Incremental Cost Effectiveness of Laser Photocoagulation for Subfoveal Choroidal Neovascularization

Transcription

1 Incremental Cost Effectiveness of Laser Photocoagulation for Subfoveal Choroidal Neovascularization Gary C. Brown, MD, MBA, 1,2 Melissa M. Brown, MD, MN, MBA, 2,3 Sanjay Sharma, MD, MSc, 2,4 Heidi Brown, 2 William Tasman, MD 1 Objective: To perform a patient preference-based, incremental cost-effectiveness analysis for laser treatment of subfoveal choroidal neovascularization. Design: Computer-based econometric modeling. Methods: The cost effectiveness of laser photocoagulation therapy was compared with the natural course of subfoveal choroidal neovascularization. The model applied long-term visual data from previous clinical trials, utility analysis (which reflects patient perceptions of quality of life associated with a health state), decision analysis with Markov modeling, and the economic principles of present value analysis with discounting to account for the time value of money. Database: Data from patients eligible for treatment of subfoveal choroidal neovascularization obtained by researchers in the Macular Photocoagulation Study were used for the analysis. Intervention: Modeled laser therapy for subfoveal choroidal neovasacularization in patients with age-related macular degeneration. Main Outcome Measure: Cost per quality-adjusted life-year ($/QALY gained) associated with laser therapy. Results: Laser photocoagulation therapy for subfoveal choroidal neovascularization, as compared with no treatment, resulted in a mean gain of QALYs per treated patient. Using a yearly discount rate of 3% to account for the time value of money and inflation, the resultant $/QALY gained was $5629. Sensitivity analysis used in the cost-effectiveness analysis resulted in a $/QALY gained of $4974 with no gained discount rate and $11,633 with a yearly discount rate of 10%. Conclusions: The incremental expense of laser therapy for the treatment of subfoveal choroidal neovascularization appears to be highly cost effective. The result, which takes into account patient preference-based utility data, compares quite favorably with other interventional therapies across different medical specialties. Ophthalmology 2000;107: by the American Academy of Ophthalmology. A large clinical trial performed by the Macular Photocoagulation Study Group has demonstrated that laser photocoagulation for subfoveal choroidal neovascularization has a beneficial therapeutic effect. 1,2 Other therapies, such as subfoveal surgery 3 and photodynamic therapy, 4 have also demonstrated potential benefit for the treatment of subfoveal choroidal neovascularization, but do not have the long-term, clinical trial followup of the subfoveal laser study. 1,2 Originally received: May 19, Accepted: March 22, Manuscript no Retina Vascular Unit, Wills Eye Hospital, Jefferson Medical College, Philadelphia, Pennsylvania. 2 Center for Evidence-Based Health Care, Flourtown, Pennsylvania. 3 Cataract and Primary Eye Care Service, Wills Eye Hospital, Jefferson Medical College, Philadelphia, Pennsylvania. 4 Departments of Ophthalmology and Epidemiology, Queen s Medical College, Kingston, Ontario, Canada. Supported in part by the Retina Research and Development Fund, Philadelphia, Pennsylvania, and the Principal s Initiative Research Fund, Kingston, Ontario, Canada. Reprint requests to Gary C. Brown, MD, MBA, Center for Evidence-Based Health Care, Suite 210, 1107 Bethlehem Pike, Flourtown, PA Eyes with subfoveal choroidal neovascular membranes no larger than 3.5 standard disc areas that were treated with laser photocoagulation had both mean improved visual acuity and better contrast threshold for larger letters as compared with control eyes randomized to no treatment. 1,2,5 The mean visual acuity for eyes with treated subfoveal lesions was 20/320 at 48 months after therapy, whereas the mean visual acuity in eyes with no laser therapy was 20/ Although visual acuity and contrast sensitivity were improved in the laser-treated group, these changes have not been related to a quantitative improvement in quality of life. Additionally, the cost effectiveness of such therapy is uncertain. Utility theory and its incorporation into medical treatments takes into account patient preferences with regard to improvement of quality of life associated with a health state. By convention, a utility value of 1.0 is assigned to perfect health and a utility value of 0.0 is assigned to death. 6 8 The closer to 1.0, the better the quality of life associated with a health state and the better a person is able to deal with the activities of everyday life. The closer to 0.0, the poorer the quality of life associated with a health state. As examples, a by the American Academy of Ophthalmology ISSN /00/$ see front matter Published by Elsevier Science Inc. PII S (00)00169-X

