Elliott, David B. PhD, FAAO; Patla, Aftab E. PhD; Furniss, Monica MSc; Adkin, Allan MSc
An aging population and ever-increasing improvements in cataract surgery and intra-ocular lens design has resulted in an increasing rate of cataract surgery in the developed world. This increasing demand has led to concerns regarding the associated health care expenditure. About one third of cataract operations are performed on the second eye with an annual cost of approximately one billion dollars in the U.S. 1 and thirty million pounds in the UK. 2 Recently, proposed practice guidelines by a major U.S. utilization review firm and a major insurance carrier suggested eliminating re-imbursement for second eye surgery. 1 In the UK, healthcare purchasers have questioned whether 2nd eye surgery should be rationed 2; since 1998, one UK Health Authority is, at least temporarily, not providing second eye surgery under the National Health Service. The purpose of this investigation was to determine whether there is a need for cataract removal in the second eye.
In 1993, the U.S. Agency for Health Care Policy and Research (AHCPR) guidelines panel on cataract in adults found no research that addressed the benefits associated with modern second eye cataract surgery and intraocular lens implant. 3 The majority of recent literature on vision changes caused by cataract surgery have concentrated on improvements in quality of life and perceived changes in functional vision. 1, 4–12 These reports and U.S. Government guidelines 3 suggest that the decision on when to extract cataract should be based more on the patient’s functional vision difficulties and less on clinical measures such as visual acuity (VA). This reflects a wide acceptance of the importance of quality of life measurements to evaluate all health care interventions. 13, 14 Muldoon et al. 14 suggested that there are two main types of quality-of-life assessment: self-reported functioning and subjective wellbeing. Most recent studies that have assessed improvements attributable to second eye cataract surgery have made assessments of self-reported functioning using questionnaires before and after surgery. 1, 6, 7, 10, 11 Functional vision was assessed in these instruments by asking questions regarding how well patients could see to drive, read, sew, play cards, etc. Two reports found a similar improvement in self-reported functional vision after first and second eye surgery. 6, 7 However, using similar instruments, later studies found a greater improvement after first eye surgery than after second eye surgery. 1, 10, 11, 15 For example, Javitt et al., 1 using the VF-14 questionnaire, 8 found the average improvement attributable to second eye surgery was 32% of that after first eye surgery. Other studies have assessed the improvements in patients’ symptoms and a wide range of clinical measurements such as VA, contrast sensitivity (CS), stereoacuity, and anisometropia for patients undergoing second eye surgery. 15,16
In an earlier study, 11 we found significant improvements in self-reported functional vision [Activities of Daily Vision Scale (ADVS) questionnaire], 7 performance-based measures of functional vision (e.g., reading speed, face perception) and binocular clinical vision tests after second eye surgery. This study expands upon the earlier one in that it uses an important additional clinical test (stereoacuity), performance-based measure (obstacle avoidance), and quality-of-life questionnaire. Applegate et al 4 have previously used physical performance measures to show the impact of cataract surgery, and it has been shown recently that performance tasks completed in the laboratory or clinic under standardized conditions correlate well with similar tasks performed in the home. 17 Other advantages of performance measures over selfreported disability include the fact that change over time can be assessed on a continuous rather than a categorical scale and that reliability and between-subjects comparisons may be better. 17, 18 The obstacle avoidance task was added because of the importance of avoiding falls in elderly people. 20 For example, several studies have shown a relationship between impaired vision and recurrent falls and hip fracture. 20–22 An additional quality-of-life questionnaire [the Self Rating Scale (SRS)23] was used to assess subjective well-being, 14 and the ADVS 7, 19 was used to assess self-reported functional vision.
