Scatterplot of the relationship between binocular visual acuity and total Assessment of Disability Related to Vision (ADREV) score (r = −0.79; P < .001).
Scatterplot of the relationship between binocular Pelli-Robson contrast sensitivity and total Assessment of Disability Related to Vision (ADREV) score (r = 0.80; P < .001).
Scatterplot of the relationship between integrated visual field and total 25-item National Eye Institute Visual Function Questionnaire (NEI–VFQ-25) score (r = −0.53; P < .001).
Richman J, Lorenzana LL, Lankaranian D, Dugar J, Mayer J, Wizov SS, Spaeth GL. Importance of Visual Acuity and Contrast Sensitivity in Patients With Glaucoma. Arch Ophthalmol. 2010;128(12):1576-1582. doi:10.1001/archophthalmol.2010.275
To determine which aspects of vision most influence the ability of patients with glaucoma to function.
A total of 192 patients with a full range of glaucomatous visual loss were selected from the Glaucoma Service of Wills Eye Institute. Patients were evaluated clinically with standard visual assessments: visual acuity, contrast sensitivity, visual field, stereopsis, the Disc Damage Likelihood Scale, and intraocular pressure. Patients were evaluated objectively using a comprehensive performance-based measure of visual function, the Assessment of Disability Related to Vision (ADREV), and subjectively with the 25-item National Eye Institute Visual Function Questionnaire. Statistical analyses, including Spearman correlation coefficients and regression analysis, were performed on the data.
Performance on the ADREV was most strongly associated with binocular visual acuity (r = −0.79; P < .001) and binocular contrast sensitivity (r = 0.80; P < .001). Monocular and binocular visual field test results correlated well with the ability to perform the ADREV tasks, but there was a significantly weaker association (P < .05) compared with visual acuity and contrast sensitivity.
The aspects of visual function that best predict the ability of a patient with glaucoma to perform activities of daily living are binocular visual acuity and contrast sensitivity.
The diagnosis and treatment of open-angle glaucoma is based primarily on 4 considerations: the anterior chamber angle, intraocular pressure, visual field, and optic disc. Although intraocular pressure is a critically important consideration for diagnosis, risk assessment, and treatment, evaluation of the optic disc and visual field also provides essential information.1- 4 The visual field has been studied extensively for diagnosis and treatment, as has the optic disc.5- 7
What is not well known is which aspects of vision influence patients with glaucoma most during day-to-day life. Many previous studies have examined which aspects of vision are most highly correlated with patients' subjective beliefs about their vision.8- 19 However, to date, less attention has been placed on how visual field test results and other standard measures of visual function correlate with objective visual performance in tasks that are important to patients with glaucoma in their daily lives.20- 28 We evaluated patients with glaucoma using a well-established, vision-specific quality-of-life questionnaire, the 25-item National Eye Institute Visual Function Questionnaire (NEI–VFQ-25), and with a third-generation, performance-based test entitled the Assessment of Disability Related to Vision (ADREV) to learn which aspects of vision most influence patients with glaucoma.
We selected 200 patients with glaucoma from the Glaucoma Service of Wills Eye Institute of Jefferson Medical College between March 1, 2006, and December 17, 2006. Glaucoma was considered present when the patient had characteristic optic nerve damage and visual field loss.29,30 Intraocular pressure was not considered in the definition of glaucoma. No case was considered to be glaucoma if there was any other reasonable cause for the optic disc or visual field abnormality.
All medical records of patients scheduled for examinations in the Glaucoma Service of Wills Eye Institute were reviewed consecutively. Inclusion criteria included age between 15 and 95 years; understanding and speaking English; being literate; and having primary open-angle glaucoma, primary angle-closure glaucoma, normal-tension glaucoma, exfoliative glaucoma, pigmentary glaucoma, inflammatory glaucoma, traumatic glaucoma, or plateau iris syndrome. Patients were excluded if they had neurological or musculoskeletal problems that would influence their performance, incisional eye surgery within the past 3 months, laser therapy within the previous month, a cataract of grade 2 or higher according to the Lens Opacities Classification System II,31 or any cause for visual reduction other than glaucoma. No patients with macular pathologic features or diabetic retinopathy, for example, were included. Patients with low intraocular pressure were eligible as long as the low pressure did not contribute to their visual loss. To ensure that patients with a full range of glaucomatous damage were included, selection was based partially on the amount of optic disc damage and visual field loss.32- 34
Patient testing was conducted at the Glaucoma Research Center of Wills Eye Institute. Testing was standardized: patients first completed the NEI–VFQ-25, followed by the clinical tests, and finished with the ADREV. The NEI–VFQ-25 was selected because it applies to patients with all visual diseases and is the most widely used and widely applicable.35 If a patient could not self-administer the questionnaire, a clinical research coordinator read the questions aloud and recorded the patient's answers. The questions were scored on a scale of 0 to 100, with 100 corresponding to the best quality of life.