2 Brown et al Subfoveal Choroidal Neovascularization person with mild angina may have a utility value of.90, 4 whereas a person with a recurrent severe stroke may have a utility value of It has been demonstrated in a large series of 325 patients that utility values associated with visual loss are proportional to the degree of loss of vision in the better-seeing eye. 10 The poorer the vision in the better-seeing eye, the lower the associated utility value. The association of utility values with visual acuity levels now allows an objective measure of patient preference-based improvement in quality of life associated with interventional ophthalmologic therapies. This improvement in quality of life can then be incorporated with discounted costs to yield cost-effectiveness for a specific therapy. 6,7,11 13 Although laser therapy for subfoveal choroidal neovascular membranes appears to yield a better long-term visual outcome as compared with no treatment, patients who are treated actually have a worse mean visual acuity in the treated eye at 3 months after treatment than those who are not treated. 1 Furthermore, controversy exists as to whether an improvement in long-term mean vision from 20/500 (untreated subfoveal net group) to 20/320 (treated subfoveal net group) is truly meaningful with regard to improvement in quality of life and cost effectiveness. The present authors therefore undertook a cost-effective analysis, incorporating patient preference-based utility values for laser therapy to treat subfoveal choroidal neovascularization. Materials and Methods Overview Software produced by Treeage Inc. 14 was used to perform decision analysis 15 to simulate the clinical situation involving laser therapy and no therapy for the eyes of patients with subfoveal choroidal neovascularization. Markov modeling, which incorporates matrix algebra to evaluate recurrent risk, was also used in the analysis. 16 The microcomputer model takes into account the mortality within the group of patients with subfoveal choroidal neovascular membranes, as well as the treatment results expressed in utility value form. The numbers of unilateral and bilateral cases are also factored into the analysis, because utility results differ dramatically, depending on whether one or both eyes are involved by the neovascular process. Utility values were used at the terminal nodes at the right side of the decision tree to provide the most probable utility value outcomes given the different possible clinical scenarios. A copy of the Markov portion of the decision tree is shown in Figure 1. The clinical data used for the analysis are shown in Table 1, and the economic assumptions undertaken in creating the model are shown in Table 2. A more complete explanation of these variables follows below. Patient Sample Patient data used in the present analysis were taken from those obtained by the Macular Photocoagulation Group for 371 patients who were randomized to either laser treatment or no treatment for subfoveal choroidal neovascularization. 1,2,17 Per the criteria of the Macular Photocoagulation Study Group, 1,2,5 randomized patients demonstrated the following characteristics: (1) subfoveal choroidal Figure 1. Markov modeling employed for after year four following randomization to laser or no laser therapy. M Markov node. neovascularization measuring 3.5 standard disc areas (1 disc area 1.77 mm 2 ) and documented by fluorescein angiography within 48 hours after randomization to treatment of no treatment, (2) age 50 years, (3) no prior photocoagulation to macula of study eye, (4) no other eye disease that could compromise visual acuity, and (5) no current or past use of systemic corticosteroids. It should be noted that, because of the smaller numbers in the subgroups, the present cost-effectiveness analysis included the overall subfoveal group 3.5 standard disc areas in size, rather than subdivisions of the subfoveal group according to initial lesion size and visual acuity. 18 Each eye randomized to treatment received either argon green or krypton red laser therapy to the area of subfoveal choroidal neovascularization. The burns were required to be intense and to extend for 100 m beyond the perimeter of the lesion. 1,5 Extensive followup information was available for 4 years after randomization to therapy or no therapy. 2 The mean age of patients who were randomized was 74.4 years. Mortality and Morbidity It was assumed that patients who were randomized to treatment or observation of subfoveal choroidal neovascularization had a normal life expectancy, although confirmatory data are not available for this cohort. Data on life expectancy were obtained using 1994 statistics obtained from the Centers for Disease Control and Prevention/National Center for Health Statistics. 