Comparison of patients with low vision and those with cataract with the reference group on mean (SEM) National Eye Institute Visual Function Questionnaire (NEI-VFQ) subscale scores. Linear regression results for 2-group comparisons with the reference group, adjusted for age, sex, race, and medical comorbidities. Asterisk indicates P≤.05.
Mangione CM, Lee PP, Pitts J, Gutierrez P, Berry S, Hays RD, for the NEI-VFQ Field Test Investigators. Psychometric Properties of the National Eye Institute Visual Function Questionnaire (NEI-VFQ). Arch Ophthalmol. 1998;116(11):1496-1504. doi:10.1001/archopht.116.11.1496
To test the reliability and validity of the 51-item Field Test Version of the National Eye Institute Visual Function Questionnaire (NEI-VFQ) across 5 common chronic eye conditions.
Prospective observational cohort study of persons with 1 of 5 chronic eye diseases who were scheduled for nonurgent visits in ophthalmology practices or had low vision from any cause, and a reference sample of persons without eye disease.
Six university-based ophthalmology practices and the National Eye Institute Clinical Center, Bethesda, Md.
Eligible participants had to have 1 of the following eye conditions: age-related cataracts, age-related macular degeneration, diabetic retinopathy, primary open-angle glaucoma, cytomegalovirus retinitis, or low vision from any cause. Each of the 7 sites also enrolled persons in a reference sample. Reference sample participants had no evidence of underlying eye disease but were scheduled for either screening eye examinations or correction of refractive error. All eligible persons had to be aged 21 years or older, English speaking, and cognitively able to give informed consent and participate in a health status interview.
Measurements and Main Results
To provide the data needed to assess the reliability and validity of the 51-item NEI-VFQ, all subjects completed an interview that consisted of the 51-item NEI-VFQ, the Medical Outcomes Study 36-Item Short-Form Health Survey, and at least 1 measure of vision-targeted functional status. Estimates of internal consistency and test-retest reproducibility indicate that the 51-item NEI-VFQ is reliable. Tests of association with other scales and clinical variables support the construct validity of the survey.
In this cross-sectional study, the 51-item NEI-VFQ seems to be reliable and valid and should be a useful tool for group-level comparisons of vision-targeted, health-related quality of life in clinical research. Additionally, the psychometric properties of the NEI-VFQ were not influenced by the type or severity of the underlying eye disease, suggesting that the measure will provide reproducible and valid data when used across multiple eye conditions.
TRADITIONAL CLINICAL measures of vision, such as Snellen visual acuity, may fail to assess many aspects of visual disability that are identified by persons as being important for their daily functioning and well-being.1- 13 Except for investigations that were specifically designed to capture outcome after cataract surgery and intraocular lens implantation,4,5,10,11,14,15 patients' perceptions of visual functioning and other aspects of health-related quality of life (HRQOL) have only recently been incorporated into ophthalmologic research and clinical care.9,16- 18 Previously developed measures of vision-targeted functioning have principally emphasized difficulty with tasks and symptoms3,6,9,10,13,16 rather than examining the influence of visual disability on other aspects of HRQOL such as emotional well-being or social function. Most surveys with published reliability and validity are designed for use among persons with cataract3,6,13 or for older persons in general.16 Surveys designed with 1 condition in mind are likely to have excellent content validity for persons with that specific condition. However, they may not fully describe the range of disability and impairment experienced by persons with other eye diseases or the influence of limitations in visual functioning on other dimensions of HRQOL.
The goals of this multi-institution psychometric field test were to assess the reliability and validity of a new vision-targeted measure of HRQOL, the 51-item National Eye Institute Visual Function Questionnaire (NEI-VFQ). The content area for the NEI-VFQ was uniquely derived from focus groups with visually impaired persons.19 This aticle presents data that examine the reliability and validity of the 51-item NEI-VFQ when it is used among persons with common chronic eye conditions.
A prospective sample of persons with chronic eye diseases and reference group participants who met eligibility criteria was recruited from 7 tertiary care ophthalmology practices between July 1995 and March 1996. Investigators at each site approached all potentially eligible subjects who met the general and condition-specific eligibility criteria described in this section. Those who refused to participate in the interview component of the study were asked to consent to a confidential medical record review to obtain information on visual acuity, presence of eye diseases, age, and sex. These data were used to monitor for selection bias between consenting and nonconsenting subjects and to ensure that consent rates were uniform across the different sites. Each clinical site was given condition-specific target sample sizes. These target sample sizes were influenced by the type of patients seen most often in each of the settings.
