Diabetes Home

The Economics of Diabetes Mellitus: An Annotated Bibliography

Historical

This webpage is archived for historical purposes and is no longer being maintained or updated.

TYPES OF INTERVENTION

Tertiary Intervention

Abstracts 65–91

Eye Care

TITLE: Cost Savings Associated with Detection and Treatment of Diabetic Eye Disease. Javitt, J.C. PharmacoEconomics. 8 (Supplement 1): 33-39. 1995.

OBJECTIVE: To estimate current and potential savings in the United States and Sweden from screening and treating retinopathy in persons with diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Societal.

CONCLUSION: Treatment of retinopathy in patients with diabetes mellitus yields substantial savings of sight years and money.

RECOMMENDATION: Eye care for patients with diabetes must emphasize patient identification, carefully maintained follow-up, and prompt, appropriate treatment.

ABSTRACT: The author describes the use of the PROPHET modeling system, a program designed to model the progression of a chronic, irreversible disease, to estimate savings from recruiting, screening, and treatment programs for diabetic eye disease. Data from cross-sectional and longitudinal studies and clinical trials are used in the model. The analysis derives the costs of screening and treatment from average Medicare charges for 1990; savings as well as costs are expressed in 1990 U.S. dollars using a discount rate of 5 percent. An annual federal expenditure of $14,296 is predicted for blind persons with diabetes under 65 years of age, just $32 (not counting Medicare, Social Security, income tax exemption) for those 65 and over. Based on studies by Klein et al. (1987) in Wisconsin, the implementation rate of eye screening is currently 60 percent. Even at this suboptimal level, screening and treatment for eye disease in patients with diabetes generates annual savings of $350 million to the federal budget and 100,000 person-years of sight. Each additional person (beyond the 60 percent level) enrolled in appropriate screening and treatment is associated with net lifetime savings of $9,571 (type 1 diabetes) or $973 (type 2 diabetes). The Swedish Council on Technology Assessment in Health Care repeated this analysis and found that 60 percent implementation of screening could potentially save 22 million SEK. Their analysis found that savings associated with detection and treatment were 10 times greater than costs. The authors of the present study found that changing the frequency of screening for patients with no or mild background retinopathy from 1 to 2 years does not reduce years of sight saved and reduces screening costs if the sensitivity of eye screening is 80 percent or greater. 3 figures, 43 references.

TITLE: Cost-Benefit Analysis of Diabetic Eye Disease. Matz, H.; Falk, M.; Gottinger, W.; Kieselbach, G. Ophthalmologica. 210(6): 348-353. 1996.

OBJECTIVE: To compare the costs of blindness caused by diabetes with the costs of screening and treatment of retinopathy associated with diabetes.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Screening for blindness saves more money than it costs.

RECOMMENDATION: All patients with diabetes should be screened for retinopathy and optimal therapy should be initiated.

ABSTRACT: The authors calculated the cost-benefit of screening and testing of diabetes-associated eye disease in the state of Tyrol, Austria. Current and new cases of severe retinopathy (proliferative retinopathy or clinically significant macular edema) were calculated to be 2,147 and 252, respectively. Without proper intervention, all 252 persons with incident cases would be expected to go blind within 10 years, vision could be retained in at least one eye in 179 cases with proper treatment, and blindness could be delayed an average of 3 years in patients for whom it is unavoidable. Costs of blindness included disability payments and allowances; exemptions from telephone, television, and radio fees; tax exemptions; financial aid based on reduced earnings and disability; and early retirement pensions. The anticipated costs of blindness were ATS 19 million overall, of which ATS 14.6 million was considered preventable, assuming 100 percent ability to diagnose severe retinopathy and macular edema and 100 percent ability to treat these conditions at an optimal time. Cost was based on 10-year life expectancy at onset of severe retinopathy, duration of blindness ranging from 2.5 to 8.5 years, sex, age, employment status, type of diabetes, and correlation of minimum retirement age with life expectancy. In Tyrol, 16,913 persons would require annual eye examinations (cost: ATS 5,191,445) and 2,147 persons would require examinations 4 times per year (ATS 3,022,547). Laser coagulation or vitrectomy plus examinations would cost the national health care system ATS 8,560,089 and private insurance ATS 2,119,566 (only 17.7 percent of the population has supplemental private insurance). Up to ATS 3.9 million can be saved through adequate screening and treatment to prevent blindness. 1 figure, 5 tables, 25 references.

TITLE: Cost-Effective Screening for Retinopathy Using a Nonmydriatic Retinal Camera in a Pre-paid Health-Care Setting. Peters, A.L.; Davidson, M.B.; Ziel, F.H. Diabetes Care. 16(8): 1193-1195. August 1993.

OBJECTIVE: To evaluate the effectiveness of a nonmydriatic retinal camera as a screening tool for serious retinopathy.

CATEGORY: Tertiary intervention.

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

CONCLUSION: The nonmydriatic retinal camera was highly sensitive in detecting serious retinopathy.

RECOMMENDATION: The nonmydriatic retinal camera should be considered an easy-to-use, inexpensive component of screening strategies for retinopathy.

ABSTRACT: The authors evaluated the effectiveness of the Canon CR4-45 nonmydriatic retinal camera as a screening tool for retinopathy. Nurse clinicians took retinal photos of the 522 patients in a diabetes program affiliated with a health maintenance organization at their initial and annual visits. The degree of retinopathy as assessed by the reader (a diabetologist) of the retinal photos was compared with results of examinations of the patients by retinal specialists. These ophthalmologists used direct and indirect ophthalmoscopy as well as slit-lamp biomicroscopy. Sensitivity and specificity values for photo interpretation were based on 189 patients who had gradable photos for both eyes and a retinal examination report available. Comparison of any retinopathy noted by the examination with any retinopathy noted by the reader yielded a sensitivity and specificity of 85 percent and 93 percent, respectively. The reader occasionally missed background retinopathy not requiring treatment. If serious retinopathy was seen on the examination, the reader always noted some retinopathy (sensitivity: 100 percent; specificity: 82 percent). Results show that the nonmydriatic retinal camera, which is easy to use and inexpensive, can be employed as a screening tool for detecting serious retinopathy. The camera might identify more patients at risk for serious retinopathy than routine referral for ophthalmologic screening, for which compliance rates are often poor (74 percent in this study). 1 table, 11 references.

TITLE: Cost-Effectiveness of Alternative Methods for Retinopathy Screening. Lairson, D.R.; Pugh, J.A.; Kapadia, A.S.; Lorimor, R.J.; Jacobson, J.; Velez, R. Diabetes Care. 15(10): 1369-1377. October 1992.

OBJECTIVE: To compare the cost-effectiveness of four approaches to screening for retinopathy.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Screening with retinal photographs in a primary care setting can be cost effective. Cost per true-positive case was lowest for retinal photography with dilation.

RECOMMENDATION: Screening could be offered as an addition to a routine visit in a primary care clinic or on a separate visit.