19 It was also assumed that laser therapy had no effect on life expectancy. Complications from laser therapy that might have affected long-term visual acuity results were already factored into the data reported by the Macular Photocoagulation Study Group papers. 1,2 Visual Results Visual results were available for 4 years after laser therapy (Table 1). 1,2 The mean visual acuities were similar in the control and treatment groups at the time of randomization at 20/125. At 3 months after randomization, the mean visual acuity in the control group was 20/200, whereas in the treatment group it decreased to 20/320. At both the 6-month and on yearly intervals, the mean visual acuity in each group was approximately 20/320. At 2 years after randomization, however, the mean visual acuity in the treatment group was 20/320, and in the control group it had decreased to 20/400. The mean vision in each group remained stable at 3 years after randomization, but by 4 years after randomization the vision in the control group had decreased to 20/500. In the treatment group, it remained 20/

3 Table 1. Clinical Data Used in the Incremental Cost-Effectiveness Model for Treatment of Subfoveal Choroidal Neovascularization Eligibility criteria for treatment: 1,5 Subfoveal choroidal neovascular membrane no larger than 3.5 standard disc areas (1 standard disc area 1.77 mm) Best corrected visual acuity in eye of no better than 20/40, but no worse than 20/320 (Snellen equivalents) Age 50 yrs No prior macular photocoagulation in treated eye No other eye disease that could compromise visual acuity Twenty-eight percent of patients with a subfoveal choroidal neovascular membrane (CNVM) have a classic choroidal neovascular membrane or scar in the opposite eye when initially seen. 1 By the end of 5 years after a patient with a subfoveal CNVM is initially seen, the second eye will be involved by a neovascular process in 48% of cases. Thus, in approximately 4%/year of the initial cohort, a CNVM will develop in the second eye over a 5-year period. 17 The mean visual acuities in treated and untreated eyes at times after initial examination are as follows: 1,2 Time (yrs) Treated Untreated 0 20/125* 20/125* /320 20/ /320 20/ /320 20/ /320 20/ /320 20/ /320 20/500 *Approximation from weighted visual acuity averages. 1 Ophthalmology Volume 107, Number 7, July 2000 It is assumed that the same vision difference holds in the treatment group and the control group indefinitely after the end of year 4. Mean age at the time of initial examination is 74.4 years. 1 The mortality rate for persons aged 74.4 is approximately 6%/year. 18 The following utility values are associated with the following visual acuities in the better-seeing eye: 10 Visual Acuity Utility Value 20/ / / / / Eyes with unilateral CNVMs at the time of initial examination have approximately 20/40 vision in the opposite eye (weighted average data). 1 In patients with subfoveal choroidal neovascularization in one eye in whom a CNVM develops in the second eye, the second eye has similar clinical characteristics and a course to that of long-term, untreated subfoveal choroidal neovascularization (mean visual acuity of 20/500). In approximately 10% of eyes in the remaining cohort with a unilateral subfoveal CNVM, a subfoveal CNVM will develop in the second eye per year starting at 4 years (mean patient age, 78.4 years) after initial examination. 1,2,17 After the end of year 4, it was assumed that the mean vision indefinitely remained 20/320 in the treated eyes and 20/500 in the untreated eyes. 1,2 The mean visual acuity in opposite eyes without exudative macular degeneration was assumed to be 20/40, 1 and remain so indefinitely unless the eye converted to exudative macular degeneration. When a second eye converted to exudative macular degeneration, it was assumed that the visual acuity in this eye dropped to a mean of 20/500, the natural long-term course of vision in untreated eyes with subfoveal choroidal neovascularization. Table 2. Economic Assumptions Used in the Incremental Cost-Effectiveness Model for Treatment of Subfoveal Choroidal Neovascularization Markov modeling assumes 7 yearly cycles after 78.4 years (4 years after randomization for subfoveal choroidal neovascularization). Thus the model takes into account the number of quality-adjusted life-years (QALYs) gained during the mean remaining life expectancy of a person treated at 74.4 years of age. A discount rate of 3%/year is used uniformly over the remaining years of life expectancy Incremental costs for laser therapy include: 22,23 CPT Code Initial outpatient consultation $126 CPT Code Photocoagulation destruction of localized lesion of choroid $733 CPT Code Fluorescein angiography with interpretation and report. An initial intravenous fluorescein angiogram to ascertain whether treatable subfoveal choroidal neovascularization is present is not included as an incremental cost. Additional intravenous fluorescein angiograms obtained at 2 to 4 weeks after treatment and 4 to 6 weeks after treatment are included. The cost is therefore 2 $94 $188 Thus the total incremental costs associated with laser therapy are: $126 $733 (2 $94) $1,047 in 1999 U.S. nominal dollars. CPT Current Procedural Terminology. 1376

