Ekici F, Loh R, Waisbourd M, Sun Y, Martinez P, Nayak N, Wizov SS, Hegarty S, Hark LA, Spaeth GL. Relationships Between Measures of the Ability to Perform Vision-Related Activities, Vision-Related Quality of Life, and Clinical Findings in Patients With Glaucoma. JAMA Ophthalmol. 2015;133(12):1377-1385. doi:10.1001/jamaophthalmol.2015.3426
To our knowledge, few studies have combined an objective measure of vision-related performance (VRP) and subjective measures of vision-related quality of life (VRQoL) with clinically related visual parameters, particularly in a large, prospective, cohort study setting.
To examine the relationships between clinical visual assessments and both a VRP and 2 self-reported VRQoL measurements.
Design, Setting, and Participants
Patients (N = 161) with moderate-stage glaucoma recruited from the Glaucoma Service at Wills Eye Hospital, Philadelphia, Pennsylvania, were enrolled from May 2012 to May 2014 in an ongoing prospective, 4-year longitudinal observational study. This report includes cross-sectional results from the baseline visit. Patients received a complete ocular examination, automated visual field (VF) test and Cirrus optical coherence tomographic scan. Contrast sensitivity was measured with the Pelli-Robson and the Spaeth-Richman Contrast Sensitivity (SPARCS) tests. Vision-related performance was assessed by the Compressed Assessment of Ability Related to Vision (CAARV) test. Vision-related QoL was assessed by the National Eye Institute Visual Function Questionnaire 25 (NEI-VFQ-25) and a modified Glaucoma Symptom Scale (MGSS).
Main Outcomes and Measures
Correlations between clinical measures and CAARV, NEI-VFQ-25, and MGSS scores.
A total of 161 patients were enrolled in the study. The strongest correlation was found between SPARCS score in the better eye and total CAARV score (r = 0.398; 95% CI, 0.235-0.537; P < .001). The CAARV score also correlated with the Pelli-Robson score (r = 0.353; 95% CI, 0.186-0.499; P = .001), VF mean deviation (r = 0.366; 95% CI, 0.200-0.510; P < .001), and VA (r = −0.326, 95% CI = −0.476 to −0.157; P = .003) in the better eye. There were more statistically significant correlations between contrast sensitivity tests and VF mean deviation with VRQoL measurements than with other clinical measures (visual acuity, intraocular pressure, Disc Damage Likelihood Scale, and mean retinal nerve fiber layer thickness). The MGSS scores were lower (worse) in women compared with men (P = .03 for binocular, P = .01 for better eye, and P = .05 for the worse eye). Structural measures (eg, Disc Damage Likelihood Scale, and retinal nerve fiber layer thickness) were generally not informative with respect to VRP or VRQoL.
Conclusions and Relevance
Contrast sensitivity tests and VF mean deviation were associated with both objective measures of the ability to act and subjective measurements of VRQoL. The strongest correlation was between SPARCS score (contrast sensitivity) in the better eye and total CAARV score. Therefore, measurement of contrast sensitivity should be considered when evaluating patients’ VRQoL. The results of this study were limited by the patient population and apply only within the bounds of the tested cohort.