ABSTRACT: The four methods of screening for retinopathy compared were (1) 45-degree retinal camera photography taken by a physician assistant or nurse practitioner without pupil dilation (n = 351), (2) the same kind of photography with pupil dilation (n = 351), (3) direct and indirect funduscopic examination (dilated pupils) by an ophthalmologist (n = 347), and (4) direct ophthalmoscopic examination (dilated pupils) by a physician assistant or nurse practitioner (n = 172). All photographs (methods [1] and [2]) were read by an ophthalmologist. Positive retinopathy was considered a level of 40 or greater on the modified Airlie House reference standard. Screening costs were incurred through Veterans Administration and Department of Defense facilities and included staff salaries and fringe benefits; equipment (amortized over expected life); supplies; participants' travel, screening time, and transportation; space; overhead; and reading fees. Interest rates were set at 8.24 percent, based on 1989 long-term government bond rates, and overhead costs were set at 29.7 percent of ambulatory care health system costs. Patient travel costs were assessed through an interview; wages for unemployed patients were calculated at $3.35 per hour. Cost-effectiveness was based on the cost per true-positive test. Sensitivity of methods 1-4, respectively, was 0.61, 0.81, 0.33, and 0.10; specificity was 0.85, 0.96, 0.99, and 0.99. Total cost (health system plus patient) per exam was $70, $74, $48, and $31, respectively. The average cost to the health care system only per case of identified retinopathy was $295 (method 2); $378 (method 1); $390 (method 3); and $794 (method 4). Adding the patient costs did not change these rankings. 1 figure, 4 tables, 22 references.

TITLE: Cost-Effectiveness of Alternative Methods for Retinopathy Screening. Wareham, N.J. Diabetes Care. 16(5): 844. May 1993.

OBJECTIVE: To point out an issue not raised in an article by Lairson et al. on screening methods for retinopathy (Diabetes Care. 15:1369-1377. 1992. Abstract 53) and offer calculations of sensitivity, specificity, and cost per true-positive for the data presented by those authors.

CATEGORY: Tertiary intervention.

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

CONCLUSION: The conclusions of cost-effectiveness drawn by Lairson et al. may not be as strong as presented.

RECOMMENDATION: None.

ABSTRACT: The author notes that Lairson et al. did not account for error in their sensitivity estimates of the four screening tests they compared. He reassesses their data by applying 95 percent confidence intervals to the sensitivity and specificity of screening patients with diabetes for retinopathy by (1) 45-degree retinal photographs without pupil dilation, (2) retinal photography with pupil dilation, (3) ophthalmologist examination, or (4) technician examination. The sensitivity (with upper and lower 95 percent confidence limits) for 1-4 was 0.61 (0.72, 0.50), 0.81 (0.90, 0.72), 0.33 (0.44, 0.22), and 0.07 (0.14, 0), respectively. The specificity for 1-4 was 0.85 (0.89, 0.81), 0.96 (0.99, 0.94), 1.00 (1.00, 0.99), and 0.99 (1.00, 0.97), respectively. System cost per true-positive diagnosis with these approaches (with upper and lower estimates) was $378 ($463, $330), $295 ($331, $265), $390 ($581, $294), and $794 (not given, $379), respectively. Patient cost per true-positive diagnosis was $171 ($209, $144), $139 ($156, $125), $306 ($454, $230), and $1,009 (not given, $481), respectively. Because of the overlap in confidence intervals, the author points out that conclusions in this paper about the cost-effectiveness of these screening methods are weakened. He also points out that sensitivity is a major determinant of the cost-effectiveness of screening for retinopathy and states that including confidence intervals in the estimate of the cost per true-positive case is critical for making policy decisions. 2 tables, 2 references.

TITLE: Cost-Effectiveness of Current Approaches to the Control of Retinopathy in Type I Diabetics. Javitt, J.C.; Canner, J.K.; Sommer, A. Ophthalmology. 96(2): 255-264. February 1989.

OBJECTIVE: To estimate the benefits (in terms of preservation of vision) and attendant costs of screening, diagnosis, follow-up, and treatment of retinopathy in patients with type 1 diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Societal.

CONCLUSION: Screening, diagnosis, follow-up, and treatment of retinopathy in patients with type 1 diabetes resulted in a cost of $966 per person-year of sight saved, which is only 14 percent of the direct cost to the federal government of a year of blindness-related disability.

RECOMMENDATION: The federal government should fund eye care for patients with type 1 diabetes to avoid the higher cost of disability from blindness.

ABSTRACT: The authors used a computer model to analyze the benefits (in terms of preservation of vision) and attendant costs of screening, diagnosis, follow-up, and treatment of retinopathy in patients with juvenile-onset type 1 diabetes. Screening recommendations were taken from the Public Health Committee of the American Academy of Ophthalmology. Current charges for screening and treatment of retinopathy were compared with current federal budgetary expenses for blindness-induced disability under the Social Security Disability program. Incidence, disease progression, and mortality data were drawn from several population-based studies. The computer model, based on a Monte Carlo simulation, was run for a hypothetical cohort of 31,000 patients with type 1 diabetes beginning at age 12.5 years and followed over a 60-year period. Model inputs were based on published reports of cross-sectional and disease incidence studies, clinical trials, and U.S. statistics. The model predicted that by age 60, background retinopathy would have developed in 98 percent of the cohort, macular edema in 42 percent, and proliferative retinopathy in 72 percent. Despite appropriate screening and treatment, ultimately 28 percent of the cohort, according to the model, would suffer severe vision loss. Over the 60 years, costs for ophthalmologic examinations, focal laser treatment, and panretinal photocoagulation were $91.0 million, $26.0 million, and $44.0 million, respectively. A total of 92,700 person-years of sight could be saved at the end of 60 years by the application of screening and panretinal photocoagulation. The cost of screening and panretinal photocoagulation per person-year of sight saved was $966, which is only 14 percent of the direct cost ($6,900) to the federal government of a year of blindness-related disability payments. 8 figures, 2 tables, 35 references.

TITLE: Cost-Effectiveness of Detecting and Treating Retinopathy. Javitt, J.C.; Aiello, L.P. Annals of Internal Medicine. 124 (1 Part 2): 164-169. January 1, 1996.

OBJECTIVE: To determine from the perspective of health insurers the cost-effectiveness of ophthalmologic screening and treatment to prevent vision loss for patients with diabetes.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Ophthalmologic screening and treatment of eye disease in patients with diabetes is a highly cost-effective approach to preventing blindness in people with diabetes.

RECOMMENDATION: Implementation of current guidelines regarding eye examinations for people with diabetes should be encouraged.

ABSTRACT: Detection and treatment of diabetic eye disease has been shown to result in cost savings. Prevention of diabetes-related blindness, on a cost-effectiveness basis, ranks above other medical interventions commonly provided. The PROPHET system, based on Monte Carlo simulation, was used to model the progression of proliferative retinopathy. Monte Carlo simulation, which uses random number generation, allows for a simple, probability-based solution of complex disease progression processes over time. Screening and treatment costs were derived from average Medicare charges in 1990, using a discount rate of 5 percent. Previous reports suggest that over 413,200 person-years of sight are currently saved, and over 710,800 person-years of sight could be saved, if all patients with diabetes had appropriate ophthalmologic screening and treatment. The cost of implementing currently recommended guidelines for screening and treatment of retinopathy was calculated to be $1,757 per person-year of sight saved. The cost per quality-adjusted life-year (QALY) associated with detecting and treating diabetic eye disease was found to range from $1,996 per QALY for those with type 1 diabetes to $3,530 per QALY for those with type 2 diabetes who do not require insulin. Overall cost of detecting and treating disease in patients with diabetes was $3,190 per QALY. Despite the high level of efficacy, clinical effectiveness, and cost-effectiveness, screening and treatment for diabetic eye disease are not universally practiced. Recommendations for annual dilated-eye examinations of patients with diabetes have now been included in the Health Plan Employer Data and Information Set (HEDIS II) quality guidelines adopted throughout the managed care industry. 1 figure, 4 tables, 55 references.

