THE BURDEN OF AGE-RELATED MACULAR DEGENERATION: A VALUE-BASED MEDICINE ANALYSIS

Abstract

Purpose

To assess the quality-of-life loss and the macroeconomic financial consequences associated with age-related macular degeneration (ARMD).

Methods

Time tradeoff utility analysis was performed to assess the quality-of-life diminution caused by ARMD (both dry and neovascular) in cohorts consisting of (1) patients with ARMD, (2) ophthalmologists asked to assume they had various degrees of severity of ARMD, (3) healthcare providers asked to assume they had various degrees of severity of ARMD, and (4) participants from the general community asked to assume they had various degrees of severity of ARMD. ARMD was classified according to vision in the better-seeing eye as (1) mild: 20/20 to 20/40, (2) moderate: 20/50 to 20/100, (3) severe: ≤ 20/200, or (4) very severe: ≤ 20/800.

Results

Mild ARMD caused a 17% decrement in the quality of life of the average patient, similar to that encountered with moderate cardiac angina or symptomatic human immunodeficiency virus syndrome. Moderate ARMD caused a 32% decrease in the average patient’s quality of life, similar to that associated with severe cardiac angina or a fractured hip. Severe ARMD caused a 53% decrease in quality, more than that of dialysis, and very severe ARMD caused a 60% decrease in the average ARMD patient’s quality of life, similar to that encountered with end-stage prostate cancer or a catastrophic stroke that leaves a person bedridden, incontinent, and requiring constant nursing care. Patients with varying degrees of severity of ARMD were found to have quality-of-life impairment ranging from 96% to 750% greater than that estimated by treating ophthalmologists for the same condition.

An economic analysis based upon losses to the gross domestic product suggests that ARMD has approximately a $30 billion annual negative impact. The return on investment is therefore potentially high for both treatment with current ARMD therapies and the research costs invested in the development of new ARMD treatment modalities.

Conclusions

ARMD is a major public health problem that has a devastating effect upon patients and marked adverse financial consequences for the economy.

INTRODUCTION

Age-related macular degeneration (ARMD) is the leading cause of legal blindness in the older population in the United States today.¹ In addition to the considerable deleterious effect it has upon the quality of life,^2–4 ARMD has adverse financial consequences upon the overall economy of the United States. To date, little study has been devoted to comparing the adverse effects of ARMD with those of other diseases.^2–4

There are two forms of ARMD: (1) dry, or atrophic, and (2) wet, or neovascular. According to the International Age-Related Maculopathy Epidemiological Study Group,⁵ the fundus appearance of dry macular degeneration is characterized by the presence of geographic atrophy, drusen, and areas of retinal pigment epithelial hyperplasia. Typically, dry ARMD occurs in an eye prior to development of the wet form.

The onset of wet macular degeneration (also called exudative macular degeneration or neovascular macular degeneration) is heralded by the appearance of subretinal blood, hard exudates, and/or subretinal fluid in the macula occurring secondary to choroidal neovascularization. Development of the wet variant of ARMD is usually accompanied by visual loss that progresses over weeks to months and longer.² The majority of eyes that progress to legal blindness (20/200 or worse visual acuity) have the wet variant. Although ARMD is the leading cause of legal blindness in the older population in the United States,⁶ the etiology is unknown.

PREVALENCE

The prevalence of ARMD increases with advancing age.^1,6 Overall, 1.47% of people over 40 years of age, or approximately 1.75 million people in the United States, have advanced macular degeneration (the neovascular variant and/or geographic atrophy variant). Among this group, approximately 1.25 million have the neovascular (wet) variant of macular degeneration (Table 1).^1,6 The number of cases of wet macular degeneration in the United States is expected to rise to 1.875 million by the year 2020.⁶

TABLE 1.

PREVALENCE OF AGE-RELATED MACULAR DEGENERATION (ARMD) IN THE UNITED STATES

AGE (YEARS)	NEOVASCULAR ARMD	GEOGRAPHIC ATROPHY ARMD*	DRUSEN (125 μ m OR LARGER)
40–49	20,000	NA	51,000
50–54	40,000	27,000	519,000
55–59	36,000	25,000	534,000
60–64	41,000	31,000	585,000
65–69	60,000	46,000	709,000
70–74	101,000	80,000	906,000
75–79	154,000	132,000	1,049,000
80+	734,000	632,000	2,154,000

NA = not applicable.

^*Geographic atrophy is an area of retinal pigment epithelial depigmentation at least 175 μm in diameter. Note that some cases can have neovascular ARMD in one eye and geographic atrophy in the fellow eye.