Glaucoma is a leading cause of irreversible blindness worldwide. In 2010, about 60.5 million individuals in the world had glaucoma; prevalence is expected to increase to 79.6 million by 2020.1 Although most affected patients do not report any specific symptoms or vision loss initially, glaucoma can adversely affect patients’ quality of life (QoL) and ability to perform visually related activities, even when they are unaware of their diagnosis.2- 5
The impact of glaucoma on individuals can be assessed in 3 different ways: (1) clinical measures (eg, visual acuity [VA], contrast sensitivity, and visual field [VF]), (2) self-reported measurements of subjective well-being and QoL, and (3) performance-based assessments of daily activities.6,7 Self-reported questionnaires, both specific to glaucoma and general visual function, are widely used to assess how vision ability affects QoL.8,9 However, to eliminate patients’ personal, emotional, and psychological differences, objective performance-based tools are also needed.10 Performance-based measures have the advantage of allowing for an objective assessment of patients’ abilities using standardized criteria.11,12 Previous studies have used performance-based measures in a clinical setting for assessing patients’ ability to perform activities of daily living.12- 15 Richman et al10 examined the relationship between clinical measures of vision, health-related QoL (HRQoL), and the ability to perform vision-related activities. The performance-based Assessment of Disability Related to Vision (ADREV) correlated more strongly with clinical measures than the self-reported National Eye Institute Visual Function Questionnaire 25 (NEI-VFQ-25).10 Although each approach provides a unique and important perspective in understanding the lives of patients with glaucoma, research that integrates all 3 methods is limited.10,16
Previous studies have investigated the associations between visual function and both vision-related QoL (VRQoL) and general HRQoL.5,17- 19 Vision-related QoL is a person’s satisfaction with his or her visual function and how visual ability impacts his or her life.20 Past research has focused primarily on clinical measurements of VF and VA. Visual deficits, as determined by these factors, have been found to negatively influence patients’ self-reported VRQoL.2,19,21- 23 Previous studies have found that binocular VF loss is associated with lower self-reported VRQoL scores than monocular loss.16,17 In people aged 65 years or older, the severity of self-reported visual impairment strongly correlates with poor HRQoL.24 Other studies examining the relationship between clinical visual measurements and performance-based assessments have found that binocular VA scores, binocular VF scores, and VA in the better eye are strongly associated with ADREV scores.12,14,16
A clinician’s ultimate goal is to address patients’ concerns, which usually are centered on improving or at least maintaining their QoL and their ability to function. With the help of both objective and subjective measures of VRQoL, clinicians can obtain a good understanding of an individual’s activity limitations and psychological changes and assess the impact of glaucoma on patients’ QoL. Meanwhile, assessment of the relationship between VRQoL and visual clinical characteristics can help link subjective patient appraisals of their QoL with objective assessments of vision function and glaucoma severity.7,12,25 This ongoing prospective cohort study investigates the relationship between the visual clinical characteristics of patients with moderate glaucoma and their self-reported QoL and vision-related performance (VRP) ability. The current study described these relationships at baseline. The ultimate goal was to study these relationships over 4 years, making it one of the largest studies, to our knowledge, to include extended follow-up to assess QoL and VRP in patients with glaucoma.
We aimed to explore the relationships between clinical visual assessments, vision-related performances, and subjective vision-related quality of life.
The strongest correlation was found between contrast sensitivity and an objective measure of the ability to perform vision-related activities.
Contrast sensitivity and visual field were associated with both objective measures of ability to act and subjective measurements of vision-related quality of life.
This present report obtained cross-sectional data from the baseline visit of an ongoing and ultimately prospective, longitudinal, observational cohort being conducted at Wills Eye Hospital in Philadelphia, Pennsylvania, in accordance with the Declaration of Helsinki and with approval from the institutional review board at Wills Eye Hospital. Informed consent was obtained from all participants. Full details of the study design and methods have been described elsewhere.26
There were 161 study participants recruited from the Glaucoma Service at Wills Eye Hospital, one of the largest tertiary glaucoma referral centers in the United States, from May 2012 to May 2014. Inclusion criteria included a 2-year minimum diagnosis of open-angle glaucoma, chronic angle-closure glaucoma, glaucomatous optic neuropathy with a Disc Damage Likelihood Scale (DDLS) of stage 5 through 8 in at least 1 eye and characteristic VF loss, age between 21 and 85 years, and the ability to understand and speak English. Patients with DDLS stages 1 through 4 often do not exhibit VF loss due to glaucoma, while patients with DDLS stages 9 and 10 have extensive disc and field damage, which makes detection of optic rim deterioration difficult. Therefore, DDLS stages 1 through 4 and 9 and 10 have been excluded from this study. Other exclusion criteria included neurological and musculoskeletal diseases that would influence performances, cognitive impairment, low availability for annual ocular examinations, incisional eye surgery within the past 3 months, laser therapy within the previous month, causes for visual impairment other than glaucoma, or any medical condition that would preclude the patient from providing reliable and valid data.
After examining NextGen electronic medical health records for eligibility criteria, patients were recruited through telephone calls or contacted in the clinic. Demographic information was collected, and baseline measurements were obtained at the time of enrollment by the Wills Eye Glaucoma Research Center.