TITLE: Cost-Effectiveness of Strategies for Detecting Retinopathy. Dasbach, E.J.; Fryback, D.G.; Newcomb, P.A.; Klein, R.; Klein, B.E. Medical Care. 29(1): 20-39. January 1991.

OBJECTIVE: To evaluate the cost-effectiveness of six different strategies for providing ophthalmologic care to patients with diabetes.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Screening for retinopathy is generally cost effective in younger-onset patients and in those older-onset patients who take insulin.

RECOMMENDATION: None.

ABSTRACT: The authors used data from the Wisconsin Epidemiologic Study of Retinopathy to create simulation models of the natural progression of retinopathy and the effects of detecting and treating retinopathy on this process. Data were modeled (by the Markov process) for three groups of 1,000 patients: younger-onset patients (under age 30 years at diagnosis) with diabetes for 5 years or more, older-onset patients (age 30 years or over at diagnosis) who required insulin, and older-onset patients who did not take insulin. The six strategies modeled for ophthalmologic care were annual or biannual use of ophthalmoscopy, annual or biannual use of fundus photography through physiologic pupil dilation (the nonmydriatic camera), and annual or biannual use of fundus photography through pharmacologic pupil dilation (mydriatic). Effectiveness was based on sight years saved with intervention. Each strategy was modeled over 10 and 60 years, with costs remaining constant; savings in years and costs were discounted by 5 percent to represent their present value. Except for the older-onset group not taking insulin, net savings for annual or biannual screening by all methods were substantial; for 10 years they ranged from $877,656 to $997,462 in the younger-onset cohort and from $19,043 to $71,986 in the older-onset group taking insulin. The younger-onset group (10-year model) saved more than 200 sight years by any of the strategies; the older-onset groups taking insulin saved 45 to 59 years; and those not taking insulin, 14 to 19 years. Sixty-year results were similar to those for 10 years; an exception was that the younger-onset cohort gained considerably more sight years and cost savings. Annual screening with fundus photography using the mydriatic camera was the most effective of the six screening strategies; however, the gain in effectiveness was small compared with annual screening with the nonmydriatic camera or with an ophthalmoscope. 7 figures, 5 tables, 32 references.

TITLE: Cost-Effectiveness of the Screening and Treatment of Retinopathy. What Are the Costs of Underutilization? Fendrick, A.M.; Javitt, J.C.; Chiang, Y.P. International Journal of Technology Assessment in Health Care. 8(4): 694-707. Fall 1992.

OBJECTIVE: To determine whether screening and treating retinopathy is cost effective in Sweden.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Annual photographic screening to detect early signs of retinal disease combined with laser photocoagulation treatment of macular edema and retinopathy leads to improved health outcomes in terms of years of sight saved and also to decreased medical expenditures.

RECOMMENDATION: None.

ABSTRACT: A computer model (PROPHET) was used to simulate the health and economic outcomes of an annual screening program for retinal disease in a group of patients with type 1 diabetes in Sweden. Many Swedish patients with type 1 diabetes do not receive optimal eye care because physicians have inconsistent practice patterns, are inexperienced in performing the eye exam, and lack knowledge of screening recommendations. Screening for retinopathy and macular edema before high-risk lesions become worse is the key to preventing vision loss. The study looked at 750 patients newly diagnosed with diabetes over a 60-year life span. Fundus photography (3 fields per eye) was used as the screening method and photocoagulation (panretinal for proliferative retinopathy and focal retinal for macular edema) was employed to treat eye disease. The model predicted that background retinopathy, a preclinical state, would be present in 98 percent of the patients after 60 years of diabetes. Macular edema and proliferative retinopathy were predicted to occur in more than 50 percent of the patients in that period. The number of person-years of vision preserved by annual screening was closely linked to patient compliance with screening recommendations. The model revealed that with a compliance rate of 70 percent, more than 45,000 screening exams were performed over the 60-year study period and 2,306 years of vision were saved. Without screening and treatment, 6,500 years of sight would be lost. The model showed that with a screening compliance rate of 60 to 100 percent, net savings, including the costs of screening and treatment, would range from $3.7 to $6 million (U.S. dollars). Although the program was costly in the beginning, its net cost decreased over time as the years of sight saved accrued and treatment sessions decreased. Sensitivity analysis was performed to evaluate variables for which there was uncertainty. In this model, screening compliance rates correlated positively with improved clinical and economic outcomes. 5 figures, 3 tables, 64 references.

TITLE: Detecting and Treating Retinopathy in Patients with Type I Diabetes Mellitus: A Health Policy Model. Javitt, J.C.; Canner, J.K.; Frank, R.G.; Steinwachs, D.M.; Sommer, A. Ophthalmology. 97(4): 483-494 (discussion: 494-495). April 1990.

OBJECTIVE: To estimate the medical and economic implications to the federal government of several screening and treatment strategies for retinopathy in patients with type 1 diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Societal.

CONCLUSION: All five retinopathy screening and treatment strategies resulted in cost savings to the federal government.

RECOMMENDATION: Public health policy must consider the enormous medical and economic benefits that can be realized by detecting and treating diabetic eye disease.

ABSTRACT: The authors used a cost-benefit model to evaluate the average net savings to the federal government of retinopathy screening and treatment in patients with type 1 diabetes. Using a PROPHET simulation system, the authors analyzed outcomes for a hypothetical cohort of 31,000 patients developing the disease at age 12.5 years; these values represent annual incidence and average age of onset for this disorder in the United States. Five screening strategies involving various schedules of dilated ophthalmoscopy with and without full fundus photographs were tested: (1) ophthalmoscopy (with the eyes dilated) every 2 years, (2) annual ophthalmoscopic exam, (3) ophthalmoscopic exam annually for patients with no retinopathy, every 6 months for those with retinopathy, (4) annual ophthalmoscopic exam with full fundus photography, and (5) annual ophthalmoscopic exam with fundus photographs for patients with no retinopathy, an exam and photos every 6 months for those with retinopathy. The discount rate was set at 5 percent; potential savings (in 1986 dollars) were based on the amounts paid by the federal government for blindness-related disability ($6,300 annually in the model). Undiscounted screening and treatment costs for the five strategies varied from $89.2 million to $290.3 million. All of the strategies resulted in substantial net annual savings to the federal government, ranging from $62.1 million to $108.6 million. The model predicted a clear medical and economic advantage for the strategy of dilated ophthalmoscopy performed annually, then semiannually upon diagnosis of retinopathy. This strategy had the second highest government savings (annual ophthalmoscopy only had the highest), but saved several thousand more person-years of sight than annual ophthalmoscopy. Positive returns were seen for all five strategies at discount rates below 10 percent. Screening and treatment were cost-saving if the value of a year of sight saved was $2,500 or more. Little advantage was demonstrated in adding routine fundus photography to screening exams. 3 figures, 8 tables, 41 references.

TITLE: Detecting and Treating Retinopathy in Patients with Type I Diabetes Mellitus: Savings Associated with Improved Implementation of Current Guidelines. Javitt, J.C.; Aiello, L.P.; Bassi, L.J.; Chiang, Y.P.; Canner, J.K. Ophthalmology. 98(10): 1565-1573 (discussion: 1574). October 1991.

OBJECTIVE: To estimate the net federal budgetary savings that might be attained with increased enrollment of patients with type 1 diabetes into appropriate ophthalmologic care.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Societal.

