Progressive vision impairment due to cataract growth can result in significant reduction in patients' quality of life (QOL). Specifically, patients may have difficulty driving, reading, and performing other daily activities that depend on vision. Cataract extraction with implantation of intraocular lens (IOLs) has become a widely used treatment for cataract patients and can result in improvements in patients' self-reported visual function and QOL.1
Because cataract surgery can change patients' QOL, self-assessment of visual function and health-related quality of life (HRQOL) are important outcome measures.2 A longitudinal study of 552 patients who had first-eye cataract surgery used a 14-item questionnaire that assesses visual function (VF-14). Eighty-nine percent of patients showed improvement 4 months after surgery. The change in the patient ratings of vision trouble and satisfaction correlated more with the change in the VF-14 than in visual acuity.3 The VF-14 was also used to demonstrate QOL benefits of cataract surgery in patients with age-related macular degeneration.4
The National Eye Institute Visual Functioning Questionnaire (NEI-VFQ), measuring multiple domains of vision-related functioning and HRQOL, has been validated for use with a wide variety of eye conditions, including cataract.5–7 Trials involving cataract patients have also used generic QOL questionnaires. Improved visual function after cataract surgery was shown to be associated with better HRQOL, as measured by the Medical Outcomes Study 36-item short form (SF-36).8
Quality-of-life measures have also been used in trials evaluating the effect of cataract extraction in the second eye. One study evaluated the effects of the surgery in the second eye compared with the first eye. Surgery in the second eye in patients with bilateral cataract was associated with clinical and statistically significant improvement in functional impairment (measured by the VF-14), satisfaction, and vision problems.9 A study of visual function and acuity revealed that patients who had surgery in both eyes experienced greater improvement in visual function than patients who had surgery in 1 eye only. The findings in this study support the recommendation that patients with bilateral cataract-induced impairment should be treated with surgery in both eyes.10
Laboratory and animal models have shown that short-wavelength blue light causes retinal damage.11–14 Intraocular lenses in use in the United States prior to the AcrySof Natural IOL (Alcon Laboratories) have been clear. Most absorb ultraviolet (UV) light, but none absorbs the blue-light spectrum (400 to 500 nm). However, the healthy human lens progressively yellows with age, filtering some blue light. The AcrySof Natural IOL was developed to approximate the light-filtration properties of the natural, healthy adult crystalline lens with respect to UV and high-energy wavelengths of the blue-light spectrum. Although this blue light–filtering IOL more closely approximates the natural, healthy lens of an adult, it is important to answer the logical question of whether the attenuation of blue light produces poorer outcomes than a clear IOL with respect to the color perception and/or vision of the patient and the ability to function in areas that are particularly vision-sensitive, such as driving.
This article reports on the HRQOL outcomes in a clinical trial in which patients with bilateral cataract were randomly assigned to bilateral implantation with a blue light–filtering IOL or a similar widely used clear IOL that does not filter blue light (AcrySof single piece, Alcon Laboratories). Among the HRQOL domains included in the study were scales measuring patient-perceived difficulties with color vision and driving.
PATIENTS AND MATERIALS
This study used a prospective randomized parallel-groups design involving bilateral cataract extraction and implantation with the new blue light–filtering IOL (AcrySof Natural) or a clear IOL (AcrySof single piece). Both lens models are of single-piece design; however, as noted in its U.S. Food and Drug Administration-approved labeling, the blue light–filtering IOL is designed to filter light in a manner that approximates the human crystalline lens in the 400 to 475 nm blue-light wavelength range. Compared with the clear lens without the chromophore, the AcrySof Natural reduces transmittance of blue-light wavelengths from 71% at 400 nm to 22% at 475 nm. Patients and HRQOL data collectors were masked to treatment; however, clinical investigators were not masked to treatment because the chromophore of the blue light–filtering IOL gives it a visible yellowish tint. The study was approved by institutional review boards at each of the clinical sites involved in the trial, and all patients signed an informed consent form before being allowed to participate. Evaluable patients who had IOL implantation in both eyes were followed for up to 6 months (120 to 180 days) after their second IOL implantation.
Questionnaires and Schedule of Observations
The HRQOL measures were administered 3 times via telephone interviews: a preoperative assessment while patients were afflicted with bilateral cataracts prior to implantation in the first eye (baseline); 30 to 60 days following implantation in the second eye (second assessment); and 120 to 180 days following implantation in the second eye (third assessment). At the completion of the HRQOL study, patients were compensated $10 for each QOL interview they completed, for a maximum of $30.