Adapted from Bird, et al.⁵

It should be noted that many ophthalmologists in clinical practice consider the presence of macular drusen to be an early form of dry ARMD. Using that definition, 7 million people in the United States with macular drusen measuring 125 μm or more in diameter could be considered to have dry ARMD.⁶

INCIDENCE

Data from the Beaver Dam Eye Study,¹ a longitudinal cross-sectional study of people living in Beaver Dam, Wisconsin, reveal that among a group aged 43 to 86 years, the incidence of early (mild) ARMD is 12.1% over a 10-year period, or 1.21% per year. Nonetheless, among people aged 75 years or older, the incidence rises to 36.7% over a 10-year period, or 3.67% per year. The incidence of late (advanced) ARMD is 2.1% over a 10-year period in the 43- to 86-year-old group and 9.5% over a 10-year period in the group aged 75 years or older. These numbers are summarized in Table 2.

TABLE 2.

INCIDENCE OF AGE-RELATED MACULAR DEGENERATION (ARMD) OVER A 10-YEAR PERIOD FROM THE BEAVER DAM STUDY

AGE GROUP (YEARS)	10-YEAR INCIDENCE OF EARLY ARMD (%)	10-YEAR INCIDENCE OF LATE ARMD (%)
43 to 54	4.1	0.1
55 to 64	10.7	1.0
65 to 74	23.6	4.4
75+	36.7	9.5
Overall	12.1	2.1

Adapted from Klein et al.¹

Extrapolation of data from the Beaver Dam Eye Study¹ and the Eye Diseases Prevalence Research Group⁶ suggests that 1.6 million new cases of dry macular degeneration and 153,000 new cases of wet macular degeneration associated with ARMD develop in the United States each year.

The relevance, or burden, of a disease from the public health perspective depends upon the incidence of the disease, the impact of the disease upon the lives of people it affects, and the cost impact. In essence, both the human burden and the economic burden give the best profile of the total burden of a disease and its public health relevance.

From the incidence viewpoint, it is already clear that ARMD is a major public health problem. The effect of ARMD upon an individual, or the human burden, can best be demonstrated using evidence-based medicine data integrated with value-based medicine methodology, whereas the economic disease impact can be assessed by multiple methods, one of which is measuring the effect of the disease upon the gross domestic product (GDP).

The objectives of this manuscript are to assess the relevance of ARMD as a public health dilemma by (1) comparing the patient-perceived quality of life associated with ARMD with quality-of-life estimates obtained from surrogate respondents asked to presume they had ARMD; (2) comparing the patient-perceived quality of life associated with ARMD with the quality of life associated with other diseases; and (3) quantifying the adverse macroeconomic effect of ARMD upon the economy of the United States.

METHODS

REVIEW OF VALUE-BASED MEDICINE CONCEPTS

Evidence-based medicine is the practice of medicine based upon the highest level of scientific evidence available. There are five levels of interventional studies, as shown in Table 3.⁷ Level 1, the randomized clinical trial, represents the highest level of evidence. Level 1 data from clinical trials are the most reproducible and allow clinicians and patients to have the greatest confidence in the repeated outcomes of their treatments. It is preferable to use Level 1 data in value-based analyses that evaluate interventions. For the evaluation of quality of life, such as the case herein, cross-sectional data are reasonable.

TABLE 3.

FIVE LEVELS OF INTERVENTIONAL EVIDENCE

LEVEL OF EVIDENCE	INTERVENTIONAL STUDY
Level 1	Randomized clinical trial with low type 1 error (≤ 0.05) and low type 2 error (≤ 20%) or meta-analysis
Level 2	Randomized clinical trial with high type 1 error (>0.05) and/or high type 2 error (>0.20)
Level 3	Uncontrolled, nonrandomized clinical trial (treatment group compared to no treatment group without randomization)
Level 4	Intervention on a series of patients with no comparison group
Level 5	Interventional case report

Adapted from Sharma.⁷

Although the data gleaned from clinical trials are invaluable, the primary evidence-based outcomes of clinical trials (eg, number of deaths, number of strokes, 3 lines of visual loss) often overlook important quality-of-life variables associated with a disease and interventions employed to treat that disease. In particular, the data do not present a mechanism by which the adverse effects associated with an intervention can be factored into the clinical value equation. Just as important, evidence-based data alone generally do not allow a comparison of the value conferred by dissimilar interventions across different specialties.

VALUE-BASED MEDICINE

Value-based medicine is the practice of medicine based upon the value conferred by interventions. This value is derived from evidence-based data that are converted to value-based data. As well as allowing a comparison of data from disparate interventions, value-based medicine also permits data from evidence-based clinical trials to be used in healthcare economic analyses. Of note is the fact that a value-based analysis is only as good as the underlying evidence-based data utilized in that analysis.

WHAT IS VALUE?

The value of a healthcare intervention is measured by quantifying the improvement it confers in (1) length of life and/or (2) the patient-perceived quality of life. Every intervention must deliver one or both, or it should be discarded from the therapeutic armamentarium. Most ophthalmologic interventions do not confer an improvement in length of life; the conferred value is therefore generally measured by quantifying the improvement in patient-perceived quality of life.