CONCLUSION: At current levels of screening and treatment implementation, it is less expensive to provide preventive eye care for patients with type 1 diabetes than to support subsequent disability. Because each 10 percent improvement in implementation over current levels would save $16.5 million annually, significant recruitment efforts could be undertaken without diminishing returns.

RECOMMENDATION: Public education should be expanded so that all possible patients are recruited into appropriate screening strategies, and primary care physicians must be enlisted to identify patients so that strict ophthalmologic follow-up may be instituted.

ABSTRACT: The authors analyzed the net federal budgetary savings achieved under current American Academy of Ophthalmology screening and treatment conditions for retinopathy in patients with type 1 diabetes and estimated savings that might be obtained by increased enrollment of patients into appropriate ophthalmologic care. A PROPHET modeling system, based on Monte Carlo techniques, was used to analyze events and costs. Disease data in the model were derived primarily from reports of cross-sectional studies and clinical trials. Screening and treatment costs were derived from average Medicare charges in 1990. The authors expressed costs and savings in 1990 U.S. dollars with a discount rate of 5 percent. They estimated that 60 percent of patients with type 1 diabetes receive retinopathy treatment and screening that meets American Academy of Ophthalmology guidelines. At this level, 47,374 person-years of sight are salvaged and $101.0 million saved annually. The model predicts that for every additional 10 percent of patients who enter screening, more than 7,966 person-years of sight and $16.5 million in further annual savings would be realized. With 100 percent screening, 79,236 person-years of sight and $167.0 million would be saved annually. Current American Academy of Ophthalmology guidelines recommend initiating retinopathy screening 5 years after diagnosis of type 1 diabetes. The additional annual financial burden of beginning screening upon diagnosis would be $3.03 million. However, the yearly increase in expense would be recovered totally if less than 1 additional patient were recruited to screening from every 56 patients with diabetes. The model predicts a loss of $17.4 million and 5,961 person-years of sight if treatment of retinopathy is delayed 1 year. 8 figures, 3 tables, 41 references.

TITLE: Detection of Sight-Threatening Diabetic Eye Disease. Leese, G.P.; Broadbent, D.M.; Harding, S.P.; Vora, J.P. Diabetic Medicine. 13(10): 850-853. October 1996.

OBJECTIVE: To review the feasibility and costs of screening methods to detect asymptomatic eye disease in patients with diabetes in the United Kingdom.

CATEGORY: Tertiary intervention.

Type of Study: Patient screening.
Methodology: Cost analysis.
Perspective: Societal.

CONCLUSION: Community-based screening programs employing fundus photography or slit-lamp biomicroscopy may offer a cost-effective alternative to ophthalmological screening for eye diseases in patients with diabetes.

RECOMMENDATION: Formal research-based measurements of the effectiveness of alternative eye disease screening methods are needed, and health services purchasers must be persuaded to implement cost-effective community-based screening programs.

ABSTRACT: The authors review the status of screening for sight-threatening eye disease among patients with diabetes in the United Kingdom and assess the feasibility of alternatives to ophthalmological examinations. Because the number of ophthalmologists in the United Kingdom is not sufficient to allow screening of all patients, other options must be considered. Screening for retinopathy and other eye diseases at hospital-based clinics, at general practice clinics, in optometry practices, and by mobile fundus photography is discussed. The per-patient cost of screening by direct ophthalmoscopy has been estimated at £ 13 when performed by community-based optometrists, £ 15 by general practitioners, and £ 27 by hospital physicians. Costs of fundus photography by a mobile unit have been estimated at between £ 10 and £ 23 per patient versus £ 19 per screen if the camera is maintained within a hospital. When using a mobile van, overall costs per potentially sight-saving treatment have been estimated by other researchers at £ 700 to £ 1,000. Screening by fundus photography has been demonstrated to be more cost effective than ophthalmoscopy because of its greater sensitivity. Screening costs must be evaluated in comparison with the costs of supporting a blind person in the community, which were estimated in 1981 to be £ 3,500 per year. 42 references.

TITLE: Evaluation of Argon Laser Treatment of Retinopathy and Its Diffusion in The Netherlands. Vondeling, H. Health Policy. 23(12): 97111. January 1993.

OBJECTIVE: To review studies on argon laser treatment for retinopathy, including its cost-effectiveness; to discuss the diffusion of this technology.

CATEGORY: Tertiary intervention.

Type of Study: Patient screening.
Methodology: Review of studies.
Perspective: Societal.

CONCLUSION: Screening for and treating retinopathy is cost effective.

RECOMMENDATION: None.

ABSTRACT: Clinical trial evidence indicates that immediate argon laser treatment can prevent blindness and stabilize retinopathy for at least 10 years in 70 percent of cases. Drummond et al. (1990) analyzed the U.S. Retinopathy Study and found a net savings from the trial for U.S. society over 22 years of laser photocoagulation of $2,816 million (including $2,585 million in lost production). Another model (Javitt and coworkers), for patients with type 1 diabetes, predicted a cost (in 1986 dollars) of $966 per person-year of vision saved from proliferative retinopathy and $1,118 per person-year of central acuity saved from macular edema; in contrast, average annual federal payments to eligible blind recipients were $6,900. A model of screening practices indicated that the most cost-effective method is dilated ophthalmoscopy performed annually for patients without retinopathy and every 6 months for those with retinopathy. Proper screening of 60 percent of persons with type 1 diabetes would result in annual savings of 47,374 person-years of sight and $101 million. In The Netherlands, the number of argon lasers increased from 10 prior to 1978 to 111 in 1992. Change in clinical practice in The Netherlands could have been implemented more quickly with more active governmental support. Adequate screening for diabetic eye disease is being promoted by the American Academy of Ophthalmology, the U.S. National Eye Institute, the World Health Organization, and the International Diabetes Federation. Further studies are needed to document accurately the cost-effectiveness of screening and early treatment programs. 3 figures, 21 references.

TITLE: The Evaluation of Mobile Screening for Retinopathy. Thompson, C.; Leese, G. Scottish Medical Journal. 40(1): 5-7. February 1995.

OBJECTIVE: To overview the use of mobile screening for retinopathy.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Mobile retinal cameras, which offer the chance to expand eye screening for retinopathy, particularly to remote and rural areas, should be considered a valuable addition to current methods.

RECOMMENDATION: None.

ABSTRACT: The authors review previous studies on screening for retinopathy in the United Kingdom. The ideal screening program for retinopathy, which has yet to be determined, should be technically accurate, cost effective, and applicable to the whole population. Mobile screening in the community is a possible alternative for patients who do not attend diabetes clinics and may be particularly valuable in remote rural areas, where fewer people attend specialty diabetes clinics or receive a regular ophthalmic examination. In the great majority of cases, mydriatic drops (to increase pupil size) are not required for the retinal cameras used. An evaluation of a rural mobile screening program (in Tayside) for persons with diabetes found that 20 percent of those not attending diabetic clinics had retinopathy, and 6.5 percent needed urgent ophthalmological assessment. It has been estimated that a comprehensive screening program that included detection, referral, treatment, and follow-up would reduce new blindness by 10 percent in persons under 70 years. A recent multicenter study on screening found the cost of diagnosis per true-positive case of sight-threatening retinopathy to be 33 to £ 1,079 when the screener was a general practitioner, £ 497 for a mobile community-based retinal camera, £ 1,546 for a hospital-based retinal camera, £ 1,028 for opticians, and £ 1,033 for hospital physicians. In the Tayside program, the cost to screen a patient was £ 10; the cost per case of newly discovered sight-threatening retinopathy was £ 50. The cost per patient receiving laser therapy for retinopathy was £ 1,000. A screening service's cost-effectiveness depends on the prior probability of detecting significant retinopathy; after the initial impact of the mobile camera it may drop because of the low annual incidence of this problem. The development of a strategy for identifying high-risk groups has been advocated. 24 references.