The HRQOL questionnaires used in this study were a 39-item version of the National Eye Institute's Visual Functioning Questionnaire (NEI VFQ-39) and the 12-item Short Form Health Survey (SF-12).
The NEI VFQ-39 is a vision-specific instrument developed to measure multiple dimensions of vision-related functioning and HRQOL. The VFQ-39 was developed using multicondition focus groups, thereby allowing its use in various eye disorders. The VFQ-39 was selected because of its proven validity and reliability in a large number of eye indications and patient populations.7 Each VFQ-39 scale is scored from 0 to 100, with a higher score indicating better vision-related functioning or HRQOL.
The SF-12 is a generic (not vision-specific) HRQOL instrument measuring 2 components: overall physical and mental health. It was selected because of its proven validity and reliability in a large number of indications and patient populations, including ophthalmic conditions.15 Both the physical and mental component summary scales use a norm-based scoring methodology. In a reference general U.S. population sample, the mean on each scale is 50 and the standard deviation is 10. A score higher than 50 indicates better health than the U.S. mean, and a score below 50 indicates worse health than the U.S. mean.
The scales included in the HRQOL measures used in the study are detailed in Table 1.
Patients and Selection of Study Sample
Patients requiring bilateral extraction of age-related cataracts with implantation of a posterior chamber IOL were enrolled from 6 clinical sites in the U.S. The participating sites performed a high volume of cataract surgeries and had previous experience with clinical trial research. Patients were required to be at least 60 years of age, in good general and ocular health, expected to achieve at least 20/40 postoperative visual acuity, and pass both the Farnsworth D-15 Panel Test and the Ishihara Color Test. Patients with other eye conditions (including color blindness or other color-vision deficiencies) or taking other medications that could interfere with the results were excluded to accurately measure any impact on color vision and other outcomes postoperatively. Also excluded were patients with alcoholism, Alzheimer's disease, or terminal cancer. All enrolled study patients who successfully completed bilateral implantation and completed a baseline HRQOL assessment and at least 1 postoperative assessment were included in the HRQOL analyses.
The study was sized to target a sample with approximately 125 patients per treatment group. This target was chosen to achieve good power (>95%) to detect medium effect sizes (0.5 SD) for the primary endpoints, which included vision-targeted HRQOL. This was based on a 2-sample 2-tailed t test (α = .05).
The study endpoint was defined as the third HRQOL assessment (120 to 180 days after the second IOL implantation) or early termination. The primary analysis consisted of mean change to the third assessment or early termination, using the last observation carried forward approach to impute missing observations; if a score was missing at the third assessment, the score from the second assessment was carried forward, but no scores were carried forward from the baseline assessment. Analyses were also conducted on mean change to the second and third assessments without last observation carried forward to ensure the robustness of the findings.
At baseline, the equivalency of the treatment groups was tested for age, sex, race, baseline visual acuity of the first operative eye, and baseline scores on all HRQOL scales. The Wilcoxon rank sum test was used to compare non-normally distributed continuous variables, analysis of variance was used to compare normally distributed continuous variables, and the Mantel-Haenszel chi-square test was used to compare categorical variables. Independent variables included in the analyses were treatment group, center (study site), and treatment group-by-center interaction.
In each treatment group, 2-tailed paired t tests were used to test for the significance of mean change from baseline to the third assessment (difference from zero) on each of the HRQOL scales. For these tests, the Bonferroni method was used to adjust for multiplicity of testing. To better represent the meaningfulness of change on each HRQOL scale, mean change was converted to a standardized effect size by dividing mean change on each scale by the observed baseline standard deviation of scores on each scale for the pooled study sample.16
The primary treatment comparison analyses focused on the vision-specific VFQ-39 composite score, color vision, and driving scales, and the generic SF-12 physical and mental component summary scales. The analyses tested treatment significance using 2-tailed 2-sample t tests of a treatment group effect on change scores from baseline to the third assessment. Secondary treatment comparison analyses were performed in a similar manner on the remaining subdomains of the VFQ-39. If baseline differences in age, sex, ethnicity, or logMAR visual acuity in the first operative eye were found, treatment comparisons of change on the HRQOL measures were analyzed using general linear models with the covariate included. The Bonferroni method was again used to adjust for multiplicity of testing.
Follow-up analyses of 4 specific VFQ items of special relevance (related to driving in different conditions and walking in dim light) were conducted as well. For these analyses, 2-tailed paired t tests were used to evaluate the significance of change in each treatment group and 2-tailed 2-sample t tests were used to compare the treatment groups.