Lighting for all tests was standardized and followed established guidelines when available. Monocular and binocular visual acuity was measured using a rear-illuminated Early Treatment Diabetic Retinopathy Study distance chart, second edition (Precision Vision, LaSalle, Illinois). Visual acuity was scored as the total number of letters identified correctly and converted to logMAR (log10 minimum angle resolvable).36 Binocular contrast sensitivity was measured using the Pelli-Robson contrast sensitivity chart and protocol.37 Visual field testing was performed with the Humphrey Field Analyzer II (Zeiss, Dublin, California) with appropriate refractive correction. The 24-2 Swedish Interactive Threshold Algorithm (SITA) standard program was used for monocular visual field testing. If patients had been tested with this method within the past 6 months, the 24-2 SITA standard program was not repeated. Patients' monocular visual fields were then merged to create an integrated visual field as described by Crabb and Viswanathan.38 Binocular visual fields were also tested with the Esterman program. Depth perception was measured using the Stereo Fly test (Stereo Optical Co, Chicago, Illinois). The Stereo Fly test was scored according to the degrees of arc of the last level correctly chosen and converted into a logarithmic scale. The optic nerves were evaluated using the Disc Damage Likelihood Scale (DDLS) (staging from 1 to 10).32,33 The intraocular pressure was measured using a calibrated Goldmann applanation tonometer (Haag-Streit, Bern, Switzerland).
The ADREV items were performed in the following order: (1) reading in reduced illumination, (2) recognizing facial expression, (3) detecting computer motion, (4) reading signs at a distance, (5) finding large and small objects spread around a room, (6) navigating an obstacle course, (7) putting a stick into holes of different sizes, (8) dialing a telephone simulation, and (9) matching socks. The patients performed the ADREV subtests with both eyes open and with their own appropriate refractive correction to simulate how each patient normally functions. Each item was scored from 0 to 7, with 7 being perfectly performed. The total ADREV score was calculated as an aggregate of the scores of the 9 items.21
The distribution and relationships of all the variables in the study were analyzed in a correlation matrix and by scatterplot. The Mann-Whitney test was used to determine differences among dichotomous variables, such as sex. The Kruskal-Wallis test was used to determine if there were differences in the NEI–VFQ-25 and ADREV scores owing to age, race, type of glaucoma, or number of comorbid conditions. The Spearman rank correlation coefficient was used to determine the relationships within the entire study population between binocular visual acuity, binocular contrast sensitivity, mean deviation (better eye and worse eye), integrated visual field, Esterman binocular visual field, stereopsis, DDLS (both eyes, better eye, and worse eye), and intraocular pressure (eye with higher pressure and that with lower pressure) and the NEI–VFQ-25 and ADREV scores.
The analyses were also conducted in groups of patients based on the severity of visual field loss and optic disc damage. Mean deviation group A included patients with a mean deviation in the better eye greater than −1.70, group B included patients with a mean deviation of −1.70 to −4.37, group C included patients with a mean deviation of −4.38 to −10.00, group D included patients with a mean deviation of −10.01 to −20.00, and group E included patients with a mean deviation in the better eye less than −20.00. The DDLS group A included patients with a DDLS score summated from both eyes between 2 and 5, DDLS group B between 6 and 9, DDLS group C between 10 and 12, DDLS group D between 13 and 16, and DDLS group E between 17 and 20. The mean deviation and DDLS groups were divided with the goal of making the groups equally sized.
Stepwise backward regressions were conducted for the ranks of the total ADREV scores to find the effect of a change in clinical test score on the ADREV score after considering other aspects of vision and demographics. The total ADREV scores were ranked for the regression to minimize the effects of skewed data. Redundant variables were excluded from the regression based on correlations of 0.9 or higher.
All correlations of statistical significance were 2-tailed, and measurements with P ≤ .05 were considered significant. Comparison of correlation coefficients to determine if 2 correlations were significantly different from each other was performed using the z score. Power calculations indicated that 190 patients would be required to detect an effect size of 0.25, at α = .05 and more than 80% power, using a 2-tailed correlation analysis. Internal consistency was determined by means of analysis with the Cronbach α. We used SPSS version 10.1 statistical software (SPSS, Inc, Chicago, Illinois) to perform the statistical analysis.
The institutional review board of Wills Eye Institute approved the study protocol, which was in accordance with the Declaration of Helsinki and Health Insurance Portability and Accountability Act regulations. Eligible patients signed an informed consent document before the study.