TITLE: Local Survey of Optometrists about Dilated Funduscopic Examinations for Patients with Diabetes: Making Use of Phone Book Yellow-Page Listings. Foster, D.T.; Wylie-Rosett, J.; Walker, E.A. Diabetes Educator. 22(6): 605-608. November-December 1996.

OBJECTIVE: To assess the knowledge, attitudes, and practices of optometrists in the Bronx, New York, area related to providing dilated funduscopic examination for patients with diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Patient management.
Methodology: Telephone survey.
Perspective: Health care system.

CONCLUSION: Dilated funduscopic examinations were available at a relatively modest cost in over half of the optometry practices listed in the NYNEX yellow pages for the Bronx.

RECOMMENDATION: Information is needed concerning how to increase education about dilated fundoscopic examinations in medically underserved areas such as the South Bronx. Campaigns to increase the rate of dilated funduscopy among patients with diabetes should consider how optometrists interact with the medical care system to achieve early detection of retinopathy and other diabetes-related eye problems.

ABSTRACT: The authors surveyed optometrists listed in the Bronx, New York, NYNEX yellow pages. Telephone interviews were conducted with 23 of the 31 optometry practices listed. Dilated funduscopic examinations were performed by 13 of the 23 practices. The primary contraindication to performing dilated funduscopy cited by respondents who performed the examination was narrow-angle glaucoma; hypertension was also mentioned as a contraindication (it is not considered one per se, and this problem is frequently associated with diabetes). Estimates by optometrists of the percentage of patients with diabetes who knew of the need for dilated funduscopy examinations ranged from 2 percent in the South Bronx to 25 percent for Westchester County (New York) practices. Billing charges for a general examination ranged from $12 to $55. The billing charge for dilated funduscopy ranged from no additional charge to a $27 extra charge. Of the 23 practices, Medicaid payment for examinations was accepted by 22, credit card by 20, and Medicare by 18. Performing dilated funduscopy could be a recruitment strategy for optometrists and might increase the number of patients who get this examination annually. 2 tables, 12 references.

TITLE: Meeting the Challenge of Diabetic Blindness in the 90's. Yeo, K.; Fan, R.; Yong, V. Singapore Medical Journal. 34(2): 128-130. April 1993.

OBJECTIVE: To describe a 5-year screening and education program for diabetes.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Blindness from retinopathy is largely preventable.

RECOMMENDATION: Institute a nationwide program of screening and treatment for diabetes.

ABSTRACT: The prevalence of diabetes in Singapore is rising, and retinopathy is a leading cause of blindness among Singapore adults. Hospital-based screening efforts in Singapore will reach only a small percentages of those with diabetes. Most Singapore patients with diabetes are treated in primary-level services (general practitioners, outpatient departments), where direct ophthalmoscopy is widely available. This approach, however, has important technical limitations. Many persons with diabetes do not know that diabetes might cause blindness, and those with good vision are often not motivated to seek a fundal examination. If blindness can be prevented in 10 percent of the people in Singapore who have diabetes, cost savings will be significant. A nationwide screening program for public education in diabetes; training of medical staff, nursing personnel, and volunteers; and providing adequate treatment and follow-up facilities are suggested to reduce blindness from diabetes, with initial screening targeting high-risk groups. As a first step, the authors suggest establishing a centralized screening clinic in a hospital or diabetic center coupled with a mobile screening service. They project that 16,800 patients can be screened the first year, with 10 percent increments each year, for a total of 102,564 patients for 5 years. The cost of these two programs is estimated at $150,000 (Singapore dollars) for the first year, with equipment accounting for the major expenditure in that year. The cost of screening is not excessive and is much below the cost of treating late-stage retinopathy and rehabilitating blind patients. The loss of economically productive persons with diabetes must also be considered. Education can be carried out through mobile exhibits, mass media, talks and seminars, and a diabetes education exhibit in the screening center. Additional laser facilities would have to be made available to treat patients with sight-threatening retinopathy. 8 references.

TITLE: Mobile Retinal Photography: A Means of Screening for Retinopathy in Aboriginal Communities. Karagiannis, A.; Newland, H. Australian and New Zealand Journal of Ophthalmology. 24(4):333-337. November 1996.

OBJECTIVE: To determine whether interpretable fundus photographs of the eye could be taken by specially trained aboriginal health workers in a mobile screening setting.

CATEGORY: Tertiary intervention.

Type of Study: Patient screening.
Methodology: Quality review.
Perspective: Health care system.

CONCLUSION: Of the 390 photographs taken of 47 patients known to have diabetes, 371 slides were of sufficient clarity to interpret for retinopathy. However, 158 of the 371 slides were suboptimal.

RECOMMENDATION: Mobile retinal photography carried out by aboriginal health workers as part of routine health visits could be an efficient, practical approach to eye screening of aboriginal diabetes patients living in remote areas and of monitoring patients with existing retinopathy.

ABSTRACT: The prevalence of diabetes in the aboriginal community in Australia is estimated to be up to 15 percent. Aborigines and Torres Strait Islanders living in remote areas of Australia and New Zealand have limited access to screening services for retinopathy. The investigators describe a pilot study to train aboriginal health workers to take fundus photographs of the eyes of aborigines with diabetes during routine health clinic visits. The health workers received 2 weeks of training from an ophthalmic photographer; 1 week took place on-site in an aboriginal community. The photographer then supervised the health workers on 2 clinic visits in a 6-month period in the community during which the health workers carried out dilated-eye examinations in 47 known diabetes patients. The health workers took an average of 8 photographs per patient (range: 4 to 27), for a total of 390 slides. A retinal specialist compared the slides with baseline photographs taken by the ophthalmic photographer at the beginning of each clinic visit and graded them for quality and interpretability. Nineteen slides, representing 11 patients, could not be read; 371 slides were of sufficient quality to detect significant eye disease. The unit cost per photograph was $1.00 (Australian dollars), not including the cost of the equipment, services of the retinal specialist, or mobile screening vehicle. 2 figures, 14 references.

TITLE: Ophthalmic Screening for Diabetics: The Importance of Physician-Ophthalmologist Collaboration in the Prevention of Blindness. Chew, S.J.; Hart, P.M.; Ang, B.C.; Lim, A. Singapore Medical Journal. 31(1): 26-29. February 1990.

OBJECTIVE: To evaluate the efficacy and cost-effectiveness of a screening program for retinopathy involving the coordination of primary physician and ophthalmologist services.

CATEGORY: Tertiary intervention.

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

CONCLUSION: The screening program accurately identified patients with varying degrees of retinopathy at a reasonable cost.

RECOMMENDATION: Screening for retinopathy based on fundus photography performed by primary care physicians and interpreted by ophthalmologists should be considered an accurate, cost-effective strategy for blindness prevention in patients with diabetes.