Patients’ current symptoms, health problems, medications, and ocular comorbidities were recorded during their visit. The complete ocular examination consisted of the Early Treatment Diabetic Retinopathy Study best-corrected VA (BCVA) test, an intraocular pressure (IOP) measurement (Goldmann applanation tonometry), a slitlamp examination of the anterior segment, and evaluation of the ocular fundus. The DDLS was used to evaluate the extent of optic disc damage caused by glaucoma (a 10-point scale).27- 31 Two VFs were obtained for each eye during the baseline visit using a Humphrey 24-2 SITA Standard perimeter (Carl Zeiss Meditec Inc). The individual VF scores for each eye were averaged to calculate a composite baseline VF score. Cirrus HD-optical coherence tomography (Spectral Domain Technology with Optic Disc Cube 200 × 200; Carl Zeiss Meditec Inc) was used to evaluate the retinal nerve fiber layer thickness (RNFLT). The Pelli-Robson and the Spaeth-Richman Contrast Sensitivity (SPARCS, a novel, internet-based test) tests were used to measure contrast sensitivity. The SPARCS test features multiple answer choices and a bracketing technique to determine the contrast threshold of patients’ central and peripheral vision. The results of SPARCS testing have been shown to be highly reproducible and reliable.32
Objective and subjective methods were used to assess 3 measures. The Compressed Assessment of Ability Related to Vision (CAARV) test is an instrument that correlates strongly with the more extensive ADREV task test, while only requiring between 10 and 15 minutes to complete.25,26 The 4 CAARV items include (1) computerized motion detection, (2) recognizing facial expressions, (3) reading street signs, and (4) finding objects in a room. Participants completed CAARV subtests with both eyes open and used their own appropriate refractive correction to simulate normal function. Participants completed 2 subtests in ambient lighting and 2 subtests in a dark room. Each item was scored from 0 to 7, with 7 being perfect performance. The cumulative CAARV score was calculated as an aggregate of the scores from the 4 subtests.
The NEI-VFQ-25 includes a series of 42 questions (25 primary questions) pertaining to vision or feelings about a visual condition. Answers are selected among a numbered list of possible responses. Subscales include general health, general vision, near vision, distance vision, driving, peripheral vision, color vision, ocular pain, role difficulties, dependency, social functioning, and mental health. Composite scores range from 0 to 100, with higher scores indicating better visual functioning.33
The Glaucoma Symptom Scale (GSS) includes 10 ocular symptoms (eg, burning/smarting/stinging, tearing, dryness, itching, and hard to see in dark places) often experienced by patients with glaucoma. This study implemented a modified GSS (MGSS) questionnaire involving an initial 4-level score (1 = very bothersome; 2 = somewhat bothersome, 3 = a little bothersome, and 4 = no/not at all bothersome). This initial score was then converted to a 0 to 100 scale where zero signifies the presence of a very bothersome problem and 100 signifies the absence of a problem. The final MGSS score was computed as an unweighted average of the responses to all 10 symptoms, averaged between the 2 eyes.34
Demographics, clinical information, VRQoL scores, and MGSS values were summarized using means, medians, and ranges or as frequencies and percentages when appropriate. For continuous variables, groups were compared using the nonparametric Kruskal-Wallis test owing to skewed data. Comparisons were made for VRQoL measurements between sexes, age groups, diagnoses, and race/ethnicity, as well as for clinical measurements between sexes. Spearman correlation coefficients were calculated for all pairs of VRQoL outcome variables and clinical characteristics from each participant’s baseline visit. For analysis of measures within individual eyes, better eye vs worse eye was determined based on the mean deviation (MD) of the VF.
Owing to the large number of clinical measures, P values for the set of comparisons within each table were adjusted using the false discovery rate where noted to control for multiple comparisons.35 Outcome variables with P ≤ .05, 2-sided, were considered to be statistically significant. Data were analyzed using SAS system version 9.4 (SAS Institute).
This study examined 322 eyes of 161 enrolled participants (75 men and 86 women). Baseline demographics and clinical characteristics of the study participants appear in Table 1 and Table 2. When examining VRQoL scores, significant differences between sexes were found for the binocular MGSS (P = .03), MGSS in the better eye (P = .01), and MGSS in the worse eye (P = .05), as well as the driving subscore of the NEI-VFQ-25 (P = .04), with men reporting better VRQoL (higher score) compared with women for all of these measures.