The improvement in length of life for most interventions can be gleaned from data in the evidence-based literature, but measuring the improvement in quality of life gained from an intervention is more difficult. It can, however, be objectively determined using health-related quality-of-life instruments.

HEALTH-RELATED QUALITY-OF-LIFE INSTRUMENTS

There are numerous health-related quality-of-life instruments in use, including the Medical Outcomes Study Short-Form-36 (SF-36),⁸ the Quality of Well-Being Scale,⁹ and the Activities of Daily Living (ADL) Scale.¹⁰ For ophthalmologic diseases, the most popular instruments are the 14-item Visual Function (VF-14)¹¹ and the 25-item National Eye Institute Visual Function Questionnaire (NEI VFQ-25).¹²

FUNCTION-BASED INSTRUMENTS

The majority of the health-related quality-of-life instruments in use are function-based, meaning that they evaluate entities such as physical function, social function, visual function, and psychological function.¹³ But by essentially evaluating function, they can miss important additional aspects that compose quality of life, such as caregiver status, the welfare of dependants, economic distress, and fear of the future.¹³ In addition, most function-based instruments are not applicable to diseases encountered across all specialties in healthcare. This is especially so for ophthalmologic instruments such as the VF-14 and the NEI VFQ-25, which are used only to evaluate ocular diseases. Function-based instrument results, in contrast to those of preference-based instruments, are generally not used in healthcare economic analyses.¹³

PREFERENCE-BASED INSTRUMENTS

Preference-based instruments require a patient to indicate a preference for the desirability of a health state (state of health). The term health state is often synonymous with disease, but it also includes the states of perfect health, death, and combinations of diseases.

There are two basic types of preference-based instruments: (1) rating scales and (2) utility analysis. Rating scales typically ask a person to rate his or her health on a scale ranging from 0 (death) to normal health (100), although other anchors, such as the poorest possible health and the best possible health, have also been used. Rating scales have been criticized for poorer reliability (reproducibility) than utility analysis, and also create patient comprehension difficulties when the lower anchor of death is used and the disease under evaluation is not lethal (as is the case with many ophthalmic, otolaryngologic, dermatologic, and plastic surgical conditions).^13,14

UTILITY ANALYSIS

Utility analysis is the health-related quality-of-life instrument preferred by most healthcare economic researchers to objectively measure the quality of life improvement conferred by medical interventions.^13,15,16 There are four basic variants: (1) time tradeoff utility analysis, (2) standard gamble utility analysis, (3) willingness-to-pay utility analysis, and (4) multi-attribute instruments such as the EuroQol and the Health Utilities Index.¹³ The time tradeoff method is among the most reproducible^13,14 of the group and has excellent construct validity (meaning that the instrument measures what it is intended to measure, in this instance the health-related quality of life associated with a disease).¹³ Those with an in-depth interest in utility analysis are referred to other writings for additional information.^13,15,16

With the time tradeoff utility analysis variant used in the current study, a respondent is first asked how long he or she theoretically expects to live. The person is next asked what is the maximum amount of that time—if any—he or she would be willing to trade for a return to normal health during the years that remain. The utility value associated with the disease is then calculated by subtracting the proportion of time traded from 1.0. For example, if a person with ARMD expects to live for 10 additional years and is willing to trade 3 of those years to be rid of the ARMD, the resultant utility value associated with the condition is 0.70 (1.00 – 0.30). If the same patient is willing to trade 6 of 10 remaining years to be rid of ARMD, the resultant utility value is 0.40 (1.00 – 0.60).

Utility analysis values obtained from a large number of patients typically have a narrow 95% confidence interval, meaning the mean utility value associated with a disease has a 95% chance of falling within that narrow range if the study was to be repeated.^17–21 Narrow confidence intervals give clinicians assurance that the results of a study are reproducible.

OPHTHALMIC UTILITY VALUES

Ophthalmic utility values have been previously shown to most highly correlate with visual acuity in the better-seeing eye, rather than the underlying cause of visual loss.^4,13 A list of ophthalmic utility values previously obtained from patients with varied visual conditions is shown in Table 4. Noteworthy is the fact that normal 20/20 visual acuity in each eye permanently is associated with a utility value of 1.00. Patients with ocular disease (eg, ARMD, diabetic retinopathy, glaucoma) and no visual loss (20/20 visual acuity bilaterally) have a utility value of 0.97 because of the fear that vision will be lost in the future. When the visual acuity is 20/20 in one eye and 20/40 or worse in the second eye, the associated utility value is 0.92. As the vision in the better-seeing eye decreases, so does the utility value. No perception of light bilaterally is associated with a utility value of 0.26, and death, the lower anchor of the scale, is associated with a utility value of 0.00

TABLE 4.