ABSTRACT: The screening program, which was offered in Singapore, included Polaroid fundus photography, noncontact tonometry, and blood pressure measurement, with primary care physicians and ophthalmologists equally responsible for patient care. Screening examinations took place at a retinal clinic, and photographs were reviewed and discussed with patients by a retinal specialist. During the first 6 months of the program, 428 patients with known diabetes were screened; retinopathy was detected in 161 eyes (18.8 percent). The prevalence of retinopathy was closely associated with duration of diabetes; age of onset was of lesser importance. Of the 161 eyes with retinopathy, 60 percent exhibited only background changes; sight-threatening retinopathy was found in 7.6 percent of all cases (59 of 856 eyes). Background retinopathy occurred in 22.5 percent of eyes among patients with type 1 diabetes and 9.7 percent of eyes among patients with type 2 diabetes. Proliferative retinopathy was 10 times more prevalent in the eyes of patients with type 1 than in the eyes of patients with type 2 diabetes. Patients were charged S$6.00 for the screening service, which included the cost of photography. 7 tables, 5 references.

TITLE: Opportunities for Cost Reduction in Retinopathy Treatment: Case Study From Mexico. Phillips, M.; del Rio, I.; Quiroz, H. Bulletin of the Pan American Health Organization. 28(1): 50-61. March 1994.

OBJECTIVE: To measure the costs of treating eye problems in patients with diabetes, to find out who bears these costs, and to determine how cost reduction could be accomplished.

CATEGORY: Tertiary intervention.

Type of Study: Patient management.
Methodology: Cost analysis.
Perspective: Societal.

CONCLUSION: The cost of treating eye problems was quite high, and the individual patient bore most of the costs for treatment. Reducing the number of visits could substantially lower the economic impact on patients.

RECOMMENDATION: Further research is needed to clarify the benefits gained by patients who have photocoagulation laser therapy.

ABSTRACT: A randomized study was conducted at the Hospital for the Prevention of Blindness in Mexico City to determine the costs of treating patients with diabetes who had retinopathy. Clinical records of a random sample of 69 patients were used to collect data on demographic and socioeconomic variables, diabetes treatment, initial eye diagnosis, and the amount and type of eye treatment provided by the hospital from 1985 to 1991. The cost of the resources used for each type of treatment provided was determined. Patient interviews provided information about treatment-associated costs incurred by patients and those accompanying them. The average age of patients was 59 years; most were poor and had little formal education. All lacked private health insurance and none was reimbursed for medical care expenses incurred. The cost of treatment per patient over 5 years was $630; the patient and family paid 83 percent of the cost. Fees accounted for 45 percent of patient costs, other direct expenses (e.g., travel and accommodations) made up another 45 percent, and lost income accounted for 10 percent. Suggestions for reducing patient costs without lowering quality of care included shortening waiting time, increasing the strength of laser treatments to decrease number of visits needed, completing more procedures in a single visit, revising the policy on fluoroangiography, and educating patients and at-risk relatives accompanying them about the need for early detection and treatment of eye problems. 5 tables, 13 references.

TITLE: Practical Community Screening for Retinopathy Using the Mobile Retinal Camera: Report of a 12 Centre Study. British Diabetic Association Mobile Retinal Screening Group. Also, Population-based Screening for Retinopathy: A Promising Start (Comment). Greenwood, R.H. Diabetic Medicine. 13(11):925-926, 946-952. November 1996.

OBJECTIVE: To report the progress of a project involving the use of mobile units equipped with retinal cameras to provide eye screening services to diabetes patients in 12 health districts in the United Kingdom.

CATEGORY: Tertiary intervention.

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

CONCLUSION: The mobile units performed 64,905 screenings; analysis of 42,803 screenings found 2,400 referrals for further evaluation, of which 516 resulted in immediate laser therapy for sight-threatening disease. During subsequent years of this multiyear project, the number of patients referred and those needing laser therapy declined.

RECOMMENDATION: Mobile eye screening programs should be designed to fit the specific needs of each district, retinal photographers should be personable and well trained, and close liaisons with local ophthalmologists should be established.

ABSTRACT: In 1989, the British Diabetic Association initiated a mobile eye screening program designed to reduce morbidity from diabetic eye disease, the leading cause of blindness in working-age adults. Twelve health centers, representing urban and rural areas, participated in the program. The vans were equipped with retinal cameras, and each center was responsible for training an operator/driver, determining how the screening service would be used in the district, and setting up systems for reporting results. Most screenings (76.5 percent) were carried out in primary care settings; the remainder were carried out in hospital-based settings. The average cost per patient screened and per patient treated was £ 13.11 and £ 1,110, respectively. Patient acceptance of the screening process was high, and in 10 of 12 districts financial responsibility for the program has been taken over by hospital trusts or district health authorities. The investigators conclude that the use of mobile vans for retinal screening is cost effective and efficient. An accompanying "Comment" notes that performance standards need to be developed by a central organization such as the British Diabetic Association before the program can be implemented on a wider scale. A table of proposed standards developed by the British Diabetic Association Retinal Screening Group is presented in the original article. The article also includes an appendix listing reports from district units. 7 tables, 1 appendix, 19 references in principal article; 15 references in Comment.

TITLE: Preventive Eye Care in People with Diabetes Is Cost-Saving to the Federal Government: Implications for Health Care Reform. Javitt, J.C.; Aiello, L.P.; Chiang, Y.; Ferris, F.L.; Canner, J.K. III; Greenfield, S. Diabetes Care. 17(8): 909-917. August 1994.

OBJECTIVE: To estimate savings to the federal government from screening and treatment of retinopathy in patients with type 2 diabetes.

CATEGORY: Tertiary intervention.

Type of Study: Epidemiological cohort model.
Methodology: Cost-benefit analysis.
Perspective: Societal.

CONCLUSION: Improvement in ophthalmologic screening of patients with type 2 diabetes will reduce vision loss and save money for the federal government.

RECOMMENDATION: Eye care for patients with type 2 diabetes must emphasize patient identification, careful follow-up, and prompt, appropriate treatment.

ABSTRACT: The authors used the PROPHET computer model system to project the incidence and the costs of screening and treatment of eye disease resulting from diabetes. PROPHET is designed to model the course of a chronic, irreversible disease. Incidence data for type 2 diabetes were applied to 1990 Medicare, Social Security Disability Insurance, and Social Security Insurance costs for screening and treatment; a 5 percent discount rate was used. Costs of screening were based on complete dilated-eye examinations ($62 each) at diagnosis and every 2, 3, or 4 years or, for those with retinopathy, every 6, 12, 18, or 24 months. Treatment (photocoagulation for both eyes) cost, including fluorescein angiograms, was $1,980. Based on 1988 population figures for the United States, 576,136 patients yearly develop type 2 diabetes. Blindness in patients with diabetes costs the federal government $14,296 annually per patient under age 65; per patient aged 65 and over, federal expenditures are $32 annually (does not include Medicare/Social Security payments or income tax exemption). Screening and treatment for eye disease in patients with type 2 diabetes saves 53,986 person-years of sight at an annual federal budget savings of $247.9 million; these results assume that 60 percent of type 2 patients receive appropriate eye care. For patients with type 2 diabetes controlled by insulin, however, savings would be $1,715 per person, versus $725 for those controlled by other means. Patients with onset prior to age 45 account for 89.1 percent of sight savings and 100 percent of cost savings. With recommended eye care, 112,730 and 94,304 person-years of sight and $624 and $472.1 million would be saved annually in all patients with diabetes and those with type 2 diabetes, respectively. 4 figures, 2 tables, 52 references.