The total CAARV score significantly correlated with SPARCS score (r = 0.398; 95% CI, 0.235-0.537; P < .001), Pelli-Robson score (r = 0.353; 95% CI, 0.186-0.499; P = .001), MD (r = 0.366; 95% CI, 0.200-0.510; P < .001), and VA (r = −0.326; 95% CI, −0.476 to −0.157; P = .003) in the better eye (Table 3). All CAARV subtests, except for reading street signs, were statistically correlated with SPARCS score, Pelli-Robson score, and MD. The CAARV motion detection subtest was the only score to correlate significantly with RNFLT in the better eye (Table 3). The CAARV scores were not found to be correlated with IOP (r = −0.023; 95% CI, −0.200 to 0.156; P > .99). The total CAARV and all CAARV subtests, except for reading street signs, correlated significantly with SPARCS score in the worse eye (Table 4).
The NEI-VFQ-25 composite scores were statistically correlated with MD in both the better (r = 0.289; 95% CI, 0.116-0.443; P = .008) and worse eye (r = 0.348; 95% CI, 0.180-0.494; P = .001), as well as with VA, Pelli-Robson score, and SPARCS score in the worse eye (Table 3 and Table 4). In addition to the NEI-VFQ-25 composite score, 9 of the 12 NEI-VFQ-25 subscales significantly correlated with the SPARCS score and MD in the worse eye, and 8 of the subscales were statistically associated with the Pelli-Robson score in the worse eye (Table 4). The NEI-VFQ-25 general vision and peripheral vision subscales correlated significantly with the DDLS score in the worse eye. The DDLS scores in the better eye did not correlate significantly with any VFQ measures. Additionally, NEI-VFQ-25 scores did not correlate with IOP or RNFLT measurements of either eye. There were no statistically significant correlations between MGSS score and any clinical measure.
The ultimate goal of treating patients with glaucoma is to preserve or even improve their VRQoL and their ability to perform visually related tasks. Clinical tests performed in the office are often useful in estimating the severity of the disease, but provide limited information about VRQoL or VRP.
In the current study, several clinical measures of vision showed statistically significant correlations with both an objective test of VRP and subjective tests assessing VRQoL. The CAARV, an objective test used to determine aspects of vision that influence the activities of daily living of patients with glaucoma, was correlated with contrast sensitivity for both the better and worse eyes. We used 2 distinct tests to measure contrast sensitivity: the Pelli-Robson chart and the SPARCS, a novel computer-based test. While both tests significantly correlated with the CAARV total score for both the better and worse eyes, the SPARCS had stronger correlations for the better eye subscores. Measuring contrast sensitivity by the SPARCS test may provide useful information on the ability of patients to perform daily activities. These findings are in line with previous studies, which highlighted the impact of contrast sensitivity on the VRQoL and VRP of patients with glaucoma.14 The correlations between clinical measures and the SPARCS test, which separately tests 5 individual areas in the VF, were greater than correlations of clinical measures with the Pelli-Robson test, which only assesses the central field. The only positive correlation between RNFLT and any of the subsets of CAARV, NEI-VFQ-25, or MGSS was with motion detection in CAARV.
Measures of contrast sensitivity and VF MD had more significant correlations with CAARV scores compared with other clinical measures including VA, IOP, and structural measures (DDLS and mean RNFLT) in the better eye. It is possible that even patients with relatively good VA may have low CAARV scores owing to profound VF defect or loss of contrast sensitivity.
Subjective VRQoL measures, as assessed with the NEI-VFQ-25 questionnaire, showed significant correlations with contrast sensitivity and VF MD compared with other clinical visual measures, especially in the worse eye. These findings differ from those of previous studies, which typically found stronger correlations between VRQoL scores and measurements in the better eye.2,16,23 This disparity in findings may be attributed, at least partially, to individual bias. Patients may be more aware or bothered by the loss of vision in their worse eye, which could result in higher correlations between worse eye measurements and subjective (self-report) VRQoL. On the other hand, the better eye may compensate for some degree of vision loss in the worse eye, thereby being the ultimate factor in affecting function in daily activities (objective VRP). However, other research has also suggested that the worse eye may have a stronger influence on VRQoL than has been previously assumed, especially if peripheral vision is the primary impairment (as opposed to central vision).36,37 In addition, patients may also have different VF defect patterns between the worse and better eyes. These variations in VF defect patterns can have largely different effects on visual abilities and VRQoL.22,36,38
It is difficult to determine whether objective or subjective measures can be accurately applied to a particular individual. Previous studies have shown good reproducibility and stability over time of the CAARV test.25 However, long-term studies have not been reported with any of the performance-based or QoL-based measures. Patients may have low scores when performing objective tasks but may also think and report that their VRQoL is relatively good and vice versa. We believe the clinician should take both types of measures into consideration. One of the strengths of this study was that it evaluated the relationship between clinical measures and both objective and subjective tests, which both provide valuable information regarding patients’ VRQoL and their visually related ability to function.