4 Brown et al Subfoveal Choroidal Neovascularization Bilaterality versus Unilaterality of Exudative Macular Degeneration At the time of entrance of patients in the Subfoveal Macular Photocoagulation Study, 1 approximately 28% of patients had involvement of both eyes by exudative macular degeneration. Approximately 10% of the remaining unilateral cases per year converted to bilateral exudative macular degeneration during a 4-year followup period. 2 It was assumed for the Markov modeling that this 10% conversion rate from unilateral to bilateral exudative macular degeneration occurred indefinitely after year 4 after randomization as well (Fig 1). The assumption was also made that when exudative macular degeneration occurred in the second eye, it occurred subfoveally (as in the first eye) and also yielded the same visual result as the untreated natural course of subfoveal choroidal neovascularization. Conversion of Visual Acuity to Utility Values It has been noted that the mean utility values of individuals with visual loss resulting from age-related macular degeneration and other ocular diseases decrease in direct proportion to the severity of visual loss in the better seeing eye. 8,20,21 It appears that the visual acuity level itself, rather than the cause of visual loss (e.g., diabetic retinopathy, age-related macular degeneration, cataract, etc.) is most closely associated with a utility value. The utility values applicable to mean visual acuity levels associated with the subfoveal macular degeneration patients used in the present analysis are shown in Table They were based on data obtained from a cohort of 325 patients with visual loss and range from 0.80 when the visual acuity is 20/40 in the best seeing eye to 0.53 when the visual acuity in the best seeing eye is 20/500. Medical Costs The costs obtained from this analysis were representative of those paid by the Health Care Finance Agency for provider services classified according to Current Procedural Terminology (CPT) data. 22 The cost data used were from urban areas of Pennsylvania. 23 The incremental costs associated with subfoveal laser therapy included the expense to the payer of an initial outpatient consultation (CPT code 99254), destruction of localized lesion of the choroid (CPT code 67220), and fluorescein angiography with interpretation and report (CPT code 92230) two times. An initial fluorescein angiogram to confirm the diagnosis of a treatable subfoveal choroidal neovascular membrane was not included in the cost analysis, because it is not necessarily an incremental cost. It is recommended, however, that fluorescein angiography be repeated at 2 to 4 weeks and again at 4 to 6 weeks after laser therapy 24 ; thus the two intravenous fluorescein angiograms. Because there is a 3-month postoperative period after laser therapy (CPT code 67220), any laser retreatment given during this period is not billable and was not considered an incremental cost. Furthermore, because the authors are unaware of clinical trial data suggesting that retreatment of the type of subfoveal choroidal neovascular membranes analyzed in this report is of clinical benefit, the costs associated with additional laser therapy after the 3-month postoperative period were not included either. Of additional note is the fact that costs such as those of traveling for treatment were not included. The total costs are listed in Table 2. The 1999 U.S. nominal dollar costs were $1047. Table 3. Net Gain (or Loss) of Quality-Adjusted Life-Years (QALYs) after Laser Therapy for Subfoveal Choroidal Neovascularization Time period Net Utility Change QALYs 0 3 mos (0.015) (0.004) 3 mos 1 yr yrs yrs yrs yrs (Markov) Variable Total no. QALYs gained Markov Markov modeling. Present Value Analysis The benefits derived from laser therapy for subfoveal choroidal neovascularion are spread out over the remaining lifetime of the treated individual, but the dollars required to deliver the laser therapy are expended early during the course of the disease, typically within the first 3 months after diagnosis. Therefore, some compensatory mechanism must be instituted to incorporate the time value of money. The time value of money accounts for the fact that the real value of a dollar changes with time. 25,26 A dollar spent today could otherwise be reinvested to yield more dollars over time. Thus a nominal dollar spent in the year 1999 will purchase more goods and services than a nominal dollar