Of the 291 patients in the starting sample, 257 patients completed both a baseline HRQOL assessment and at least 1 of the 2 assessments following IOL implantation in the second eye and thus were eligible for the HRQOL analyses. Of the 34 ineligible patients, 6 failed to complete any of the 3 HRQOL assessments, 14 did not complete a baseline assessment, and 14 completed only a baseline assessment. Nineteen of the 34 ineligible patients had been implanted with the blue light–filtering IOL, while the other 15 had been implanted with the clear IOL. With the exception of 1 patient in her late 40s, all 257 patients in the HRQOL population were at least 60 years old; and the mean age was 72.2 years ± 6.4 (SD). Ninety-seven percent of the patients were white, and 65% were women. The mean baseline visual acuity in the first operative eye was 20/97 ± 1.9 lines. As shown in Table 2, there were no treatment group differences at baseline with respect to age, ethnicity, or logMAR visual acuity lines in the first operative eye. However, there was a statistically higher percentage of women in the blue light–filtering IOL group than in the clear IOL group.
Baseline VFQ-39 and SF-12 scores are presented in Table 3. There were no statistically significant differences between groups on 13 of the 15 HRQOL scales. The blue light–filtering IOL group had a slightly higher mean score on the VFQ-39 social functioning scale (although both groups had relatively high scores), while the clear IOL group had a slightly higher mean score on the SF-12 physical component summary. If adjustments for the effects of multiple testing had been made, these differences would not have been statistically significant. Numerically, the lowest baseline HRQOL scores were seen in general vision, driving, and mental health. These decrements were reflected in the SF-12 physical component summary, resulting in impairments approximately one-half standard deviation below normative values. The SF-12 mental component summary, however, showed that patients were in good mental status and, on average, reported better mental health than observed in the general U.S. population.
Improvements were observed in both treatment groups from baseline (bilateral cataracts) to the third assessment (120 to 180 days after second-eye implantation) on all vision-specific HRQOL domains (Table 4). A Bonferroni adjustment was used to maintain a family-wise Type I error rate of 0.05 for the full set of 30 t tests (2 treatments × 15 HRQOL domains). It resulted in an adjusted per-comparison significance level of P<.0017 (0.05 of 30). Using this adjusted level of significance, statistically nonsignificant results were observed for the VFQ-39 general health and SF-12 mental component summary scales in both groups and for the SF-12 physical component summary scale in the clear IOL group. Among the 3 generic HRQOL domains, only the mean observed improvement of 3.47 on the SF-12 physical component summary for the blue light–filtering IOL remained statistically significant after adjusting the per-comparison value threshold to 0.0017. Numerically, the largest improvements were seen in the VFQ composite score, general vision, near activities, distance activities, mental health, role difficulties, driving, and peripheral vision scales. These domains all approached or exceeded improvements of 20 points from baseline. The magnitude of these mean changes becomes even clearer after they were converted to standardized effect sizes, as seen in Figure 1.
The primary treatment comparisons were performed for change in the VFQ composite score, color vision, and driving scales and the SF-12 physical and mental component summary scales. For these 5 domains, both pairwise t tests and general linear models with sex included as a covariate indicated no statistically significant differences in outcomes between the treatment groups (Table 4). Secondary treatment comparisons were performed for change in the remaining 10 subdomains of the VFQ-39. For change on 9 out of the other 10 VFQ-39 subdomains, no statistically significant treatment differences were found. The only exception was a marginally statistically significant result for mean change on the VFQ social functioning scale (t test, P<.0420; general linear model with sex included as covariate, P<.0432). However, treatment comparisons for change in VFQ-39 scales other than the VFQ-39 composite, color vision, and driving scales were considered secondary analyses. For these sets of treatment comparisons, applying a Bonferroni adjustment to maintain a familywise Type I error rate of 0.05 resulted in statistical nonsignificance for treatment differences in the VFQ social functioning scale as well.
Although similar improvements were observed for each lens type on the VFQ driving scale, treatment comparisons were performed for the 3 items in the driving scale (Table 5) to ascertain whether there was any difference between the lens types in change in difficulty driving: (1) in the daytime; (2) at night; and (3) in difficult conditions (bad weather, heavy traffic, etc.). In all 3 individual driving items, statistically significant improvements were observed with both lenses (P<.0001 in all cases) and there were no statistically significant differences between lens types (P = .8954 to P = .9834). In addition, since it was believed that the VFQ item asking about “difficulty going down steps, stairs, or curbs in dim light or at night” would be particularly relevant for comparing the lens models, this item was analyzed separately (Table 5). On this item, both lens models produced statistically significant improvement (P<.0001) and the difference between lens models was not statistically significant (P = .2593).