Women had significantly worse (lower) MGSS scores compared with men. The MGSS score included a number of symptoms that can be attributed to dry eye and blepharitis (eg, blurry/dim vision, itching, feeling of something in the eye, and dryness), and these results are in line with others who showed higher rates of dry eye and blepharitis in women.39- 41 Surprisingly, the clinical measures used in this study were not significantly correlated with any of the MGSS scores. This is in contrast to other studies, which reported significant correlations between the severity of glaucoma on VF MD and the degree of dry eye symptoms.42 Because this study had a homogeneous population of patients with moderate-stage glaucoma, differences in the severity of dry eye syndrome may be smaller and more difficult to detect and, therefore, the correlations between clinical measures and MGSS were less significant.
Finally, although many correlations between clinical measures and VRQoL tests were significant, they were not strongly correlated (Spearman correlation coefficient were typically between 0.2-0.4). This result may be owing to the inevitable variability that characterizes individuals even within a relatively homogeneous population, as different individuals consider their vision and its impact on activities of daily life differently. It is also plausible, even likely, that the precision of CAARV, NEI-VFQ-25, and MGSS is limited.
There were several limitations of this current report. First, the data reported here are cross-sectional in nature. Second, this study focused only on patients with moderate glaucoma, and the population was primarily white and African American (Table 1). Therefore, any associations and results from this study apply only to the cohort population examined and may not correlate with advanced glaucoma or other demographics. Future directions of this ongoing study are to evaluate prospectively a VRP-based measure and 2 VRQoL assessments in a cohort of patients with glaucoma over a 4-year period to search for factors that may influence the progression of clinical measures of glaucoma and to establish how clinical measures longitudinally relate to patients’ VRQoL and VRP.
Although MD and VA affect VRQoL, the present study suggests that contrast sensitivity may also play an important role in its effects on the daily functioning and activities of patients with glaucoma. Therefore, measurement of contrast sensitivity, particularly using the SPARCS test, should be considered when investigating patients’ VRQoL. In addition, owing to the influence of worse eye VF and contrast sensitivity, clinicians should also aim to improve visual abilities in the worse eye rather than focusing all efforts solely on the better eye.
Corresponding Author: George L. Spaeth, MD, Glaucoma Research Center, Wills Eye Hospital, 840 Walnut St, Ste 1110, Philadelphia, PA 19107 (email@example.com).
Submitted for Publication: April 7, 2015; final revision received July 21, 2015; accepted July 30, 2015.
Published Online: October 1, 2015. doi:10.1001/jamaophthalmol.2015.3426.
Author Contributions: Drs Ekici and Spaeth had full access to all of 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: Hark, Spaeth.
Acquisition, analysis, or interpretation of data: Ekici, Loh, Waisbourd, Sun, Martinez, Nayak, Wizov, Hegarty.
Drafting of the manuscript: Ekici, Loh, Waisbourd.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hegarty.
Obtained funding: Hark, Spaeth.
Administrative, technical, or material support: Ekici, Loh, Waisbourd, Sun, Nayak, Wizov, Hark, Spaeth.
Study supervision: Waisbourd, Hark, Spaeth.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Martinez reported receiving support from Merck, Sharp & Dohme Corp and the Glaucoma Service Foundation. Dr Wizov reported receiving a grant from Merck, Sharp & Dohme Corp. Dr Spaeth is the holder of a patent on the Spaeth/Richman Contrast Sensitivity Test. No other disclosures were reported.
Funding/Support: This study was supported by funding from Merck, Sharp & Dohme Corp and the Glaucoma Service Foundation.
Role of the Funder/Sponsor: The funders 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 decision to submit the manuscript for publication.
Additional Contributions: We thank Benjamin Leiby, PhD (Division of Biostatistics, Department of Pharmacology and Experimental Therapeutics, Sidney Kimmel Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania), for performing the statistical analysis for this study and Elif Erdem, MD (Glaucoma Research Center, Wills Eye Hospital, Philadelphia, Pennsylvania), for assisting with patient recruitment and data collection. Dr Leiby received compensation from a funding sponsor for his contribution.