TITLE: A Relative Cost-Effectiveness Analysis of Different Methods of Screening for Retinopathy. Sculpher, M.J.; Buxton, M.J.; Ferguson, B.A.; Humphreys, J.E.; Altman, J.F.; Spiegelhalter, D.J.; Kirby, A.J.; Jacob, J.S.; Bacon, H.; Dudbridge, S.B.; Stead, J.W.; Feest, T.G.; Cheng, H.; Franklin, S.L.; Courtney, P.; Talbot, J.F.; Ahmed, R.; Dabbs, T.R. Diabetic Medicine. 8(7): 644650. August/September 1991.

OBJECTIVE: To determine the relative cost effectiveness of various screening methods for sight-threatening retinopathy in terms of cost per true-positive case detected.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Cost per true-positive case ranged from £ 633 to £ 1,079 for general practitioners; it was £ 1,033 for a hospital physician; £ 784 for an ophthalmic optician; £ 441 to £ 609 for an ophthalmological clinical assistant; £ 601 to £ 1,546 for a hospital-based camera; and £ 497 to £ 747 for a traveling camera. Except for the ophthalmological clinical assistant (the reference standard), the sensitivity of all screening methods was low (0.35 to 0.67). Relative Cost-effectiveness changes if the screening can take place without requiring an additional patient visit and is strongly related to the relative sensitivity of the screening methods and to the prior probability (prevalence or incidence) of retinopathy in the population with diabetes.

RECOMMENDATION: Approaches to screening, which are discussed in this report, may improve sensitivity without reducing specificity or increasing cost per true-positive case substantially.

ABSTRACT: The authors report on screening for sight-threatening retinopathy of five patient groups (n = 3,318) in three British centers. Patients' fundi were assessed by a primary screener (a hospital physician, a general practitioner, or an ophthalmic optician) using ophthalmoscopy with mydriasis. Additionally, all patients had their fundi photographed by a nonmydriatic fundus camera, and they received an ophthalmoscopic examination with mydriasis by an ophthalmological clinical assistant (the reference standard). The costs per true-positive case for the primary screeners ranged from £ 633 to £ 1,079 ( £ 1,033 for the hospital physician), for the clinical assistant from £ 441 to £ 609, and for photography from £ 497 for a camera that is taken to general practices in one center to £ 1,546 for a hospital-based camera. The cost for true-positive case for hospital physicians would drop from £ 1,033 to £ 353 if an additional visit to the hospital were not required. Similarly, if ophthalmoscopy is part of a general assessment by a general practitioner rather than requiring an additional visit, the cost per true-positive case would drop dramatically (to £ 245 to £ 362). Total costs per patient screened by primary screeners ranged from £ 19.31 for ophthalmic opticians to £ 37.77 for hospital physicians. The low sensitivities of primary screeners and of photography (35 to 67 percent) may indicate that none of these methods would be acceptable in routine clinical practice, despite their relatively high specificities (86 to 98 percent). Alternatives that may improve sensitivity (while avoiding a reduction in specificity or increase in cost per true-positive case detected) include (1) clinical assistant screening of both hospital and community-based patients; (2) use of combined screening strategies (e.g., general practitioner, ophthalmoscopy and fundus photography); and (3) identification of risk factors for retinopathy that could be used, exclusively or in combination with single or joint screening methods, as forms of screening in themselves (e.g., patients with 10+ years of type 1 diabetes could be referred directly to an ophthalmologist without prior screening). 1 figure, 3 tables, 25 references.

TITLE: Retinopathy in the West of Scotland: Its Detection and Prevalence, and the Cost-Effectiveness of a Proposed Screening Programme. Foulds, W.S.; McCuish, A.; Barrie, T.; Green, F.; Scobie, I.N.; Ghafour, I.M.; McClure, E.; Barber, J.H. Health Bulletin. 41(6): 318-326. November 1983.

OBJECTIVE: To assess the prevalence of retinopathy as diagnosed by ophthalmoscopy in the West of Scotland and to evaluate the financial implications of implementing a screening program for retinopathy in that part of Scotland.

CATEGORY: Tertiary intervention.

Type of Study: Patient screening.
Methodology: Cost-effectiveness analysis.
Perspective: Societal.

CONCLUSION: A projected screening program for retinopathy appeared to be cost savings in terms of costs of patient identification and screening relative to savings associated with the prevention of blindness.

RECOMMENDATION: Universal screening for early detection of serious retinopathy should be performed.

ABSTRACT: The authors assessed the prevalence of retinopathy in the West of Scotland. Based on ophthalmoscopic examination of 1,147 patients with diabetes, the authors estimated the prevalence of retinopathy to be 26 to 35 percent; of serious retinopathy, 9.5 to 11 percent. The potential cost-effectiveness of a proposed screening program involving annual ophthalmoscopic examination of all patients with diabetes in the West of Scotland (population about 2.5 million) was assessed. Projected annual cost associated with physician/ophthalmologist examination plus nurse time was £ 51,800. Total annual cost to identify those patients with serious retinopathy was estimated to be £ 55,300 , or £ 183 per patient with serious retinopathy identified. Total cost per annum of identifying and treating patients at risk for blindness was estimated to be £ 387 per patient treated. Costs associated with identifying and treating a backlog of patients during the first year of the program were estimated to be £ 86 per patient treated. Projected savings in blind welfare services and state benefits per case of prevented blindness were calculated to be £ 3,575. It is estimated that 60 percent of blindness from retinopathy could be prevented by appropriate laser therapy. Total savings to the state per annum in treating and preventing blindness in such a percentage of at-risk patients were estimated to be £ 193,050. One-time savings associated with treating the backlog of patients with serious retinopathy were estimated to be £ 1.6 million. Appropriate laser therapy was estimated to result in a net savings of £ 135,025 annually based on projected treatment costs and savings for the state associated with blindness prevention. In addition, the 90 patients annually prevented from going blind would be expected to earn £ 4.67 million. 1 table, 10 references.

TITLE: Retinopathy — Need and Demand for Photocoagulation and Its Cost-Effectiveness: Evaluation Based on Services in the United Kingdom. Savolainen, E.A.; Lee, Q.P. Diabetologia. 23(2): 138-140. August 1982.

OBJECTIVE: To assess both need for photocoagulation and the cost-effectiveness of using this therapy in patients with retinopathy.

CATEGORY: Tertiary intervention.

Type of Study: Patient management.
Methodology: Patient care model.
Perspective: Societal.

CONCLUSION: The cost of photocoagulation and follow-up is less than the indirect costs to maintain a blind person for 1 year.

RECOMMENDATION: Criteria for the need for photocoagulation are urgently needed.

ABSTRACT: The authors used published literature from 1962 to 1978 to estimate the number of patients with diabetes needing photocoagulation for retinopathy in two regional Health Authorities in England. They also reviewed case notes (only 141 of 272 sampled were available) from 9 of 10 photocoagulation centers in the region. Interviews were conducted with consultants at the centers, and information on manpower, equipment, and practices with different patient groups was obtained to develop a model for estimating consultant hours and treatments needed per year. It was estimated that 10,608 eyes were in need of photocoagulation, which would require 21,417 consultant hours per year and 14,496 treatments. The estimated number of actual treatments for 1979 was only 3,080. The annual outpatient cost for photocoagulation and follow-up was £ 100 per patient ( £ 170 in 1981 to 1982 prices). The cost of maintaining one blind person for 1 year (considering lost earnings and Social Security payments) was estimated to be £ 1,751 ( £ 2,871 at the end of 1981). As two of the nine centers in the sample accept patients from outside the regions, these data indicate that fewer than 20 percent of the patient need for photocoagulation was met. Patients may be undiagnosed or diagnosed too late for treatment to be effective. Criteria are needed to guide early detection, adequate treatment, and follow-up of retinopathy in patients with diabetes. Interested physicians, and possibly paramedical staff, will require training in photocoagulation to meet the need for this service. 1 figure, 1 table, 10 references.