All analyses were repeated for change to the second assessment and change to the third assessment without using the last observation carried forward approach. Results of these analyses agreed overwhelmingly with the primary set of analyses presented here for change to the third assessment using last observation carried forward (data not shown).
The finding that there is no difference between lens models in observed improvement on the VFQ color vision and driving scales shows that the blue light–filtering IOL does not adversely affect patients' color vision and, compared with bilateral cataracts, improves color vision and driving in a manner similar to a widely used clear IOL. These findings are consistent with Farnsworth D-15 panel color perception test results for this study, which show no differences between the 2 IOLs 120 to 180 days after the implantation of the second eye (AcrySof Natural IOL model SB30AL product insert; R.J. Cionni, MD, “Clinical Study Results of the AcrySof Natural IOL,” presented at the ASCRS Symposium on Cataract, IOL and Refractive Surgery, San Francisco, California, USA, April 2003). These findings are also consistent with the update Cionni presented in 2004 (“Color Perception in Patients with UV or Blue Light-Filtering IOLs,” presented at the ASCRS Symposium on Cataract, IOL and Refractive Surgery, San Diego, California, USA, May 2004). There he shared more recent results from another study based on the Farnsworth 100 Hue Test showing no statistically significant differences among patients implanted bilaterally with the AcrySof Natural or the AcrySof single-piece IOL or in patients with an AcrySof Natural in 1 eye and an AcrySof single piece in the fellow eye. This more sensitive color vision test demonstrated that AcrySof Natural IOLs do not negatively impact color perception of patients regardless of whether the patient receives this IOL unilaterally or bilaterally.
For other HRQOL domains as well, the results presented here demonstrate that bilateral implantation with the blue light–filtering or clear IOL produces similar improvements in relation to bilateral cataracts. According to commonly used guidelines, an effect size of 0.8 or greater is considered important, 0.5 is moderate, and 0.2 is small.16,17 Both IOL models produced improvements with large standardized effect sizes (near or above 0.8) on eight of the 12 vision-specific VFQ-39 scales, including the VFQ-39 composite, general vision, near activities, distance activities, mental health, role difficulties, driving, and peripheral vision scales. On the remaining 4 vision-specific VFQ-39 scales (ocular pain, social function, dependency, and color vision), the improvements were at least moderate in magnitude (standardized effect size of 0.5). However, even the moderate effect sizes on 3 of these scales (social function, dependency, and color vision) are impressive when viewed in light of that fact that the relatively high baseline scores on these scales compared with other VFQ scales imposed considerable constraint on the range of possible improvement.
The marginally statistically significant finding that the clear IOL produced a slightly larger improvement on the VFQ social functioning scale might be explained by the fact that although the baseline scores in both groups were fairly high, the mean baseline score in the blue light–filtering IOL group was slightly higher than in the clear IOL group. This left less room at the ceiling of the scale for the blue light–filtering IOL to demonstrate improvement. From between 120 days and 180 days following the second implantation, mean scores on the VFQ social functioning scale for both IOL models were in the high 90s and not significantly different from one another. In addition, as noted in the results section, applying an adjustment for the effects of multiple testing causes this marginally statistically significant difference to become statistically nonsignificant, indicating that it is likely an artifact of multiple testing.
Overall, the HRQOL results presented here confirm the results of previous research of the benefits of cataract surgery, which have demonstrated the largest improvements in vision-related functioning and HRQOL and smaller gains in more distal outcomes such as overall physical and mental health and life satisfaction.8,18 A particularly good example of this effect is the difference in results obtained for the vision-specific VFQ mental health scale compared with the generic SF-12 mental component summary. Large improvements were observed with both IOLs on the vision-specific VFQ mental health scale, which asks about frustration, worry, and irritability due to vision problems, whereas improvements were very small and not statistically significant on the SF-12 mental component summary, which is truly generic and reflects overall mental health and well-being. The lack of significant improvement in the SF-12 mental component summary may be due to the fact that before surgery, patients in this study already had mean mental component summary scores slightly higher than in the general U.S. population. However, although the effect sizes were smaller than for the vision-related scales of the VFQ, the observed improvement (3.47 points, P<.0001) on the SF-12 physical component summary for the blue light filtering IOL exceeded a level considered clinically meaningful on that scale (2.5 points)19–22 while treatment with the clear IOL approached a clinically meaningful improvement (2.36 points, P<.0054). The results also illustrate how a multidimensional measure such as the VFQ provides a fuller picture of the impact of treatment on patients' vision-related functioning and HRQOL than does a 1-dimensional measure of visual function such as the VF-14.
