.; Morley, J.E.; Rosenthal, M.J. American Journal of Medicine. 85(3): 360-364. September 1988.

OBJECTIVE: To determine the utility of a serum fructosamine assay as a measure of glycemia.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Statistical analysis.
Perspective: Health care system.

CONCLUSION: Serum fructosamine was comparable to other measures of glycemia.

RECOMMENDATION: Serum fructosamine should be considered superior to measures of other glycosylated proteins because of its automaticity, reproducibility, and lower cost.

ABSTRACT: The authors evaluated the utility of serum fructosamine as a measure of glycemia by comparing it with other standard glycemic indices: glycosylated hemoglobin, glycosylated albumin, and fasting glucose. Participants were 145 adults in six categories, including two groups of healthy persons (aged 22 to 50 and 64 to 86) without diabetes; one group with chronic debilitating illness (aged 60 to 91); otherwise healthy persons in a maintenance methadone program; and two groups (mean ages: 48 and 70 years) of patients with diabetes. Blood samples were collected at 8 a.m. after a 12-hour fast. Serum fructosamine detected glycemia as well as the established tests, and fructosamine concentrations correlated well with serum glycosylated hemoglobin, glycosylated albumin, and fasting glucose. Fructosamine concentrations were not significantly affected by any of a large battery of medications or by large differences in plasma lipid levels. Age did not affect serum fructosamine concentration, which makes this a particularly useful test for older persons with diabetes. Serum fructosamine is also amenable to automated determination and is far simpler and less expensive than assays for glycosylated proteins. The current charge to patients for assays is $6.50 for plasma glucose, $8.00 for fructosamine, $26.00 for glycosylated hemoglobin, and $31.00 for glycosylated albumin. 2 figures, 3 tables, 27 references.

TITLE: Variation in Office-Based Quality: A Claims-Based Profile of Care Provided to Medicare Patients with Diabetes. Weiner, J.P.; Parente, S.T.; Garnick, D.W.; Fowles, J.; Lawthers, A.G.; Palmer, R.H. Journal of the American Medical Association (JAMA). 273(19): 1503-1508. May 17, 1995.

OBJECTIVE: To demonstrate that profiling of claims data can be used as an ongoing method to support ambulatory care quality improvement, to measure the quality of office-based care provided to elderly patients with diabetes, and to identify factors associated with higher or lower conformance to recommended criteria of care.

CATEGORY: Secondary intervention.

Type of Study: Patient management.
Methodology: Claims review.
Perspective: Health care system.

CONCLUSION: The care of elderly patients with diabetes is not meeting nationally recommended guidelines; screening rates are particularly low in rural areas. Examining Medicare claims has great potential for measuring patterns of care.

RECOMMENDATION: Programs should be established to disseminate clinical practice guidelines relevant to primary care of diabetes and further research should be conducted to delineate variations in office practice patterns.

ABSTRACT: The authors assessed the quality of office-based care provided to elderly patients with diabetes. Data were derived from review of all Medicare claims (Parts A and B) in Alabama, Iowa, and Maryland for July 1990 to June 1991 for patients aged 65 years and over. Four services/procedures were identified as quality indicators for diabetes: three (hemoglobin A1c measurement, ophthalmologic exam, and total cholesterol measurement) were to be performed at least annually; the fourth, blood glucose measurement, was identified as a limited-use procedure. Across the three states, only 16.3 percent of patients with diabetes underwent hemoglobin A1c measurement, 45.9 percent received an ophthalmologic exam, and 55.1 percent underwent cholesterol measurement at least once during the year. In contrast, four-fifths (80.5 percent) of patients underwent blood glucose measurement. In almost all instances, practice patterns varied significantly across the three states, even after controlling for practice and patient characteristics. With the exception of ophthalmologic exams, rates of procedures were significantly lower in rural areas. Results demonstrate that elderly patients with diabetes are not receiving optimal care; a large gap exists between national practice guidelines and actual primary care practice. Clinical practice guidelines need to be more widely disseminated. Further research is needed to identify the causes of variation across office practices in order to design quality improvement programs. 3 tables, 38 references.

Back to Economics of Diabetes Mellitus Indexes

Top of page

Historical
Page last modified: December 20, 2005
Content Source: National Center for Chronic Disease Prevention and Health Promotion
Division of Diabetes Translation

CDC Home | Search | Health Topics A-Z

United States Department of Health and Human Services
Centers for Disease Control and Prevention
National Center for Chronic Disease Prevention and Health Promotion
Division of Diabetes Translation

Publications and Products

The Economics of Diabetes Mellitus: An Annotated Bibliography

The Economics of Diabetes Mellitus:
An Annotated Bibliography