Research staff reviewed the medical records of all persons who were scheduled for nonurgent appointments to identify potential participants. To be eligible for participation, patients had to be at least 21 years old, English speaking, and had to pass an abbreviated version of the Folstein Mini-Mental State examination,20 in which persons were first asked the orientation, registration, attention, and calculation items. Only those who made an error on these were asked to complete the remainder of the mental status examination. To reduce bias from referrals and loss to follow-up, only persons who had been followed up for at least 3 months or had had at least 2 previous visits were eligible for participation. All subjects underwent a comprehensive dilated eye examination after completing the interview portion of the investigation. The study protocol was approved by all appropriate institutional review boards and participants gave written informed consent prior to enrollment.
Participants had 1 of the following eye conditions: age-related cataracts, age-related macular degeneration (ARMD), diabetic retinopathy, primary open-angle glaucoma (POAG), cytomegalovirus (CMV) retinitis, or low vision from any cause, including the previous diagnoses. Additionally, each site enrolled age-matched reference participants who had no evidence of underlying eye disease but were scheduled for either screening dilated eye examinations or correction of refractive error.
Condition-specific eligibility criteria were imposed to ensure adequate representation of disease severity as follows:
Half of the subjects with diabetic retinopathy had to have at least bilateral mild to severe nonproliferative diabetic retinopathy with habitual visual acuity in the better eye of 20/30 or worse, and half had past treatment with scatter (panretinal) photocoagulation in both eyes for proliferative retinopathy with or without previous focal laser treatment for macular edema. Persons with vitrectomy in either eye during the previous 3 months were excluded.
Those with ARMD had to have evidence of retinal pigment epithelium (RPE) changes, choroidal neovascular membrane, central foveal drusen 125 µm or larger in diameter, or 1 of the following in each eye: disciform scar, past laser treatment within 500 µm of the fovea, RPE detachment, or geographic atrophy involving the fovea.
Those with age-related cataract had to have bilateral involvement exceeding the Age-Related Eye Disease Study (AREDS) nuclear or cortical standard 2, or exceeding posterior subcapsular standard 1, and currently corrected visual acuity in the better eye that was 20/30 or worse.
Those with POAG had to have evidence of at least a nasal step or abnormal hemifield test in each eye from a automated perimetry test (Humphrey Field Analyzer 24-2 or 30-2, Humphrey Instruments, San Leandro, Calif) obtained during any visit within the previous 6 months, evidence of optic nerve or nerve fiber layer defects on direct ophthalmoscopy, or 78- or 90-diopter binocular biomicroscopy, and at least 1 previous or current intraocular pressure measurement in each eye that was greater than 21 mm Hg. Patients with a history of laser surgery, filtering or seton surgery, or cyclodestructive procedures within 3 months of enrollment were excluded.
Participants with CMV retinitis had to have at least monocular necrotic and exudative changes visible with ophthalmoscopy.
Persons were eligible for the low vision group if they had a best-corrected visual acuity in the better eye that was 20/70 or worse or a central visual field of 20° or less. To ensure that legally blind persons were included in the study, 50% of the low vision participants had to have a best-corrected visual acuity in the better eye that was 20/200 or worse.
Persons enrolled in the reference group had to have no evidence of underlying eye disease except for refractive error correctable to at least 20/25 in their worse eye. Reference group members could have subclinical morphologic retinal or macular changes that did not meet criteria for ARMD and early lens opacities that were less than AREDS cortical or nuclear grade 2 or posterior subcapsular grade 1.
Additionally, a matching procedure was used to ensure that the reference sample had a similar age distribution as the participants with diseases. To accomplish this, the centers were instructed to identify and enroll reference participants whose age was within 5 years of the average age for the previous 3 persons who were enrolled into condition-specific groups. During the enrollment phase of the project, demographic characteristics were monitored by the coordinating center to ensure that overall at least 50% of the cohort was female and 30% were from underrepresented minority groups.