spent in the year To account for the changing time value of money, the authors discounted the quality-adjusted life-years (QALYs) obtained yearly from the therapy, over the life expectancy of the average individual, to their present value. Alternatively, the future value of the dollars spent at the time of therapy also could have been calculated over the mean life expectancy of the patients (leaving the initial QALYs derived a constant) to arrive at a similar analysis. The discount rate of choice is variable, depending on the discretion of investors or other interested stakeholders. For the purpose of the present analysis, the authors chose to use a discount rate representative of the yearly return on funds that an investor could obtain from the safest investment of capital over and above the yearly inflation level. This accounts for the Fisher effect, 26 which adjusts the nominal return on investment for inflation, thereby yielding the real return. The approximate nominal return on, among the safest of investments, a United States Government 1-year Treasury Bill at the time of the preparation of this manuscript was 4.5%/year, 27 and the current inflation rate in the United States was approximately 1.48%/year. 28 Thus the yearly discount rate, or return on investment in real dollars used in our analysis, was 1.045/ , or 3%/year. Results Decision Analysis and Markov Modeling When the utility change attributable to an interventional therapy is multiplied by the time over which the therapy has a benefit or detriment, the number of QALYs gained or lost from the treatment is the result. 6 9,29 Because the visual acuity as a result of both the natural course of, as well as of laser therapy for, subfoveal choroidal neovascularization varies during the first 4 years of treatment, the number of QALYs gained or lost during this time are shown in Table 3. Of note is the fact that at 3 months after 1377

5 Ophthalmology Volume 107, Number 7, July 2000 Table 4. Sensitivity Analysis Performed on No. of QALYs Gained by Laser Therapy for Subfoveal Choroidal Neovascularization at Various Yearly Discount Rates Discount Rate (%) No. QALYs Gained Table 5. Sensitivity Analysis of $/QALY Gained by Laser Treatment of Subfoveal Choroidal Neovascularization at Various Yearly Discount Rates Discount Rate (%) $/QALY Gained ,633 QALY quality-adjusted life-years. $/QALY incremental cost per quality-adjusted life-years. randomization for treatment, the mean visual acuity was poorer in treated eyes than in untreated eyes. Thus there was a net loss of QALYs during this time. Subtracting the QALYs lost at 3 months after treatment, the total number of QALYs gained during the first 4 years in the treated group was After year 4, Markov modeling was used to calculate the remaining number of QALYs gained from treatment. The mean remaining life expectancy of the surviving cohort at this time is approximately 7 years. The Markov modeling takes into account the complex issue of recurrent risk (and subsequent change in utility values) of subfoveal choroidal neovascularization converting from unilaterality to bilaterality. Over the 7-year cycle (remaining mean life expectancy of the cohort starting at year 5 after randomization to treatment or no treatment), the Markov model calculated that the total number of QALYs gained by treatment was Added to the QALYs gained during the first 4 years after randomization, the total number of QALYs gained per patient after treatment for subfoveal choroidal neovascularization was A sensivity analysis was performed on the number of QALYs gained using different discount rates over the 11-year period representing the mean life expectancy of the cohort of patients after randomization (Table 4). The number of QALYs gained from laser therapy varied from with no discounting to using a discount rate of 10%/year. Cost Effectiveness With the preferred discount rate of 3% per year selected by the authors, the incremental cost per QALY ($/QALY) for laser treatment of subfoveal choroidal neovascularization in 1999 U.S. nominal dollars was $5629. Sensitivity analyses of the $/QALY at various discount rates are shown in Table 5. The $/QALY ranged from $4074 with no discount rate to $11,633 at a yearly discount rate of 10%. Discussion The present study found that laser treatment for subfoveal choroidal neovascularization that meets the treatment criteria set forth in the Macular Photocoagulation Study 1,2,5 appears to be cost effective when compared with other treatment methods in medicine. Although the criteria for cost effectiveness are arbitrary, it has been suggested by Laupacis et al 13 that a treatment costing less than $20,000/ QALY appears to be cost effective, whereas a treatment costing more than $100,000/QALY is likely not to be cost effective. At a cost of $5629/QALY, the laser