TIME TRADEOFF UTILITY VALUES ASSOCIATED WITH VISUAL ACUITY LEVELS IN THE BETTER-SEEING EYE

VISUAL ACUITY	UTILITY VALUE
20/20 in each eye permanently	1.00
20/20 (20/20 to 20/25 in the other eye)	0.97
20/20 (≤ 20/40 in the other eye)	0.92
20/25	0.87
20/30	0.84
20/40	0.80
20/50	0.77
20/70	0.74
20/100	0.67
20/200	0.66
20/300	0.63
20/400	0.54
Counting fingers (20/800)	0.52
Hand motions	0.35
Light perception	0.35
No light perception	0.26
Death	0.00

Adapted from Brown MM, et al.⁴

IMPROVEMENT IN QUALITY OF LIFE GAINED FROM AN INTERVENTION

Just as utility analysis measures the quality of life associated with a health state, it can also measure the improvement in the quality of life conferred by an intervention. Using the above example, assuming that pretreatment ARMD is associated with a visual acuity of 20/400 (utility value = 0.54) and posttreatment ARMD is associated with a visual acuity of 20/40 (utility value = 0.80), the improvement in utility value gained from the intervention is +0.26. The improvement in quality of life (value) in this instance is 48% (0.26/0.54).

TOTAL VALUE GAINED FROM AN INTERVENTION

Once the utility value improvement gained from an intervention is ascertained, the total value conferred to a patient can be calculated. The total value gained is measured by using the outcome of the quality-adjusted life-year (QALY).

People accrue QALYs with time. If a patient’s utility value remains unchanged for a period of time, the formula for the number of QALYs accrued during that time is:

$$\text{Utility value}\times \text{time}(\text{years})$$

Thus, a person in perfect health with a utility value of 1.0 accrues 1.0 QALY during 1 year. If a person with ARMD lives at a utility value of 0.8 for 1 year, that person accrues 0.8 QALY over the year. A person with a utility value of 0.8 who lives for 20 years accrues 16.0 (0.8 × 20) QALYs during that time. The 16.0 QALYs represent the total value of the person’s life accrued during that time.

$$\begin{array}{c}\text{The formula to measure the total value gained from an intervention in QALYs is the sum of}:\\ (\text{utility value gain from intervention})\times (\text{duration of treatment benefit in years})\\ +\\ (\text{posttreatment utility value})\times (\text{additional years of life gained from the intervention})\end{array}$$

For the great majority of ophthalmologic interventions, no additional years of life are gained. The value conferred by ophthalmologic interventions is therefore generally calculated by multiplying the utility value gain conferred by an intervention by the duration of treatment benefit. Assuming that treatment for macular degeneration improves a person’s utility value from 0.54 to 0.80 for 20 years, the total QALY gain is 5.2 (0.26 × 20 years). For a nonocular intervention, such as a carotid endarterectomy that improves the utility from 0.70 to 0.90 and adds 2 more years to an expected 5-year life expectancy, the gain in value is 2.8 QALYS [(0.2 × 5 years) + (0.9 × 2 years)].

Conversely, the value loss in a patient’s life incurred from a disease can also be measured. For example, if a person develops permanent, moderate macular degeneration and drops to an ocular utility value of 0.60 from 1.00, this person experiences a 40% diminution in quality of life. This can alternatively be looked at as a 40% decrement in the person’s remaining value of life.

SOURCE OF UTILITY VALUES

A primary tenet of value-based medicine is the principle that patient-based utility values should be utilized in quality-of-life analyses and healthcare economic analyses.¹³ It has been demonstrated repeatedly that the utility values of surrogate respondents often differ considerably from those of patients for the same disease.^3,4,17,20,21 Measures of quality of life other than utility analysis have also shown a considerable difference between the perceptions of patients who have lived with a disease and treating physicians asked to estimate the quality of life associated with that disease.^22–25 Patient-based utility values are generally the criterion, or gold standard, for the quality-of-life measures used in value-based medicine, because patients who live with a disease on a firsthand basis are those best qualified to assess the quality of life associated with that disease.^4,13,21,22

STUDY DESIGN

The study herein was approved by the Wills Eye Hospital Institutional Review Board. The quality of life of patients with ARMD and that of patients with other diseases are addressed, as is the loss to the GDP from ARMD.

For the purpose of evaluating the quality of life associated with ARMD in this study, primary data from previous studies^2,3,22 were re-analyzed. Time tradeoff utility analysis data were gathered from 82 consecutive subjects with ARMD,² as well as 142 community members,³ 62 nonophthalmologist clinicians,³ and 46 ophthalmologists who treat patients with ARMD.²² Included in the nonophthalmologist clinician group were residents and medical students, and included within the ophthalmologist group were attending physicians and residents. Subjects in the community, the nonophthalmologist clinician group, and the ophthalmologist group were all asked to assume that they had mild, moderate, or severe ARMD. The criteria for the visual strata are as listed below for the medical personnel (ophthalmologists and nonophthalmic physicians), whereas the community participants were given scenarios of mild, moderate, and severe visual loss, rather than the actual Snellen visual acuities.