With other AcrySof lenses, the outcome results found in clinical development have generalized well to clinical practice. One might expect the current findings to be observed in the clinic as well.
1. Desai P, Reidy A, Minassian DC, et al. Gains from cataract surgery: visual function and quality of life. Br J Ophthalmol 1996; 80:868-873
2. Lawrence DJ, Brogan C, Benjamin L, et al. Measuring the effectiveness of cataract surgery: the reliability and validity of a visual function outcomes instrument. Br J Ophthalmol 1999; 83:66-70
3. Steinberg EP, Tielsch JM, Schein OD, et al. National study of cataract surgery outcomes; variation in 4-month postoperative outcomes as reflected in multiple outcome measures. Ophthalmology 1994; 101:1131-1140; discussion by DM O'Day, 1140–1141
4. Armbrecht AM, Findlay C, Aspinall PA, et al. Cataract surgery in patients with age-related macular degeneration; one-year outcomes. J Cataract Refract Surg 2003; 29:686-693
5. Mangione CM, Berry S, Spritzer K, et al. Identifying the content area for the 51-item National Eye Institute Vision Function Questionnaire; results from focus groups with visually impaired persons. Arch Ophthalmol 1998; 116:227-233
6. Mangione CM, Lee PP, Gutierrez PR, et al. Development of the 25-item National Eye Institute Visual Function Questionnaire. Arch Ophthalmol 2001; 119:1050-1058
7. Mangione CM, Lee PP, Pitts J, et al. Psychometric properties of the National Eye Institute Visual Function Questionnaire, the NEI-VFQ. Arch Ophthalmol 1998; 116:1496-1504
8. Mangione CM, Phillips RS, Lawrence MG, et al. Improved visual function and attenuation of declines in health-related quality of life after cataract extraction. Arch Ophthalmol 1994; 112:1419-1425
9. Javitt JC, Steinberg EP, Sharkey P, et al. Cataract surgery in one eye or both; a billion dollar per year issue. Ophthalmology 1995; 102:1583-1592; discussion by DM O'Day 1592–1593.
10. Javitt JC, Brenner MH, Curbow B, et al. Outcomes of cataract surgery; improvement in visual acuity and subjective visual function after surgery in the first, second, and both eyes. Arch Ophthalmol 1993; 111:686-691
11. Grimm C, Wenzel A, Williams TP, et al. Rhodopsin-mediated blue-light damage to the rat retina: effect of photoreversal of bleaching. Invest Ophthalmol Vis Sci 2001; 42:497-505
12. Wihlmark U, Wrigstad A, Roberg K, et al. Lipofuscin accumulation in cultured retinal pigment epithelial cells causes enhanced sensitivity to blue light irradiation. Free Radic Biol Med 1997; 22:1229-1234
13. Nilsson SEG, Textorius O, Andersson BE, Swenson B. Does a blue light absorbing IOL material protect the neuro-retina and pigment epithelium better than currently used materials? Lasers Light Ophthalmol 1990; 3:1-10
14. Sparrow JR, Miller AS, Zhou J. Blue light-absorbing intraocular lens and retinal pigment epithelium protection in vitro. J Cataract Refract Surg 2004; 30:873-878
15. Globe DR, Levin S, Chang TS, et al. Validity of the SF-12 quality of life instrument in patients with retinal diseases. Ophthalmology 2002; 109:1793-1798
16. Cohen J. Statistical Power Analysis for the Behavioral Sciences, 2nd ed. Hillsdale, NJ, Lawrence Earlbaum, 1988
17. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care 1989; 27(3 suppl):S178-S189
18. Brenner MH, Curbow B, Legro MW. The proximal-distal continuum of multiple health outcome measures: The case of cataract surgery. Med Care 1995; 33(4 suppl):AS236-AS244
19. Thumboo J, Fong KY, Ng TP, et al. Validation of the MOS SF-36 for quality of life assessment of patients with systemic lupus erythematosus in Singapore. J Rheumatol 1999; 26:97-102
20. Kosinski M, Zhao SZ, Dedhiya S, et al. Determining minimally important changes in generic and disease-specific health-related quality of life questionnaires in clinical trials of rheumatoid arthritis. Arthritis Rheum 2000; 43:1478-1487
21. Kujawski SC, Kosinski M, Martin R, et al. Determining meaningful improvement in SF-36 scale scores for treatment studies of early, active RA [abstract]. Arthritis Rheum 2000; 43:S140
22. Samsa G, Edelman D, Rothman ML, et al. Determining clinically important differences in health status measures: a general approach with illustration to the Health Utilities Index Mark II. Pharmacoeconomics 1999; 15:141-155