We included 192 patients in the study (6 patients had incomplete data and 2 potentially had visual loss not caused by glaucoma). Complete demographics were presented in another article.39 The mean age of the patients was 67.2 years (range, 24-93 years). Both sexes were almost equally represented. Slightly more than half the patients were of European extraction, and more than one-third were African American. Three-quarters had primary open-angle glaucoma. Clinical characteristics are given in Table 1. All severities of glaucomatous visual field and optic disc damage were well represented. There was no difference in the ADREV or NEI–VFQ-25 scores owing to age, sex, type of glaucoma, or number of comorbid medical conditions. White patients had significantly higher scores than African Americans on the ADREV (P < .001) and NEI–VFQ-25 (P = .02), but African Americans in this study had more severe binocular visual field loss (P < .001). For the 9 ADREV items, Cronbach α = 0.915. In another study,21 a Rasch analysis of the ADREV was performed, which showed that all 9 ADREV tasks were of discriminating value.
When considering the entire study population, binocular visual acuity (r = −0.79; P < .001) and binocular contrast sensitivity (r = 0.80; P < .001) had the strongest correlations with the total ADREV score (Figure 1, Figure 2, and Table 2). Binocular visual acuity and contrast sensitivity had significantly higher correlations with the total ADREV score (P < .05) than all of the other clinical tests. When the patients were grouped by severity of glaucoma, binocular visual acuity and contrast sensitivity had the highest correlations of the clinical tests at all 5 levels of severity based on the mean deviation in the better eye and at 4 of the 5 levels of optic disc severity (Table 3 and Table 4). Each ADREV task had its highest correlation with either visual acuity or contrast sensitivity (Table 2).
In the entire study population, the highest correlations between the NEI–VFQ-25 and clinical tests were with the mean deviation in the better eye (r = 0.54; P < .001), mean deviation in the worse eye (r = 0.53; P < .001), integrated visual field (r = −0.53; P < .001; Figure 3), and Esterman field (r = 0.53; P < .001). The visual field test results did not have significantly higher correlations than visual acuity, contrast sensitivity, or stereopsis with the total NEI–VFQ-25 score or any of its subscale scores (Table 5). Mean deviation in the worse eye had a significantly higher correlation (P < .05) than visual acuity with the peripheral vision subscale score of the NEI–VFQ-25. Multiple clinical tests had significantly higher correlations than the DDLS score in both eyes with the NEI–VFQ-25.
Binocular contrast sensitivity, binocular visual acuity, stereopsis, and binocular visual field test results had significant independent influences on ADREV performance in the stepwise backward regression of the ranks of the total ADREV score after taking into account sex, age, race, and number of comorbid medical conditions. For every progression to the next set of triplets on the Pelli-Robson contrast sensitivity chart (a 0.15-point decrease in contrast sensitivity score), the ADREV score decreased by 5.7 points (9.0%). For every 0.1-point increase in logMAR (eg, 20/40 OU to 20/50 OU or 20/80 OU to 20/100 OU), performance on the ADREV decreased by 4.16 points (6.6%). For every 10-point decrease in the integrated visual field score, the ADREV score decreased by 2.88 points (4.6%). For every 0.15-point increase in the logarithm of the stereopsis score on the Stereo Fly test (a mean interval because the score change from animal to animal or group of circles to the next group was not uniform), the ADREV score decreased by 2.83 points (4.5%). When mean deviation in the better eye was substituted for integrated visual field as the visual field component, a 3-dB decrease in mean deviation resulted in a decrease in ADREV performance by 4.71 points (7.5%). When mean deviation in the worse eye was substituted for integrated visual field, a 3-dB decrease in mean deviation resulted in a decrease in ADREV performance by 1.32 points (2.1%).
Understanding how glaucoma affects people in their daily lives is critical when making treatment decisions. Several studies using quality-of-life questionnaires have revealed which aspects of vision most influence patients' beliefs about their vision.8- 10,12- 19 Some correlation between clinical measures of vision and quality of life would be expected. The Los Angeles Latino Eye Study,18,19 for example, showed a general trend in the Latino population of Los Angeles, California, that worse visual fields lead to worse quality of life.