Subjects were classified as having mild ARMD (group 1) if they had a visual acuity of 20/20 to 20/40 in the better-seeing eye, moderate ARMD (group 2) if they had a visual acuity of 20/50 to 20/100 in the better-seeing eye, and severe ARMD (group 3) if the acuity was 20/200 or worse in the better-seeing eye. Data on patients with very severe ARMD (group 4), with a visual acuity of 20/800 or less in the better-seeing eye, were also available, but only for the ARMD patient group and the ophthalmologist group.

The means of the utility value groups were compared with one-way analysis of variance, and post hoc testing was performed using the least significant difference test for comparisons. Significance was presumed to occur at the P = .05 level.

In addition to the patient quality of life, an analysis was undertaken on the macroeconomic burden of ARMD. This analysis did not include aspects such as disability payments or caregiver payments, but rather was based on the annual loss in the US GDP occurring as a result of ARMD. The GDP reflects the total income (wages, rents, income, and profits) produced in the United States on a yearly basis and thus reflects the patient loss in salary incurred by ARMD.

RESULTS

THE QUALITY-OF-LIFE BURDEN OF ARMD

The time tradeoff utility analysis results are shown in Table 5. They demonstrate significant differences between the mean ARMD utility values of the patient cohort and those of the community, nonophthalmologist clinician, and patient cohorts, respectively (P < .001).

TABLE 5.

UTILITY VALUES OF PATIENTS WITH AGE-RELATED MACULAR DEGENERATION (ARMD) AND RESPONDENT GROUPS

ARMD GROUPINGS*	UTILITY VALUE (SD; 95% CI)				PVALUE
	PATIENTS WITH ARMD†2,3(N = 82)	COMMUNITY3(N = 142)	CLINICIANS3(N = 62)	OPHTHALMOLOGISTS22(N = 46)
Mild (20/20 to 20/40)	0.83 (.19; .77–.89) (n = 42)	0.96 (.06; .95–.97) (n = 142)	0.93 (.10; .90–.96) (n = 62)	0.98 (.03; .97–.99) (n = 46)	<.001*
Moderate (20/50 to 20/100)	0.68 (.21; .59–.77) (n = 22)	0.92 (.10; .90–.94) (n = 142)	0.88 (.12; .85–.91) (n = 62)	0.89 (.10; .86–.92) (n = 46)	<.001*
Severe (20/200 or worse)	0.47 (.18; .39–.55) (n = 18)	0.86 (.15; .84–.88) (n = 142)	0.82 (.14; .78–.86) (n = 62)	0.73 (.19; .68–.78) (n = 46)	<.001*
Very severe (≤ 20/800)	0.40 (.13; .31–.49) (n = 8)	NA	NA	0.67 (.18; .62–.72) (n = 46)	<.001*

ARMD = age-related macular degeneration; NA = not available.

^*Snellen visual acuity ranges for the better-seeing eye are listed below the Mild, Moderate, Severe, and Very severe groupings.

^†The patient values are significantly different from those of the community, clinician, and ophthalmologist groups in all categorical ARMD groupings, by one-way analysis of variance (P <.0001.)

Especially dramatic are the differences between the group of patients with ARMD and the group of ophthalmologists who treat ARMD. For mild ARMD (visual acuity 20/20 to 20/40 in the better-seeing eye), the mean patient utility value was 0.83, which was 750% worse than the mean estimate from ophthalmologists (utility value of 0.98) who treat patients with ARMD. Overall, the condition produced a 17% decrement in the average ARMD patient’s remaining quality of life, whereas the ophthalmologists estimated a mean 2% quality-of-life decrement for the same condition.

For moderate ARMD (visual acuity 20/50 to 20/100 in the better-seeing eye), the mean patient utility value was 0.68, which was 191% worse than the treating ophthalmologists’ mean estimate (utility value = 0.89). Patients in this group experienced a mean 32% loss in quality of life from the condition.

For severe ARMD (visual acuity ≤ 20/200 in the better-seeing eye), the mean patient utility value was 0.47, which was 96% worse than the treating ophthalmologists’ mean estimate (utility value = 0.73). Patients in this group experienced a mean 53% diminution in quality of life from the condition.

For very severe ARMD (visual acuity ≤ 20/800 in the better-seeing eye), the mean patient utility value was 0.40, which was 97% worse than the treating ophthalmologists’ mean estimate (utility value = 0.67). Patients in this group experienced a 60% loss in quality of life from the condition.