treatment reported herein appears to be a highly cost-effective interventional therapy using a patient preference-based methodology. Even using a discount rate as high as 10%/year, the treatment costs $11,633/QALY, and thus is still highly cost effective. Utility analysis allows a comparison of quality of life associated with health states, and subsequently cost effectiveness, across medical specialties. As noted by Redelmeier and Detsky, 29 utility analysis for the evaluation of quality of life associated with a health state encompasses all related parameters that comprise quality of life, including physical disability, pain, psychiatric overlay, relationship to employment, family support, and so forth. Thus interventional therapies for disease entities within ophthalmology and outside ophthalmology can be compared using the system used in this study. Such a comparison is very difficult using organ-specific analyses such as the VF A utility analysis, incremental cost effectiveness of various medical interventional therapies, is shown in Table 6. Of note is the fact that expensive therapies from the point of view of an initial outlay of dollars may actually be cost effective because of the improvement in quality of life or length of life conferred by the therapy. An example of such is coronary artery bypass surgery for an obstruction of the left anterior descending artery. Although the initial outlay of dollars for this therapy is approximately $25,000 using 1999 Health Care Financing Agency rates, 22 the treatment is highly cost effective at approximately $7000/QALY. In contrast, interventional therapies that require a relatively minor outlay of dollars initially may not be cost effective because they fail to improve quality of life or length of life. An example of this phenomenon is antibiotic prophylaxis before dental procedures in patients with prosthetic joints. The initial cost of the treatment is typically less than $100, but the cost effectiveness is almost $1.4 million/qaly. Numerous studies have evaluated the costs associated with medical interventional therapies. Most, however, fail to address the cost effectiveness of an interventional therapy from the patient preference point of view. In the past, utility values have been obtained from patients, but also from alternative respondents such as physicians, family, the general public, and others. 29 Indeed, in some instances, such as the inability of a patient to understand questions or communicate, alternate respondents are necessary. Nonetheless, the authors of the present study and others 12,13,15,16,31 33 believe that utility values obtained from patients with the health state in question are generally the most valuable and 1378

6 Brown et al Subfoveal Choroidal Neovascularization Table 6. Incremental Cost Effectiveness of Therapeutic Interventions Adjusted to Convert Nominal Costs to 1999 Real Costs in U.S. Dollars* Intervention $/QALY Laser therapy for threshold ROP $ 688 Cryotherapy for threshold ROP $ 1828 Coronary artery bypass for left main coronary artery disease $ 6985 Neonatal ICU treatment ( grams birth weight) $ 11,510 Cochlear implant in deaf individual $ 17,153 Chemoprophylaxis after occupational exposure to HIV $ 37,555 Treatment of hypertension (diastolic mmhg) in males, aged 40 $ 55,754 Neonatal ICU treatment ( grams birth weight) $ 81,338 Continuous ambulatory peritoneal dialysis $ 86,529 Liver transplant $ 332,413 Three-day chemoprophylaxis for prosthetic joint patients during dental treatment $1,370,250 HIV human immunodeficiency virus; ICU intensive care unit; QALY quality-adjusted life-years; ROP retinopathy of prematurity. *Adapted from Brown. 10 Adjusted for inflation using the Consumer Price Index from the U.S. Department of Labor Bureau of Labor Statistics. 28 pertinent concerning the effects of therapy on improving quality of life. In support of a patient preference-based approach over that of alternative respondents is the fact that patient perceptions of the quality of life associated with their health state often differ substantially from the perceptions of physicians who treat them. 10,32,33 This phenomenon has been noted in ophthalmology, neurology, psychiatry, and other areas of medicine. Furthermore, just as utility values obtained from patients can differ from those obtained from nonpatient groups for the same health state, there are a number of methods for obtaining utility values from patients. Included among these are the time tradeoff method, the standard gamble method, and the willingness-to-pay method. 