Our study was consistent with many previous studies showing that several clinical measures of vision have significant correlations with the NEI–VFQ-25. When considering the entire study population, the level of visual field damage had a slightly stronger correlation with patients' subjective beliefs about their vision than contrast sensitivity or visual acuity. Nevertheless, all of these relationships have a great deal of scatter (Figure 3), which also seems to be consistent with many previous studies.8,12,16,39,40
Although quality-of-life instruments may be valid for assessing a patient at the individual level, there is a large amount of variation in a study population. This may be partly because people value their vision differently.12,41 What one patient considers good vision may be inadequate for the visual demands of another. The complex causes of scatter in responses to quality-of-life questionnaires can be due to patients' emotions, personalities, psychosocial considerations, and varying abilities to adapt to visual impairments.42- 44
To minimize the effects of subjective questionnaires, objective performance-based tests have gained interest in recent years.20- 28,39 Performance-based instruments test ability directly using standardized criteria and have produced results that are similar to when patients are tested at home.45 Studies from the Salisbury Eye Evaluation Project24,25 have shown that patients with bilateral glaucoma have a significantly slower walking speed, bump into more objects, and have a slower spoken reading speed. Turano et al26 evaluated 47 patients with glaucoma and found that mean deviation had a slightly higher correlation with walking speed than contrast sensitivity or visual acuity. Haymes et al27,28 evaluated a heterogeneous group of 120 patients (57% with macular degeneration and 3% with glaucoma) with the Melbourne Low-Vision ADL (activities of daily living) Index. This index is composed of a 9-item questionnaire and a 16-item performance-based test. In that group of patients, visual acuity and contrast sensitivity had the highest correlations with the overall performance-based score.
The purpose of testing with the ADREV was to determine which aspects of vision most influence the ability of patients with glaucoma to function in their daily lives. The 9 ADREV items were designed to test various aspects of visual function. The visual acuity of these patients and their ability to detect contrast were the most important factors in determining how well they performed the ADREV activities of daily living. This is mirrored by the earlier version of this performance-based test, the Assessment of Function Related to Vision, which also found strong correlations with contrast sensitivity and visual acuity.20 Among patients with glaucoma, contrast sensitivity had either the highest or second highest correlation with all 9 items, and visual acuity had either the highest or second highest with 6 of the 9 items.
It is surprising that the degree of visual field impairment and optic disc damage did not have stronger correlations with ADREV performance. The ADREV was specifically designed to test the activities of daily living considered by patients to be most relevant. Some of the ADREV items primarily evaluated central vision and would be expected to have poorer correlations with patients' visual fields. Nevertheless, the items that involved detecting motion, navigating an obstacle course, and finding objects evaluated peripheral vision, and the mean deviation, integrated visual field, and Esterman field did not have the highest correlation with these ADREV items (Table 2). In addition, the weaker correlations of visual field and optic disc damage with ADREV performance compared with visual acuity and contrast sensitivity in the 5 mean deviation groups and 4 of the 5 DDLS groups imply that visual field loss and optic disc damage are not the best predictors of the ability of patients with glaucoma to perform daily activities during the course of the disease (Tables 3 and 4).
Although measuring visual acuity can provide valuable insight into understanding how well patients with glaucoma are able to function, glaucoma does not cause a decrease in visual acuity until late in the disease course; therefore, monitoring acuity is not a useful way to determine the progress of glaucoma until the terminal stages of the condition. At that point, however, progressive visual loss leading to worsening of visual acuity can be expected to cause disabling effects on the patient.
Changes in contrast sensitivity, on the other hand, occur early and provide highly valuable insight into how well patients with glaucoma are able to function.46,47 In our study, a simple, quick, low-technology method of assessing contrast sensitivity was used: asking patients to read letters of decreasing contrast. The close correlation between contrast sensitivity and the ability to perform activities of daily living suggests that this inexpensive test may be a highly sensitive method of assessing how glaucomatous nerve damage actually affects what people can do.
In this study, contrast sensitivity was measured binocularly to simulate real life. Monocular testing would presumably be more sensitive and more specific and is currently being studied. In patients with glaucoma, assessing monocular contrast sensitivity might be more useful in monitoring the progression of functional visual loss than the more expensive, time-consuming, and difficult evaluation of visual fields.
Correspondence: Jesse Richman, MD, Wills Eye Institute of Jefferson Medical College, 840 Walnut St, Ste 1110, Philadelphia, PA 19107 (email@example.com).
Submitted for Publication: March 5, 2009; final revision received January 19, 2010; accepted January 26, 2010.
Financial Disclosure: None reported.
Funding/Support: This study was supported by an unrestricted grant from Pfizer Inc, the Perelman Research Fund at Wills Eye Institute, the Pearle Vision Foundation, and the Glaucoma Service Foundation to Prevent Blindness.
Role of the Sponsor: Pfizer Inc reviewed the design of the study.
Previous Presentation: This study was presented as a poster at the 77th Annual Meeting of the Association for Research in Vision and Ophthalmology; May 7, 2007; Fort Lauderdale, Florida.
Additional Contributions: We thank Ben Leiby, PhD, of Jefferson Medical College for statistical support.