Although the ophthalmologist mean utility value was closer to the mean patient value for the very severe cases of ARMD than were the community and the nonophthalmologist mean utility values, the disparity compared with the mean patient utility value was still remarkable. The average ophthalmologist asked to assess the quality of life associated with very severe ARMD was willing to trade approximately 3.3 of every 10 remaining years for a return to normal vision, whereas the average patient with very severe ARMD was willing to trade 6.1 of every 10 remaining years for the same result.

QUALITY OF LIFE ASSOCIATED WITH ARMD VERSUS QUALITY OF LIFE ASSOCIATED WITH OTHER DISEASES

Utility analysis has an advantage over other health-related quality-of-life instruments in that it can effectively compare the quality of life associated with any health state and the quality-of-life improvement conferred by any healthcare intervention. For the comparison to be valid, however, the utility values must be obtained from the same respondent group (patients) using the same methodology of utility analysis (time tradeoff).^4,13

Patient-based time tradeoff utility values associated with systemic diseases and different levels of severity of ARMD are shown in Table 6. For the purpose of clarity, the utility values listed are divided into three groups: (1) 0.75 to 1.00: mild to moderate changes, (2) 0.50 to 0.74: moderate to severe changes, and (3) <0.50: severe to very severe changes.

TABLE 6.

PATIENT-BASED, TIME TRADEOFF UTILITY VALUES ASSOCIATED WITH AGE-RELATED MACULAR DEGENERATION (ARMD) AND SYSTEMIC HEALTH STATES

CONDITION	TTO UTILITY VALUE
GROUP 1
HIV, asymptomatic	0.9423
Status post–myocardial infarction, no symptoms	0.9324
Cancer, all	0.9225
Osteoporosis	0.9126
Stroke, mild (able to perform usual activities)	0.9027
Impotence	0.8528
Gout	0.8629
ARMD, mild*	0.83*
Vertebral fracture	0.8226
HIV, symptomatic	0.8223
Angina, moderate	0.8030
Status post–myocardial infarction, some residual angina and congestive heart failure	0.7824
Prostate cancer (no pain; normal bladder, bowel, and sexual function)	0.7831
GROUP 2
Claudication, severe	0.7432
Prostate cancer (pain controlled; ± bladder bowel, and sexual function, ± energy ± depression)	0.7231
AIDS	0.7023
Stroke, moderate (requiring some help, but able to walk without assistance)	0.6927
ARMD, moderate*	0.68*
Fractured hip	0.6326
Tuberculosis, hospitalized	0.6033
Angina, severe	0.5830
Ulcerative colitis, requiring surgery	0.5834
GROUP 3
Dialysis, home	0.5633
ARMD, severe*	0.47*
GROUP 4
ARMD, very severe*	0.39*
Prostate cancer, advanced (uncontrolled pain; bladder, bowel and sexual function abnormal; depression, severe fatigue)	0.3531
Stroke, severe (bedridden, incontinent, and requiring constant care, at 6 months)	0.3427
Total blindness (NLP OU)	0.2635
Stroke, severe, with aphasia	0.2636
Stroke, severe, total paralysis (at 10 years)	0.2037

AIDS = acquired immunodeficiency syndrome; ARMD = age-related macular degeneration; HIV = human immunodeficiency virus; NLP = no light perception; TTO = time tradeoff.

^*Bold indicates current study.

In group 1,^23–31 mild ARMD (mean utility value = 0.83) is associated with a greater quality-of-life decrement than that encountered with cancer (average among several types),²⁵ a mild stroke,²⁷ impotence,²⁸ or gout.²⁹ The quality of life associated with it is similar to that associated with having a vertebral fracture²⁶ or symptomatic HIV syndrome²³ and very close to that of moderate angina.³⁰

In group 2,^32–34 moderate ARMD (mean utility value = 0.68) is associated with a quality of life similar to that following a moderate stroke,²⁷ after which a patient requires considerable help with daily functions. The quality of life is also similar to that associated with AIDS.²³

In group 3, severe ARMD (mean utility value = 0.47) is associated with a quality of life similar to that of a patient with total renal failure on home dialysis.³³

In group 4,^35–37 patients with the very severe macular degeneration (utility value = 0.39), the associated quality of life is comparable to that encountered with a severe stroke that causes a person to be bedridden, incontinent, and in need of constant nursing care.²⁷ Another comparable condition is advanced prostate cancer with uncontrollable pain.³¹ Thus, the quality of life associated with very severe ARMD is on a par with some of the most harsh health states imaginable.

THE ECONOMIC BURDEN OF ARMD

The GDP can be calculated by adding the income of wages, rents, interest, and profits received by all factors in the production of all goods and services produced in the United States during a calendar year.³⁸ A loss of wages due to the sequelae of ARMD will consequently result in a diminution of the GDP. Thus, a major component of the macroeconomic burden of ARMD is the amount by which it decreases the GDP.

THE ECONOMIC BURDEN OF NEOVASCULAR ARMD

People with visual loss from ARMD experience considerable difficulty in obtaining employment. Data from the Bureau of Labor and Statistics provide a wealth of information in this arena.³⁹ The employment rate for those with ARMD and severe visual limitation, such as encountered with advanced ARMD (neovascular and/or geographic atrophy⁵), is 30.6%, whereas the employment rate for people with mild visual limitation is 44.1%. The comparable employment rate for unaffected people aged 16 to 63 years is 78.2%.³⁹ People with neovascular ARMD are considered to have severe visual loss for the purpose of the economic calculations presented herein.

In addition to having difficulty finding employment, people with severe visual limitation (advanced ARMD) have decreased earnings compared with those who have no disabilities. The 1997 mean wage for a person with no disabilities was $31,182, whereas that for a person with mild visual loss was $21,804, or 30% less ($9,378) than that for a person with no disabilities. In the same year, the average wage for a person with severe visual loss, such as associated with advanced ARMD, was $19,326, or 38% less ($11,856) than that for a person with no disabilities.³⁹

Incorporating the above data with the assumptions that 10% of the 36 million Medicare population (65 years and older) have full-time employment^40,41 and 2.9% of this group have neovascular ARMD (Table 1), the annual loss from the GDP caused by neovascular ARMD and the consequent reduction in salary is $1.238 billion ($11,856 × 36 million × 10% × 2.9%). Conservatively, it is likely that another 5% of remaining seniors with neovascular ARMD would be in the labor market if they did not have severe visual loss.^40,41 This latter subgroup accounts for an additional $1.628 million ($31,182 × 36 million × 5% × 2.9%) loss from the GDP. The sequelae of neovascular ARMD therefore account for a combined $2.866 billion GDP loss in the Medicare population.

Among the 137,000 people under the age of 65 years with neovascular ARMD (Table 1), the loss in salary is calculated by considering the percentage of the workforce unable to obtain a job due to severe visual loss (78.2% − 30.6% = 47.6%) and the salary decrease ($11,856) attributed to 30.6% of employed people with severe visual loss. The loss to the GDP in this group is therefore (47.6% × 137,000 × $31,182) + (30.6% × 137,000 × $11,856) = $2.033 billion + $0.497 billion = $2.530 billion.

The total yearly loss to the GDP due to lost wages from neovascular ARMD among all affected people is therefore $2.866 billion + $2.530 billion = $5.396 billion. Other costs, as yet poorly quantified, such as transportation costs, caregiver costs, and injury costs, make the overall economic burden from ARMD higher yet. A summary of the adverse economic sequelae occurring as a result of ARMD is shown in Table 7.

TABLE 7.

YEARLY ECONOMIC LOSS TO THE GROSS DOMESTIC PRODUCT (GDP) IN THE UNITED STATES FROM THE SEQUELAE OF AGE-RELATED MACULAR DEGENERATION (ARMD )

CATEGORY	PATIENT ≥ 65 YEARS OLD	PATIENTS <65 YEARS OLD
A. WET (NEOVASCULAR) ARMD
Per capita salary loss for those employed	$11,856	$11,856
People employed	104,400	41,922
Total salary reduction loss	$1.237 billion	$497 million
Jobs lost from neovascular ARMD	52,200	65,212
Average job salary	$31,182	$31,182
Total salary loss	$1.628 billion	$2.033 billion
Subtotal all losses	$2.866 billion	$2.530 billion
Total GDP loss from wet ARMD =$5.396 billion
B. DRY ARMD
Per capita salary loss for those employed	$9,378	$9,378
Employed with mild visual loss	240,900	548,824
Total salary reduction loss	$2.259 billion	$5.147 billion
Jobs lost from dry ARMD and mild visual loss	120,450	424,375
Average job salary	$31,182	$31,182
Total salary loss	$3.756 billion	$13.233 billion
Subtotal all losses	$6.015 billion	$18.380 billion
Total GDP loss from dry ARMD =$24.395 billion
TOTAL GDP LOSS FROM WET + DRY ARMD =$29.791 BILLION

Data from references 1, 6, and 39 through ⁴².

THE ECONOMIC BURDEN OF DRY ARMD

If the assumption is made that half of the 7.3 million people with dry ARMD and drusen greater than 125 μm across (Table 1) also have mild visual limitation,^39,42 the additional loss to the GDP can be calculated in a manner similar to that for neovascular macular degeneration

Among the 2,489,000 people under the age of 65 years with dry ARMD (Table 1), the loss in salary is calculated by considering the percentage of the workforce unable to obtain a job due to mild visual loss (78.2% − 44.1% = 34.1%) and the salary decrease ($9,378) attributed to 44.1% of employed people with severe visual loss. The loss to the GDP in this group is therefore (0.5 × 34.1% × 2,489,000 × $31,182) + (0.5 × 44.1% × 2,489,000 × $9,378) = $13.233 billion + $5.147 billion = $18.380 billion.

Among the 4,818,000 people with dry ARMD aged 65 years or older (Table 1), the loss in salary is calculated by considering the percentage of the workforce unable to obtain a job due to mild visual loss (78.2 – 44.1 = 34.1%) and the salary decrease ($9,378) attributed to 44.1% of employed people with severe visual loss. The loss to the GDP in this group is therefore (0.5 × 0.1 × 34.1% × 4,818,000 × $31,182) + (0.5 × 0.1 × 44.1% × 2,489,000 × $9,378) = $3.756 billion + $2.259 billion = $6.015 billion.

The total GDP loss from dry ARMD, assuming that 50% with dry ARMD have mild visual loss, is $18.380 billion + $6.015 billion = $24.395 billion.

THE COMBINED ECONOMIC BURDEN OF NEOVASCULAR AND DRY ARMD

When the $24.395 billion GDP loss from dry ARMD is combined with the loss of $5.396 billion from neovascular ARMD, the total loss to the GDP from both forms of ARMD is $29.791 billion. The annual GDP of the United States for 2003 was $10.988 trillion.⁴³ The potential loss of close to $30 billion, while only 0.27% of the GDP, is nonetheless a formidable sum. Not only does ARMD cause substantial human loss of quality of life, but its economic consequences in a population that grows older every year are severe.

DISCUSSION

The data herein suggest that ARMD has a considerable adverse effect upon quality of life. They also suggest that this adverse effect is considerably underestimated by physicians, by the general community, and by ophthalmologists who actually treat patients with ARMD. Although ophthalmologists dramatically underestimated the quality-of-life loss associated with macular degeneration, treating physicians in other specialties have also underestimated the degree of patient quality-of-life impairment caused by diseases they treat.^44–46 The phenomenon appears to be widespread and reinforces the concept that patient-based values should be used to calculate the value of interventions and to calculate the cost-utility associated with interventions.¹³

Of note, 30% of persons with ARMD also have clinical depression.⁴⁷ This can contribute to the low utility values encountered with ARMD, especially among persons with the more severe types of ARMD. Although it is not the intention of this paper to discuss the treatment of ARMD, clinicians should not lose sight of the fact that, in addition to treatment to preserve vision, treatment ofdepression in those with ARMD can result in a considerable improvement in quality of life.⁴⁷

As is the case with any study, the current study has shortcomings. The fact that community participants may not understand the disability associated with decreased vision should be considered. Nonetheless, the underestimations suggest that society overall underappreciates the quality-of-life loss sustained by ARMD patients.

A detailed description of the therapeutic modalities currently used to treat ARMD^48–52 is beyond the scope of this article. (For details on current treatments, please see references 48 through ⁵².) However, laser therapy⁵² for extrafoveal choroidal neovascularization with neovascular ARMD and photodynamic therapy⁵¹ for subfoveal choroidal neovascularization with neovascular ARMD are both cost-effective modalities that provide a great deal of value to patients. Each of these modalities is considerably more cost-effective than treatment for systemic arterial hypertension, one of the most common healthcare interventions in the United States.^51,52

The needs to apply current ARMD therapies properly^42,48–52 and pursue the discovery of more effective ARMD therapies are great. The education of patients and the public regarding the severely debilitating nature of ARMD will hopefully stimulate awareness among affected patients that treatments are available and more effective when the disease is caught at earlier stages.^48–52 Whereas value-based medicine data can be difficult to navigate, the percent improvement in quality of life conferred by an intervention is a concept that most people understand. This percent improvement in quality of life can be even more germane when interventions can also be compared across disparate specialties. Knowing the percentage improvement conferred in quality of life will likely help patients to better appreciate the value of interventions they are undertaking or considering to undertake.

Of great importance is the fact that value-based medicine will differentiate the value and the cost-utility (dollars spent for this value) among all interventions utilized for the treatment of ARMD when evidence-based clinical data alone do not allow this form of differentiation.¹³ Both clinicians and patients will readily realize which treatments provide the most value, thus empowering both physicians and patients in an era of considerable marketing outside the medical arena.

The improvements in quality of life for US citizens and the financial gains for the country conferred by current and future therapies for ARMD are immense.^51,52 The effect upon GDP discussed herein, although it takes into account wages lost from ARMD, does not include disability payments, healthcare costs, caregiver costs, transportation costs, and family caregiver costs (if they prevent earning a salary). Thus, the numbers presented herein are lower than if these other costs were included.

In summary, the substantial public health burden of ARMD includes both its adverse effects upon quality of life and upon the economy. ARMD causes a marked decrease in quality of life, the diminution of which is underestimated by multiple sectors of society, including physicians who regularly care for ARMD patients. The adverse effect of ARMD upon the GDP is also considerable. Interventions that improve the morbidity caused by ARMD have the potential to greatly benefit the quality of life of individual patients^51,52 as well as the overall economic well-being of the country.