TITLE: Screening for Retinopathy in a Clinical Setting: A Comparison of Direct Ophthalmoscopy by Primary Care Physicians with Fundus Photography. Griffith, S.P.; Freeman, W.L.; Shaw, C.J.; Mitchell, W.H.; Olden, C.R.; Figgs, L.D.; Kinyoun, J.L.; Underwood, D.L.; Will, J.C. Journal of Family Practice. 37(1): 49-56. July 1993.

OBJECTIVE: To compare the accuracy and cost-effectiveness of two approaches to screening for retinopathy in a clinical setting: (1) ophthalmoscopy by trained primary care physicians followed by referral to ophthalmologists as indicated, and (2) seven-view nonstereoscopic, mydriatic fundal photographs read by general ophthalmologists and retinal specialists.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Dilated ophthalmoscopic screening by primary care physicians followed by referral to an ophthalmologist if indicated was at least as accurate as and more cost effective than nonstereoscopic mydriatic fundus photographs read by ophthalmologists in screening for retinopathy.

RECOMMENDATION: Future studies should measure the absolute sensitivity, specificity, and predictive value of referral decisions for various diabetic retinal screening strategies and should estimate their costs. Similar trials should be conducted in other clinical settings to assess physicians' referral decisions.

ABSTRACT: The two screening strategies were implemented at a rural clinic in Toppenish, Washington, that served more than 400 Native Americans with diabetes. During the 2.5-year study period, 243 clinic visits were recorded; 93 referrals were made, of which 83 were completed. The primary care physicians were first given a 2-hour update about retinopathy and their role as screeners; they were told to refer every patient with marked retinopathy to an ophthalmologist. The "primary physician method" referred all 17 patients ultimately diagnosed with significant retinopathy. Estimated maximum sensitivity in diagnosing retinopathy was 100 percent for primary physicians; for the general ophthalmologists and retinal specialists reading photographs, it was 94 and 100 percent, respectively. Estimated maximum specificity was 93 percent for the primary physician, 82 percent for the general ophthalmologists, and 64 percent for the retinal specialists. Projected costs, including personnel and material costs plus examination charges (according to the American Academy of Ophthalmology), for screening and diagnosing 100 patients by these methods were $3,132 for ophthalmoscopic screening by primary providers, $4,942 to $5,734 for screening by retinal photography, and $8,800 for referring all patients for full annual examination by an ophthalmologist. Transportation, training, equipment, and other direct or indirect costs were not included. 3 tables, 34 references.

TITLE: Screening for Retinopathy in South Africa with 60° Retinal Colour Photography. Joannou, J.; Kalk, W.J.; Mahomed, I.; Ntsepo, S.; Berzin, M.; Joffe, B.I.; Raal, F.J.; Sachs, E.; Van Der Merwe, M.T.; Wing, J.R. Journal of Internal Medicine. 239(1): 4347. January 1996.

OBJECTIVE: To assess the use of a 60° mydriatic fundal camera to screen for retinopathy.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Mydriatic retinal photography with a 60° field was more sensitive and diagnostically more accurate than funduscopy by clinic doctors and compared well with screening by an ophthalmologist. Screening for treatable retinopathy by 60° mydriatic retinal photography is likely to be cost effective.

RECOMMENDATION: Both retinae should be screened.

ABSTRACT: Patients attending a diabetes clinic in South Africa were screened for retinopathy by mydriatic fundal photography with a 60° camera. Selected eyes were evaluated by an ophthalmologist. Randomized photographs were assessed through single or two overlapping 45° fields (by masking the slides) and at 60°. The authors found that 92 percent to 94 percent of photographs were diagnostically useful (80 percent were excellent quality), which compares well with nonmydriatic cameras. Compared with an ophthalmologist's assessment, retinal photography had a sensitivity of 93 percent and a specificity of 89 percent for any retinopathy, and 100 percent and 75 percent, respectively, for severe retinopathy. Funduscopy missed 28 percent of affected eyes, compared with only 5.5 percent missed by photography. A single 45° field missed 31 percent and two overlapping 45° fields missed 11 percent of retinopathy as compared with that detected by a 60° field camera. For the 122 eyes assessed at the three field areas, the mean scores increased significantly as the field area increased (p < 0.0001 for each comparison). The costs of screening were calculated from the price of film and processing (but not the camera) and from related staff salaries for the first 663 patients screened. The basic expenditure (in U.S. dollars) was determined to be $5.85 per patient screened, $13.55 per patient with retinopathy (n = 286), and $37.03 for each patient referred for formal ophthalmological assessment (n = 103, 15.5 percent). 2 figures, 1 table, 30 references.

TITLE: Use of Mobile Screening Unit for Retinopathy in Rural and Urban Areas. Leese, G.P.; Ahmed, S.; Newton, R.W.; Jung, R.T.; Ellingford, A.; Baines, P.; Roxburgh, S.; Coleiro, J. British Medical Journal. 306(6871): 187-189. January 16, 1993.

OBJECTIVE: To compare the rate of retinopathy detected by a mobile screening unit equipped with a nonmydriatic Polaroid between rural and urban areas; to identify the cost associated with the service.

CATEGORY: Tertiary intervention.

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

CONCLUSION: Mobile eye-screening units seem to be particularly effective at identifying previously unrecognized advanced retinopathy in rural patients with diabetes; these patients were more likely than urban patients to need urgent laser photocoagulation. The cost per patient with mobile eye screening units was relatively low.

RECOMMENDATION: Patients with diabetes living in rural areas are less likely to seek help at hospital clinics and would benefit from an expanded mobile-eye-screening program to detect and treat retinopathy.

ABSTRACT: Researchers compared the effective-ness of screening for retinopathy using mobile eye-screening units in rural versus urban patients with diabetes during 2 years in the Tayside region of Scotland. They estimated that 64 to 77 percent of the population with diabetes was screened: 1,225 urban and 961 rural patients were photographed with a nonmydriatic fundal camera. Rural patients were less likely to attend a hospital clinic than urban patients (46 percent versus 86 percent, p < 0.001) and were less likely to be receiving insulin (27 percent versus 34 percent, p < 0.001). Advanced retinopathy was greater among rural than urban patients (13 percent versus 7 percent, p < 0.001), and more rural patients required urgent laser photocoagulation (1.4 percent versus 0.5 percent, p < 0.02). Direct and indirect costs, estimated for screening 1,800 patients a year, included the salary of the ophthalmic photographer; the purchasing and processing of film; and the servicing, running, and depreciation costs of the van and camera. The cost of the screening program per patient was £ 10, which is cheaper than all alternatives. This cost is equivalent to £ 350 per patient with previously unrecognized disease and £ 1,000 per patient receiving laser treatment. These costs are low compared with alternatives and could be further reduced by screening more patients per unit. 5 tables, 23 references.

Back to Economics of Diabetes Mellitus Indexes

Top of Page