All participants were interviewed with the 51-item NEI-VFQ to assess vision-targeted HRQOL and the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) to measure multidimensional HRQOL. To avoid a possible bias from information learned during the dilated eye examination, patients completed the interview prior to seeing the ophthalmologist. The 51-item NEI-VFQ is a vision-targeted survey that assesses the influence of visual disability on HRQOL. The content of the NEI-VFQ was empirically derived from a multicondition focus group process.19 This measure includes the following: multi-item scales to rate overall health on a 5-level scale that ranges from excellent to poor and overall vision on a 6-level scale that ranges from excellent to blind; multi-item scales that assess difficulty with near vision activities, difficulty with distance vision activities, limitations in social functioning due to vision, role limitations due to vision, dependency on others due to vision, mental health symptoms due to vision, future expectations for vision, driving difficulties, and pain and discomfort in or around the eyes; and single items to assess limitations with peripheral vision and color vision (Table 1). Scales were scored on a 0 to 100 scale, in which 100 represents the best possible score on the measure and 0 represents the worst.
To provide data for the planned tests of correlational validity, each participant also completed the SF-36, a 36-item generic HRQOL measure designed for chronically ill medical patients.21- 23 Responses on the SF-36 were aggregated to generate a physical health component score and a mental health component score.24 The composite scores are standardized into z scores using means and SDs from a general US population sample. Each score is then transformed linearly to a Student t score where the mean is 50 and the SD is 10. Inclusion of the SF-36 provided the data needed to assess the validity of specific NEI-VFQ scales that were designed to capture the influence of vision loss on physical and mental dimensions of health. Participants also completed a 16-item medical comorbidity checklist taken from the Medical Outcomes Study.25
To provide the needed data for additional tests of construct validity, subjects with either ARMD or cataract and reference group members were randomly assigned to complete 1 of 3 surveys of vision-targeted functioning. These measures included the Visual Activities Questionnaire (VAQ), a measure of visual functioning designed for older persons,16 the VF-14,13 and the Activities of Daily Vision Scale (ADVS).6 The VF-14 and the ADVS are short questionnaires designed to assess vision-targeted disability from cataract and to assess improvements after surgery. To enrich the responses of those with other eye conditions, a randomly assigned 50% of the low vision subjects also completed the VAQ.
A convenience sample of persons from all condition-specific groups except for cataract were asked to return to the participating site 2 weeks (±1 week) after enrollment to complete a second interview, which consisted of the 51-item NEI-VFQ alone. These data were used to assess the test-retest reliability of the survey among persons with stable eye disease.
All participants underwent a comprehensive dilated examination that included an assessment of current eye diseases and previous history of ophthalmic surgical procedures. Binocular and monocular Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity was measured while patients were wearing their current or "walking about" correction.26 Patients with visual acuity so poor that they could not read any of the largest letters at 4 m were tested at 1 m and for light perception. Manifest refraction was also done using a logarithmic visual acuity chart. The presence and severity of cataracts were graded during a slitlamp examination using the AREDS reference standards (Phase II Manual of Operations, December 11, 1997, EMMES Corp, Potomac, Md). Each participant had a complete fundus examination to rule out significant additional ocular abnormalities and to grade the severity of diabetic retinopathy, ARMD, POAG, and CMV retinitis. The fundus had to be visualized for patients to qualify for the study. For persons with POAG, data from automated perimetry were obtained either at the enrollment visit or within 6 months before enrollment. Patients with other diseases that cause visual field loss, such as optic neuritis, were excluded. Only "reliable" visual fields, defined by false-positive results, false-negative results, or fixation losses not exceeding 33% were used.27
Advanced Glaucoma Intervention Study (AGIS) scores were calculated for each eye based on the deviation plot from the automated perimetry visual fields.27 The AGIS scores represent the number and depth of depressed visual field sites found in less than 5% of normal values and an index for quantifying visual field defects for the entire eye. The AGIS scores can range from 0, indicating no defects, to 20, indicating near or complete visual field loss. Scores are calculated from points awarded to 3 areas: a maximum of 9 points can be awarded to the upper and lower fields, while the nasal area can be awarded a maximum of 2 points. An AGIS score was calculated for each participant's better and worse eye.
The distribution of age, sex, race, education, work status, and medical comorbidities was compared by condition with analysis of variance for continuous variables and χ2 tests for polychotomous variables. All tests of association were considered to be statistically significant at P≤.05. Mean (SD) 51-item NEI-VFQ scale scores were calculated by diagnosis and for the reference group. The proportion of persons with NEI-VFQ scores at the ceilings and floors for each scale was also examined.
To estimate the internal consistency reliability of the NEI-VFQ, we calculated the Cronbach α28 for the multi-item scales. Coefficient α increases with the number of scale items and the size of their covariances. We also compared the magnitude of the coefficient α by condition to determine whether the reliability of the survey varied by diagnosis.
To estimate the test-retest reliability of each 51-item NEI-VFQ scale, we calculated intraclass correlations (2-way random effects model) using data collected from 96 participants who completed the survey 2 ± 1 weeks apart. Additionally, we determined whether there were statistically significant differences in mean scores for each NEI-VFQ scale across the 2-week retest interval.
Validity assesses how well a measure adequately represents the domains or constructs of interest. To evaluate the construct validity of the observed NEI-VFQ scores across eye conditions, we estimated the statistical significance of a series of 3 comparisons.
The first comparisons assess the extent to which NEI-VFQ scales correlate with other scales measuring similar or different dimensions of health such as scores from published surveys that measure (1) overall vision-targeted functioning (VF-14 and ADVS) and specific dimensions of vision-targeted functioning such as the VAQ subscales that assess difficulty with peripheral vision and color vision; (2) physical health as estimated by the SF-36 physical component score; and (3) emotional well-being as estimated by the SF-36 mental component score.
The second set of comparisons assesses whether participants with poorer vision have significantly lower scores on the NEI-VFQ than those with better vision. Although condition-specific means were calculated for all groups, to avoid finding significant differences by chance alone, only selected preplanned comparisons were made. These statistical comparisons of adjusted mean scores included (1) scores from persons with low vision vs the reference group; (2) scores from persons with visually significant cataract vs the reference group; and (3) comparisons of selected subscales, such as the near vision scale for persons with ARMD vs the reference group and the peripheral vision score for persons with glaucoma vs the reference group. Linear regression analyses were used to adjust for between-group differences in characteristics likely to influence HRQOL, such as age, sex, race, and number of medical comorbidities.
The third set of statistical tests assesses the magnitude of correlations between responses on the NEI-VFQ and performance-based measures that are part of the ophthalmologic examination. Correlations between ETDRS visual acuity and responses on the NEI-VFQ should be greatest for the scales that capture difficulty with activities that require central visual acuity, such as the near vision scale and the driving scale, and lowest for NEI-VFQ scales that capture other aspects of vision-targeted HRQOL, such as ocular pain or expectations for future visual function. To test the significance of these associations, Spearman correlation coefficients were calculated between mean NEI-VFQ scores and the ordered categorical visual acuity and visual field in the better eye, worse eye, and both eyes. Finally, to determine the dimensionality of the scale scores, the interscale nonparametric correlations are reported.
Seven hundred eighteen eligible persons were approached for participation and 598 (83%) gave informed consent and completed the health status interview and clinical examination. Consent rates were similar across sites. The mean age, sex distribution, and median visual acuity in the better eye were statistically similar for those who participated and those who refused. A convenience sample of 96 participants returned to the practice to complete the 51-item NEI-VFQ a second time 1 to 3 weeks after enrollment. These paired data from participants were used to assess test-retest reliability.
Sixty percent of the participants were female, with the greatest proportion among those with low vision (68%) and the smallest proportion among those with CMV retinitis (5%). Mean age by condition varied from 76 years for those with ARMD to 39 years for those with CMV retinitis. For the cohort overall, 62% were white and 30% were African American. Persons with eye diseases that affect older persons had more coexisting medical conditions and were more likely to be retired. The mean time required to complete the interviewer-administered version of the 51-item NEI-VFQ was 15.5 minutes (SD, 6.1 minutes).
Members of the low vision group had the poorest visual acuity in their better eye with a mean ETDRS score of 30 points (visual acuity equivalent approximately 20/250), and those in the reference group and the CMV retinitis group had the best visual acuity, with a mean ETDRS score of 85 points (visual acuity equivalent approximately 20/20). Except for CMV retinitis, most of the participants had had their eye condition for longer than 2 years (Table 2).
Internal consistency estimates for the 11 multi-item NEI-VFQ scales were between 0.94 and 0.66 (Table 3). Among those with eye diseases, 86% of the internal consistency estimates were greater than or equal to 0.7, indicating that the measure has acceptable reliability for group-level comparisons.29 The NEI-VFQ subscale test-retest reliability as estimated by intraclass correlations were between 0.91 and 0.68 for all scales. Except for the "expectations for visual functioning" scale, mean scores over time were not significantly different.
The mean 51-item NEI-VFQ scale scores varied by condition (Table 4). Scales that are likely to be influenced by deficits in central acuity (near, distance, and driving) were lowest for those in the low vision group and for those with ARMD. The reference group participants had the best scale scores across all dimensions of vision-targeted HRQOL captured by the NEI-VFQ. Persons in the low vision group reported the most difficulty with peripheral vision and color vision. Very few participants from any group were at the floor of the scales. However, a sizable proportion of persons, especially those with milder eye diseases, had scores at the ceiling for specific scales. This was particularly true for dimensions of vision-targeted HRQOL that were unlikely to be affected by these conditions, such as ocular pain or difficulty with color vision. Persons with low vision reported the greatest dependency on others and the greatest mental distress due to their vision.