7,8,10,20,29 Utility values for the same health state have been shown to differ substantially depending on the methodology used. 10,20,21 The above factors should be taken into account when comparing utility analyses or cost-effective analyses based on utility values. The data presented herein are based on quality-of-life preferences associated with a health state in this instance, visual loss obtained from patients. As mentioned, previous data have shown that decreasing levels of vision in the better seeing eye correlate with decreasing utility values. 10 In particular, there appears to be a marked drop in utility value, and associated quality of life, when the vision decreases from 20/200 to 20/500. This marked decrease in quality of life helps to explain why treatment of subfoveal choroidal neovascular membranes is especially cost effective. Although the authors cannot be certain, through interviews with patients it seems that a drop in vision from 20/200 to 20/500 makes many of the activities of daily living (reading with a visual aid, writing checks, cooking, watching television) substantially more difficult. The fact that utility values appear to be dependent primarily on the visual acuity level in the better-seeing eye does not indicate that treatment in eyes with visual loss should not be undertaken if one eye has good vision. On the contrary, most diseases that cause visual loss (age-related macular degeneration, glaucoma, cataract) are, or become, bilateral at some point. Decision analysis with Markov modeling incorporates this unilateral versus bilateral phenomenon. Despite the predominantly unilateral nature of the choroidal neovascular process at the outset, the treatment is still highly cost effective from the patient preference-based point of view, particularly because of the drastic drop in utility value that occurs when the second eye becomes involved. The analysis presented in this study takes into account the medical incremental costs associated with treatment of subfoveal choroidal neovascularization. Thus it does not incorporate the medical costs of screening for choroidal neovascularization or the associated societal costs such as loss of employment and benefits paid for incurred disability. Additionally, costs such as travel for treatment were not included. It is the incremental costs, however, that are quite comparable from one specific interventional procedure to another across different medical specialties. The comparability of the cost effectiveness of interventional therapies in general merits at least limited discussion. As seen in the sensitivity analysis performed in the present study, the $/QALY can vary dramatically, depending on the discount rate used to take into account the time value of money. When comparing the cost effectiveness of different therapies, therefore, the rates used to discount QALYs or costs should be comparable. Moreover, cost-effective analyses performed in different years should be appropriately adjusted for inflation, and those performed in different countries should be corrected for currency exchange differentials. 34 The cost comparisons in different countries should also reflect what payment scale is used, an example being the Relative Value Scale in use by the Health Care Financing Administration in the United States. 23 We believe that the most accurate comparisons of cost effectiveness occur when all of the above cost parameters, as well as the aforementioned utility parameters (utility study group and utility elicitation methods), are standardized. In this regard, it should be noted that the incremental cost effectiveness of the interventions listed in Table 6, although adjusted for inflation, may not be exactly comparable because some of the above parameters mentioned could not be ascertained from the studies from which the data were derived. As with any study, there are inherent biases in the present 1379

7 Ophthalmology Volume 107, Number 7, July 2000 one. The 10% rate of conversion of the second eye to subfoveal choroidal neovascularization after 5 years after treatment is an assumption. Additionally, the utility values used for the various visual acuity levels were derived from patients with multiple causes of visual loss. 10 Nevertheless, it has been demonstrated that the degree of visual loss in the better eye, rather than the cause of visual loss, most closely correlates with time tradeoff utility values. 10 Despite the fact that laser therapy for the treatment of subfoveal choroidal neovascularization initially causes a mean decrease in vision in treated patients, the long-term visual benefit in terms of associated quality of life appears to outweigh substantially this initial detrimental treatment effect. Combining discounted costs with the decision analysis outcome of subfoveal laser treatment demonstrates that this therapy appears to have a positive benefit for individual patients and society in general. References 1. Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions in age-related macular degeneration. Arch Ophthalmol 1991;109: Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions of age-related macular degeneration. Updated findings from two clinical trials. Arch Ophthalmol 1993;111: Bressler NM. Submacular surgery. Are randomized trials necessary? [editorial]. Arch Ophthalmol 1995;113: Photodynamic therapy of subfoveal choroidal neovascularization in age-related macular degeneration with verteporfin: one-year results of 2 randomized clinical trials TAP report. Treatment of Age-related Macular Degeneration with Photodynamic therapy (TAP) Study Group. Arch Ophthalmol 1999; 117: Macular Photocoagulation Study Group. Subfoveal neovascular lesions in age-related macular degeneration. Guidelines for evaluation and treatment in the Macular Photocoagulation Study. Arch Ophthalmol 1991;109: Torrance GW, Feeny D. Utilities and quality-adjusted life years. Int J Technol Assess Health Care 1989;5: Brown MM, Brown GC, Sharma S, Garrett S. Evidence-based medicine, utilities, and quality of life [review]. Curr Opin Ophthalmol 1999;10: Brown GC, Sharma S, Brown MM, Garrett S. Evidence-based medicine and cost-effectiveness. J Health Care Finance 1999; 26: Gage BF, Cardinalli AB, Albers GW, Owens DK. Costeffectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation. JAMA 1995; 274: Brown GC. Vision and quality-of-life. Trans Am Ophthalmol Soc 1999;97: Drummond MF, Richardson WS, O Brien BJ, et al. Evidence- Based Medicine Working Group. Users guides to the medical literature. XIII. How to use an article on economic analysis of clinical practice. A. Are the results of the study valid? JAMA 1997;277: O Brien BJ, Heyland D, Richardson WS, et al. Evidence- Based Medicine Working Group. Users guides to the medical literature. XIII. How to use an article on economic analysis of clinical practice. B. What are the results and will they help me in caring for my patients? JAMA 1997;277: Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. Can Med Assoc J 1992;146: Data 3.5 Healthcare User s Manual. Williamstown, MA: Treeage Software, Inc., Detsky AS, Naglie G, Krahn MD, et al. Primer on medical decision analysis: part 2 Building a tree. Med Decis Making 1997;17: Naimark D, Krahn MD, Naglie G, et al. Primer on medical decision analysis: part 5 Working with Markov processes. Med Decis Making 1997;17: Risk factors for choroidal neovascularization in the second eye of patients with juxtafoveal or subfoveal choroidal neovascularization secondary to age-related macular degeneration. Macular Photocoagulation Study Group. Arch Ophthalmol 1997;115: Visual outcome after laser photocoagulation for subfoveal choroidal neovascularization secondary to age-related macular degeneration. The influence of initial lesion size and initial visual acuity. Macular Photocoagulation Study Group. Arch Ophthalmol 1994;112: Vital Statistics of United States, 1994: Preprint of Volume II, Mortality, Part A, Section 6: Life Tables. Hyattsville, MD: U.S. Department of Health and Human Services, CDC, NCHS, DHHS Publication No. (PHS) , Brown MM, Brown GC, Sharma S, Shah G. Utility values and diabetic retinopathy. Am J Ophthalmol 1999;128: Brown GC, Sharma S, Brown MM, Kistler J. Utility values and age-related macular degeneration. Arch Ophthalmol 2000; 118: Kirschner CG, Davis SJ, Duffy T, et al. CPT 1999: Current Procedural Terminology. Chicago: American Medical Association, 1998;20 4, Relative Value Studies, Inc. The Complete RBRVS. Reston, VA: St. Anthony Publishers, 1998; American Academy of Ophthalmology. Age-related macular degeneration. Preferred practice pattern. San Francisco, CA: American Academy of Ophthalmology, 1998; Damodaran A. Corporate Finance. Theory and Practice. New York: John Wiley & Sons 1997; chap 3, Ross SA, Westerfield RW, Jordan BD. Fundamentals of Corporate Finance, 3rd ed., Chicago: Irwin, 1995; chap 5, Barron s: The Dow Jones Business and Financial Weekly. 1999, March 15; MW Consumer Price Index. Washington, D.C.: U.S. Department of Labor Bureau of Labor Statistics [cited 10/18/98]. Available from: ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt. 29. Redelmeier DA, Detsky AS. A clinician s guide to utility measurement. In Bergus GR, Cantor SB, eds. Medical Decision Making. Philadelphia: Saunders, 1995; Steinberg EP, Tielsch JM, Schein OD, et al. The VF-14. An index of functional impairment in patients with cataract. Arch Ophthalmol 1994;112: Kassirer JP. Incorporating patients preferences into medical decisions [editorial]. N Eng J Med 1994;330: Schrader GD. Subjective and objective assessments of medical comorbidity in chronic depression. Psychother Psychosom 1997;666: Rothwell PM, McDowell Z, Wong CK, Dorman PJ. Doctors and patients don t agree: cross sectional study of patients and doctors perceptions and assessments of disability in multiple sclerosis. Br Med J 1997;314: Drummond MF. Economic aspects of cataract. Ophthalmology 1988;95: