Customize your JAMA Network experience by selecting one or more topics from the list below.
Neely DC, Bray KJ, Huisingh CE, Clark ME, McGwin G, Owsley C. Prevalence of Undiagnosed Age-Related Macular Degeneration in Primary Eye Care. JAMA Ophthalmol. 2017;135(6):570–575. doi:10.1001/jamaophthalmol.2017.0830
To what extent is age-related macular degeneration (AMD) undiagnosed by primary eye care physicians when AMD is actually present?
In this cross-sectional study, 320 of 1288 eyes had AMD despite no diagnosis of AMD in the primary eye care medical record, including 30.0% with undiagnosed large drusen.
As treatments and monitoring strategies for early AMD are refined in the future, these data suggest that improvements for correct, prompt identification of AMD seem to be warranted if subsequent interventions for early AMD safely avoid vision loss.
Age-related macular degeneration (AMD) is the leading cause of irreversible vision impairment in older adults in the United States, yet little is known about whether AMD is appropriately diagnosed in primary eye care.
To examine the prevalence of eyes with AMD in patients seen in primary eye care clinics who purportedly have normal macular health per their medical record and the association of AMD with patient and physician characteristics.
Design, Setting, and Participants
In this cross-sectional study of primary eye care practices in Birmingham, Alabama, 644 persons 60 years or older with normal macular health per medical record based on their most recent dilated comprehensive eye examination by a primary eye care ophthalmologist or optometrist were enrolled from May 1, 2009, through December 31, 2011. Data analysis was performed from May 1, 2016, through December 20, 2016.
Main Outcomes and Measures
Presence of AMD as defined by the Clinical Age-Related Maculopathy Staging system based on color fundus photography and a masked grader. Types of AMD-associated lesions were noted. Patient health and physician characteristics were collected.
The sample consisted of 1288 eyes from 644 participants (231 [35.9%] male and 413 [64.1%] female; mean [SD] age, 69.4 [6.1] years; 611 white [94.9%]) seen by 31 primary eye care ophthalmologists or optometrists. A total of 968 eyes (75.2%) had no AMD, in agreement with their medical record; 320 (24.8%) had AMD despite no diagnosis of AMD in the medical record. Among eyes with undiagnosed AMD, 32 (10.0%) had hyperpigmentation, 43 (13.4%) had hypopigmentation, 249 (77.8%) had small drusen, 250 (78.1%) had intermediate drusen, and 96 (30.0%) had large drusen. Undiagnosed AMD was associated with older patient age (odds ratio [OR], 1.06; 95% CI, 1.04-1.09; P < .001), male sex (age-adjusted OR, 1.39; 95% CI, 1.02-1.91; P = .04), and less than a high school education (age-adjusted OR, 2.40; 95% CI, 1.03-5.62; P = .04). Prevalence of undiagnosed AMD was not different for ophthalmologists and optometrists (age adjusted OR, 0.99; 95% CI, 0.71-1.36; P = .94).
Conclusions and Relevance
Approximately 25.0% of eyes deemed to be normal based on dilated eye examination by primary eye care physicians had macular characteristics that indicated AMD revealed by fundus photography and trained raters. A total of 30.0% of eyes with undiagnosed AMD had AMD with large drusen that would have been treatable with nutritional supplements had it been diagnosed. Improved AMD detection strategies may be needed in primary eye care as more effective treatment strategies for early AMD become available in the coming years.
Age-related macular degeneration (AMD) is a complex, multifactorial disease that affects older adults in which central retinal photoreceptors are lost by atrophic or neovascular processes, resulting in progressive loss of central acuity. Approximately 14 million Americans have AMD, and as the baby boomer population ages, this public health problem is expected to worsen.1,2 Even in early AMD, there is a significant personal burden attributable to reading difficulty,3,4 driving cessation,5,6 decreased health-related quality of life,7 depression,8 and anxiety about possible blindness.9 Given the high prevalence of AMD in the older adult population and the availability of treatment options to slow some types of AMD, it is worthwhile to examine the extent to which AMD is detected in the primary eye care clinic and to what extent AMD may be underdiagnosed in this setting. Although there are currently no proven effective treatments for early AMD, the next and emerging frontier for AMD treatment development is targeted at early AMD. Within this context, early detection of AMD will be beneficial for the individual and public health. Ophthalmologists and optometrists whose practices focus on primary care will ultimately be at the center of this process, tasked with the responsibility of early AMD detection in the increasing numbers of older eyes at risk for AMD.
This study uses a large sample of older adults who were judged to have normal macular health in both eyes by their primary eye care ophthalmologist or optometrist based on a dilated eye examination.10,11 We report the prevalence of eyes with AMD in this sample for which color fundus photography and photographic grading using a clinical AMD classification system were implemented to confirm the presence vs absence of AMD. How patient and physician characteristics relate to the prevalence of eyes with undiagnosed AMD is also reported.
Study participants consisted of 644 adults 60 years or older enrolled in the Alabama Study on Early Age-Related Macular Degeneration (ALSTAR) from May 1, 2009, through December 31, 2011.10,11 Data analysis was performed from May 1, 2016, through December 20, 2016. Participants were recruited from primary eye care practices in Birmingham, Alabama; these practices care for patients with Medicare, Medicaid, or private health insurance. Written informed consent was obtained from participants after the nature and purpose of the study were described. All data were deidentified. The participants received a stipend. This study was approved by the Institutional Review Board of the University of Alabama at Birmingham and followed the tenets of the Declaration of Helsinki.12
To be eligible, the person’s medical record from the most recent comprehensive dilated examination could not indicate a diagnosis of AMD in either eye and the medical record notes could not contain terms that signified the signs of AMD (eg, drusen, hyperpigmentation or hypopigmentation, geographic atrophy [GA], choroidal neovascularization). ALSTAR exclusion criteria were no previous diagnoses of glaucoma, other retinal conditions, optic nerve conditions, corneal disease, diabetes, Alzheimer disease, Parkinson disease, brain injury, or other neurologic or psychiatric conditions as revealed by the medical record or by self-report. For the purposes of the study, 3-field, digital color fundus photography (Carl Zeiss Meditec 450 Plus camera) was completed on each eye after dilation with tropicamide, 1%, and phenylephrine hydrochloride, 2.5%. A trained and experienced grader (M.E.C.) evaluated fundus photographs for the presence and severity of AMD by using the Clinical Age-Related Maculopathy Staging (CARMS) system.13 Digital fundus imaging software (Sonomed Escalon’s OphthaVision, version 4.0) was used to place a 6-mm Early Treatment Diabetic Retinopathy Study grid centered on the fovea in field 2 (macula) to approximate a 2-mm disc diameter (based on a 1.5-mm disc diameter) distance from the fovea for grading. Drusen sizing circles were available to determine drusen sizes of 63 µm or less, 63 µm or greater but 125 µm or less, and 125 µm or greater (C0 and C1 circle sizes in the Wisconsin grading protocol) to determine drusen load. Drusen size and number; retinal pigmentation level; retinal pigment epithelial detachment size, type, and location; choroidal neovascular membranes; and/or disciform scar were all assessed according to the CARMS designations. The CARMS was used in this study because it was designed specifically for use in clinical practice in which usual care consists of a dilated fundus examination using a slitlamp. Furthermore, the CARMS has been validated against color fundus photographs evaluated by expert graders. Thus, the CARMS was well suited for this study because the focus is on the identification of AMD by primary eye care physicians. The CARMS consists of a 5-category scale, briefly summarized here. Stage 1 of the CARMS is no drusen or fewer than 10 small drusen defined as less than 63 μm in diameter, which serves as our case definition of normal macular health; stage 2 is approximately 10 or more small drusen or fewer than 15 intermediate drusen or pigment abnormalities associated with AMD; stage 3 is approximately 15 or more intermediate drusen or any large drusen; stage 4 is GA of the macula center or noncentral GA at least 350 μm in size; and stage 5 is exudative AMD. In performing CARMS grading, the grader noted when an eye had any of the following characteristics or lesions: hyperpigmentation, hypopigmentation, more than 10 small drusen, intermediate drusen (≥63 μm but <125 μm in diameter), large drusen (≥125 μm in diameter), drusenoid retinal pigment epithelial detachment, serous retinal pigment epithelial detachment, GA, choroidal neovascularization, or disciform scar.
Information on demographic characteristics, medical comorbidities, smoking status, family history of AMD, and the use of Age-Related Eye Disease Study (AREDS) formulation nutritional supplements was obtained through participant interview. General cognitive status was assessed using the Mini-Mental State Examination.14 Visual acuity, pseudophakia status, the date of the dilated comprehensive eye examination, and whether the primary eye care physician was an ophthalmologist or optometrist were abstracted from the medical record. Logistic regression accounting for clustering of eyes within patients using generalized estimating equations was used to compare characteristics between those with and without undiagnosed AMD and to generate crude and age-adjusted odds ratios (ORs) and corresponding 95% CIs. Statistical significance was defined as P < .05 (2-tailed).
The sample consisted of 1288 eyes from 644 participants (231 [35.9%] male and 413 [64.1%] female; mean [SD] age, 69.4 [6.1] years; 611 white [94.9%]) seen by 31 primary eye care ophthalmologists or optometrists. Table 1 gives the demographic characteristics of the participants. Most participants were aged in their 60s (380 [59.0%]) or 70s (226 [35.1%]), with 38 (5.9%) of the sample in their 80s. Participants were recruited from 31 eye care physicians (17 ophthalmologists, 14 optometrists).
There were 968 eyes (75.2%) with no AMD according to the CARMS system in agreement with their medical record; 320 eyes (24.8%) had AMD based on the CARMS system even though there was no AMD diagnosis in the medical record (Table 2). The AMD stage for these 320 eyes was as follows: 220 (68.8%) had early AMD, 99 (30.9%) had intermediate AMD, and 1 (0.3%) had GA. For these undiagnosed AMD eyes, Table 3 lists the types and number of macular characteristics indicative of AMD noted during fundus photograph grading. Approximately three-fourths of the 320 undiagnosed eyes had 10 or more small drusen (249 [77.8%]) and/or intermediate drusen (250 [78.1%]), with 96 (30.0%) of undiagnosed eyes having large drusen. Older adults were more likely to have small, intermediate, and large drusen than younger adults. Hyperpigmentation (32 eyes [10.0%]) and hypopigmentation (43 eyes [13.4%]) were less commonly represented. Older adults were also more likely to have hyperpigmentation than younger age groups. No eyes had serous retinal pigment epithelial detachment, choroidal neovascular membrane, or disciform scar.
Eyes more likely to be undiagnosed for AMD in the medical record were from older persons, men, those with less than a high school education, those with worse visual acuity, those who were pseudophakic, or those who had hypertension (Table 2). Undiagnosed AMD eyes were slightly more common in persons with better mental status (Table 2). Other medical comorbidities, smoking status, and family history of AMD were unrelated to whether an eye was undiagnosed. Time since the last dilated eye examination was not different between those with and without undiagnosed AMD. The prevalence of undiagnosed AMD was not different for ophthalmologists and optometrists. The number of eyes in patients using AREDS nutritional supplements was the same regardless of whether the eyes were in normal macular health (21 [2.2%]) or had AMD that was undiagnosed (7 [2.2%]).
After adjusting for age, men had significantly higher odds of having undiagnosed AMD than women (age-adjusted OR, 1.39; 95% CI, 1.02-1.91) (Table 4). Compared with those with some college or more, those with less than a high school education were significantly more likely to have undiagnosed AMD (age-adjusted OR, 2.40; 95% CI, 1.03-5.62) (Table 4).
Age-related macular degeneration represents a significant public health concern given its high prevalence in the older adult population and its threat to central vision. This study suggests that AMD is sometimes not diagnosed in older adults receiving a dilated comprehensive eye examination in primary eye care despite its presence. We found that 25% of eyes with no diagnosis of AMD in the medical record or notations about AMD characteristics in the fundus examination actually had AMD according to a clinical classification staging system. Furthermore, approximately 30% of those undiagnosed eyes had large drusen, indicating these patients would have been candidates for therapeutic intervention with nutritional supplements15; however, only 3 (3.0%) of 99 eyes with intermediate AMD were from persons taking AREDS formulation nutritional supplements.
It is important to consider factors that could be contributing to the absence of an AMD diagnosis. The most common AMD funduscopic characteristics that were not noted in the medical record but were present were drusen, with 84.4% of undiagnosed eyes having intermediate or large lesions. Drusen are small targets to visually discern in a fundus examination and often appear sparsely or even in isolation in earlier stages of AMD, making visual detection challenging during brief inspection. In contrast, hyperpigmentation and hypopigmentation, which present as more spatially distributed patterns, were less commonly present in undiagnosed eyes. In our current era of emphasizing improved efficiency of eye care yet maintaining high quality, it is critical that sufficient attention be devoted to the dilated fundus examination during the comprehensive eye examination for older adults so that treatable disease, such as AMD, is not overlooked. Another factor that could impede detection of AMD lesions is the presence of cataract, which could hamper visualization of the fundus even when cataract is in its early stages. However, the prevalence of AMD for undiagnosed eyes was similar for phakic and pseudophakic eyes in our sample, suggesting that a poor view of the fundus was not a contributing factor to nondiagnosis.
Age-related macular degeneration was more likely to be present yet undiagnosed in older eyes. Older eyes were also more likely to have drusen of all sizes. It is widely known that risk of AMD increases with increasing age in late adulthood; thus, it is puzzling that primary eye care physicians would be less likely to identify AMD in older eyes. Instead, one would surmise that a patient’s advanced age would lead primary eye care physicians to more closely scrutinize the fundus for signs of AMD. Men were more likely to have undiagnosed AMD. This finding is the opposite of a large body of research on sex-related health disparities in chronic diseases of aging that indicates that women are more likely to have underdiagnosed conditions (eg, heart disease, chronic obstructive pulmonary disease16-19). Persons with less than a high school education were more likely to have undiagnosed conditions even though AMD was present. This finding is reminiscent of a study20 that found that educational attainment is closely tied to health.
We found that ophthalmologists and optometrists working in primary eye care were equally likely to miss a diagnosis of AMD. This finding suggests that both types of physicians could benefit from better training in identifying AMD in the dilated fundus examination. Ultimately, given the rapid advancements in retinal imaging technologies, AMD underdiagnosis could potentially dissipate if affordable yet high-quality retinal imaging modalities are implemented in primary eye care settings, an issue worthy of further investigation.
A common approach to identifying eyes as normal control eyes in studies on AMD is to recruit patients whose dilated eye examinations by primary eye care physicians do not reveal diagnoses or impressions of AMD. Our study suggests that this approach is likely to be problematic because all the eyes in this study had such a designation by the primary eye care physician. Our results suggest that many of these so-called normal control eyes will actually have early or intermediate AMD. A more objective and rigorous approach in defining normality (ie, the absence of AMD) in AMD studies is the use of fundus photography and standard and accepted AMD classification systems used by trained and masked graders. Clear case definitions of the absence of AMD will also facilitate comparison across studies.21
In using fundus grading systems such as the CARMS in AMD research when distinctions need to be made between normal macular health and AMD presence, a cut point must be selected to create a case definition of normality beyond which the eye is considered to have AMD. Setting a cut point is somewhat arbitrary, for example, in terms of drusen number and size. In the present study, stage 1 of the CARMS was defined as normal macular health; a strong rationale for adopting this definition was that stage 1 specifies no or minimal presence of small drusen and no pigmentary changes. Moving the cut point of normality to stage 2, which signifies the presence of pigmentary changes, more small drusen, and the presence of fewer than 15 intermediate drusen, would correspondingly reduce the percentage of eyes that were undiagnosed. However, defining eyes as normal only when they have no drusen, a stricter definition, would increase the undiagnosed rate. Thus, readers need to be aware of the subtleties in definitions of normality and disease when interpreting and comparing studies on underdiagnosis.
A strength of this study is that it is the first report, to our knowledge, to address the extent to which AMD is underdiagnosed by primary eye care physicians in the United States. The sample is large, with 1288 eyes. The presence of AMD was determined by a recognized clinical classification system and a grader with established reliability of judgment. Factors related to nondiagnosis of AMD were evaluated, identifying 2 intriguing factors that increase the risk of nondiagnosis, namely, advanced age and male sex. Limitations should also be acknowledged. Although the study focused on the prevalence of undiagnosed AMD among those whose primary eye care physicians indicated they did not have AMD, the study design allowed us to examine the converse: the prevalence of AMD among those who have been diagnosed with AMD in primary eye care but do not have it. The time that elapsed between the dilated comprehensive eye examination and the date of fundus photography varied in the sample. Age-related macular degeneration may not have been present on the date of the eye examination but may have developed in the ensuing months. However, the prevalence of nondiagnosis was highly similar regardless of how much time elapsed between eye examination and photography. The results of this study are based on the primary eye care physicians who referred patients to ALSTAR; the extent to which these findings generalize to other eye care physicians whose patient case mix may be different remains to be determined.
This study suggests that underdiagnosis of AMD is not uncommon in primary eye care. One of 4 eyes studied was not diagnosed with AMD in a dilated fundus examination performed by primary care ophthalmologists and optometrists despite these eyes having macular characteristics indicative of AMD. This finding is concerning because 30.0% of eyes had drusen characteristics consistent with intermediate AMD, which is treatable with nutritional supplementation. The reasons underlying AMD underdiagnosis in primary eye care remain unclear. As treatments for the earliest stages of AMD are developed in the coming years, correct identification of AMD in primary eye care will be critical for routing patients to treatment as soon as possible so that the disease can be treated in its earliest phases and central vision loss avoided.
Corresponding Author: David C. Neely, MD, Department of Ophthalmology, University of Alabama at Birmingham, 700 S 18th St, Ste 601, Birmingham, AL 35294-0009 (email@example.com).
Accepted for Publication: March 10, 2017.
Published Online: April 27, 2017. doi:10.1001/jamaophthalmol.2017.0830
Author Contributions: Drs Owsley and Huisingh had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Neely, Huisingh, McGwin, Owsley.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Neely, Huisingh, McGwin, Owsley.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Bray, Huisingh, McGwin, Owsley.
Obtained funding: Owsley.
Administrative, technical, or material support: Neely, Clark, Owsley.
Study supervision: Neely, Owsley.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This research was funded by grant R01AG04212 from the National Institute on Aging, National Institutes of Health (Dr Owsley); the Dorsett Davis Discovery Fund (Dr Owsley); the Alfreda J. Schueler Trust (Dr Owsley), the EyeSight Foundation of Alabama (Dr Owsley); Research to Prevent Blindness (Dr Owsley); and grant P30EY003039 from the National Eye Institute, National Institutes of Health (Dr Owsley).
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.