The strongest evidence of construct validity was the high correlations between the NEI-VFQ scales that are activity-oriented such as the near, distance, and driving scales, and the other vision-targeted measures that assess difficulty with vision-specific activities (the VF-14 and ADVS, r=0.9-0.6, P=.001) (Table 5). Of note, responses to the 6-level general vision rating questions also were highly correlated with the VF-14 and ADVS scores (r=0.7; P<.001). We observed unexpected high correlations between the NEI-VFQ scale designed to capture mental distress from visual disability and the VF-14 and ADVS scores (r=0.7, P=.001). The peripheral and color vision questions from the VAQ correlated significantly with the NEI-VFQ items designed to assess these 2 dimensions of vision and had lower correlations with NEI-VFQ scales that capture other dimensions of vision-targeted HRQOL. The SF-36 physical component score had low correlations with most NEI-VFQ vision-related subscales, yet had a high correlation with the NEI-VFQ general health rating scale, suggesting that the SF-36 physical component score is more strongly influenced by poorer general health than limitations with vision-targeted HRQOL. Finally, the SF-36 mental component score had the highest correlation with the NEI-VFQ scale that captures mental distress due to vision (r=0.4, P=.001).
For 11 of 13 NEI-VFQ scales, we found that the adjusted mean scores for persons in the reference group were significantly higher than scores for those with either low vision or visually significant cataract (Figure 1). Patients with cataracts had higher expectations for future visual functioning preoperatively than those in either the reference or low vision groups. Levels of ocular pain were similar for all 3 groups. Adjusted mean (SEM) scores for the NEI-VFQ near and distance vision scales were higher for reference group participants (90  and 93 , respectively) than for those in the ARMD group (58  and 66 , respectively). Finally, adjusted mean (SEM) scores for the NEI-VFQ peripheral vision question for reference group members vs those with glaucoma were 97 (2) and 76 (2), respectively. All of these means were significantly different at P≤.05.
Correlations between responses on the NEI-VFQ and ETDRS visual acuity were moderate for the scales that reflect the degree of difficulty that a person has with common visual activities. Correlations between each of the scales and visual acuity in the better and worse eyes were similar in magnitude (Table 6). Among those with glaucoma, the magnitude of visual field loss correlated moderately with limitations in social function and dependency on others due to vision (Table 7). Although statistically significant, correlations were lower for field loss and the distance, near, and peripheral vision scales.
Interscale correlations suggest that the 12 descriptive scales represent multiple dimensions of vision-targeted HRQOL (Table 8). The high correlations between the activity-oriented scales such as near and distance vision suggest a vision-targeted physical dimension. The low correlations between the expectations for future visual functioning and the ocular pain questions with any of the other scales supports that these scales represent unique additional dimensions of vision-targeted HRQOL.
The focus group method used during the first phase of the NEI-VFQ development was selected to maximize the content validity of the questionnaire for patients across a wide range of chronic diseases.19 In this report, we provide additional evidence of reliability and validity for the NEI-VFQ when used among persons with 1 of 5 chronic conditions or low vision from any cause. The strength of the intraclass correlations and the predominantly nonsignificant differences in mean scores when the NEI-VFQ is administered at 2 time points indicate that NEI-VFQ provides an estimate of vision-targeted HRQOL that has high test-retest reproducibility. This characteristic is critical if the NEI-VFQ is to be useful in studies designed to measure change in vision-targeted HRQOL over time.
The evidence for validity of the questionnaire includes comparisons of between-group differences in NEI-VFQ scores for persons with different eye diseases of varying severity, moderate to high correlations with other measures that are designed to measure similar constructs, and evidence of significant correlations between 12 of the 13 NEI-VFQ scales and visual acuity across conditions. Previous publications with larger numbers of persons with glaucoma have reported moderate to high correlations between loss of visual field and poorer dimension-specific NEI-VFQ scores.17,18 Specifically, this article demonstrates that the NEI-VFQ is sensitive to the influence of age-related cataract, ARMD, glaucomatous field loss, and low vision from any cause. These correlations with clinical markers of disease severity support the contention that the NEI-VFQ is a clinically valid measure. Whether the NEI-VFQ will be sensitive to other condition-specific markers of disease severity, such as the severity of retinopathy among persons with diabetic retinopathy, will require additional analyses of these and other data. However, across all conditions, participants report difficulty with at least some of the specific aspects of vision-targeted HRQOL as captured by the NEI-VFQ.
The literature indicates that reports of visual functioning, generic HRQOL such as the SF-36, and clinical measures of visual function such as Snellen visual acuity, reflect different dimensions of vision.5,11,14,15 The previously observed correlations between visual acuity and visual functioning with measures such as the VF-14, for example, are on the order of 0.3 and 0.4. In this investigation, we observed correlations between NEI-VFQ scale scores and visual acuity that were considerably higher (0.7) for the near and distance vision scales. However, it is important to note that the observed correlations between the majority of the NEI-VFQ scales and our generic measure of HRQOL, the SF-36, were below 0.3. This suggests that the NEI-VFQ and SF-36 are capturing different dimensions of health. Because the SF-36 permits comparisons with other populations, the optimal approach to measuring HRQOL in vision research is to use both a vision-targeted measure such as the NEI-VFQ and a generic questionnaire such as the SF-36.
When interpreting these field test results, it is important to consider the following limitations. First, although persons across a large number of conditions and geographic regions were recruited for this study, to minimize the possibility of enrolling persons with false-positive diagnoses, the condition-specific enrollment criteria selected persons with moderate to severe disease. For this reason, we do not know whether the NEI-VFQ will be sensitive to the visual disability that is associated with earlier and milder forms of these conditions. Participants in the study with well-preserved visual acuity and visual fields were at the ceiling of specific scales. This finding suggests that the NEI-VFQ may not detect improvements in vision-targeted HRQOL for persons who start out with very mild impairment. Additionally, this article includes cross-sectional data collected by interview from persons who received care at tertiary centers. Further investigations are needed to establish the responsiveness of the NEI-VFQ in longitudinal studies before it can be fully integrated as an outcome measure in clinical studies. Data to assess test-retest reliability were not collected from persons with cataracts; however, it is likely that these results would be similar among this subgroup. The feasibility of integrating the NEI-VFQ into clinical trials and practice settings will be greatly influenced by the length of the measure. For this reason, research is underway to determine whether a smaller subset of NEI-VFQ items can adequately capture the information contained in the 51-item field test version. In summary, the findings from this psychometric field test provide substantial evidence of the reliability and validity of the NEI-VFQ when used among persons with various common eye conditions and diseases of varying severity.
Accepted for publication July 10, 1998.
This work was supported by contract C950424 from the National Eye Institute, Bethesda, Md, and by the Research Division of Merck Pharmaceuticals, Whitehouse Station, NJ. Dr Mangione is a recipient of Clinical Investigator Award 1K08-AG00605 from the National Institute on Aging, Bethesda, and is an awardee of the Robert Wood Johnson Foundation Generalist Physicians Faculty Scholars Program (029250), Princeton, NJ. Efforts at the Vanderbilt School of Medicine (Nashville, Tenn) were supported in part by an unrestricted grant from Research to Prevent Blindness Inc, New York, NY.
We acknowledge and thank Leon Ellwein, PhD; Frederick Ferris III, MD; Argye Hillis, PhD; Ron Klein, MD; Cynthia Owsley, MD; and Donald Patrick, PhD, for their methodological guidance and participation on the Scientific Advisory Panel for the NEI-VFQ Psychometric Field Test.
Department of Ophthalmology, Georgetown University School of Medicine, Washington, DC: Jonathan C. Javitt, MD, Fang Wang, MD, PhD. Henry Ford Health Systems, Detroit, Mich: Julian Nussbaum, MD, Rhett M. Schiffman, MD, MS. National Eye Institute, National Institutes of Health, Bethesda, Md: Emily Chew, MD. Scheie Eye Institute, University of Pennsylvania, Philadelphia: Janet DeBerry Steinberg, OD. Department of Ophthalmology and Vision Sciences, Vanderbilt School of Medicine, Nashville, Tenn: Amy Chomsky, MD, Denis O'Day, MD. Department of Ophthalmology, Washington University Medical School, St Louis, Mo: Mae Gordon, PhD, Anthony Lubniewsky, MD.
Reprints: Carol M. Mangione, MD, MSPH, UCLA Department of Medicine, Division of GIM and HSR, 10833 Le Conte Ave, Los Angeles, CA 90095-1736.