Cataract subjects were volunteers recruited over a 2.5-year period from four local ophthalmologists who performed almost all the cataract extractions in the Waterloo (Canada) area. Informed consent was obtained, as was approval of the Office of Human Research of the institution, and the tenets of the Declaration of Helsinki were followed. Inclusion criteria were that subjects were scheduled for cataract surgery within the next month and had no signs of comorbid ocular disease or significant neuromuscular, skeletal, or cardiovascular disorders that could interfere with mobility orientation. Second eye surgery subjects were included only if the pseudophakic eye had no significant surgical complications and VA better than 6/9. Cataract patients were informally screened for inclusion by the ophthalmologists and possible subjects were provided information about the study and asked to give written consent for further contact. One hundred patients from a pool of approximately 2500 were sent further information and/or contacted by telephone. Those not contacted were generally older with significant systemic disease (particularly diabetes), various disorders affecting mobility (particularly arthritis), or had comorbid ocular disease. Sixty-six of the 100 who were contacted made appointments for subsequent screening. Of the 66, 19 (29%) were excluded. Of these, seven had surgery rescheduled beyond the duration of the study, four had surgery performed before assessments could be made, three had significant comorbid eye disease, three had significant general health problems that restricted mobility, and two developed (or their partners developed) significant general health problems before their assessment. Four patients did not attend their second visits for various reasons. Pre- and postoperative data from 25 patients (mean age 71.3 ± 9.5) undergoing second eye surgery and 18 (mean age 74.3 ± 6.1) undergoing first eye surgery were obtained. Surgery consisted of phakoemulsification with intraocular lens implantation in all cases. The time between preoperative and postoperative testing averaged 10.8 ± 6.0 weeks for the second eye surgery group and 12.2 ± 5.8 weeks for the first eye group (range, 5 to 27 weeks). In addition, data were obtained from 25 age-matched control subjects (mean age 70.6 ± 4.6) who were tested twice with a mean test-retest time of 13.6 ± 4.7 weeks. Control patients were phakic with no ocular disease or significant neuromuscular, skeletal, or cardiovascular disorders that could interfere with mobility orientation and had visual acuities of 6/9 or better in both eyes. There were no significant differences between the ages of the three groups (F2,65=1.49, p > 0.1) or the times between visits (F2,65=1.53, p > 0.1). These very strict exclusion criteria led to the relatively small sample size, although the number was disappointingly smaller than predicted from pilot studies.
The strict exclusion criteria used ensured recruitment of those second eye surgery patients who were “least likely” to benefit from surgery. Claridge et al. 2 suggested that the only second eye surgeries for which there was a rationale not to provide them through the UK National Health Service were those with no other ocular pathology who had had successful unilateral surgery. They reasoned that there is no reasonable clinical rationale behind excluding surgery from patients with poor vision in the operated eye after first eye surgery or patients with coexisting ocular pathology (such as glaucoma or diabetic retinopathy) that required clear optical media for disease treatment and/or monitoring. Using this rationale, only subjects who had successful first eye surgery and no comorbid eye disease (including lens-induced disease 3) were recruited for the second eye surgery group. This is, therefore, a very conservative study of the possible improvements after second eye surgery. In addition, because of the importance placed on performance measures of mobility orientation, exclusion criteria included any systemic disorders that could interfere with mobility orientation, such as arthritis. This was to ensure that any differences between mobility orientation in the cataract groups and control were caused by vision, and that any pre- to postoperative mobility orientation changes were caused by changes in vision (not changes in arthritis, etc.).
As has been the case in previous studies, 1, 6, 7, 10, 15 a randomized clinical trial was deemed too difficult and expensive to conduct. Because the usefulness of first eye surgery is well established, 3 the study design was a comparison of improvements after surgery for two groups: one undergoing first eye surgery and the other group undergoing second eye surgery. An age-matched healthy group was chosen as the control group so that we could determine whether surgery improved the various facets of vision to normal age-matched values in addition to checking for test-retest effects. Both the pre- and postoperative assessments generally consisted of one visit split between two centers: the first part of the visit was to the School of Optometry (University of Waterloo) and included the patient screening, clinical vision testing, and the performance-based measures of face identity and face expression recognition and reading speed. The second part of the visit was to the Department of Kinesiology and included the performance-based measures of mobility orientation and obstacle avoidance and the completion of the two quality-of-life questionnaires. Subjects were paid traveling expenses and a small honorarium.
Clinical vision testing
Clinical tests were selected on the basis that they had established reliability and validity and that they allowed binocular measurement. 24, 25 The tests and methodology are described in greater detail in an earlier report. 11 Measurements included monocular and binocular high-contrast VA, CS, and disability glare, anisometropia, and stereoacuity. All clinical vision measurements were made using natural pupils and the optimal refractive correction for the test distance, which was determined after a full subjective refraction. The optimal correction was used so that any improvement after surgery reflected the removal of cataract alone rather than the removal of cataract plus any associated uncorrected refractive error. Visual acuity was measured using a Bailey-Lovie logarithm of the minimum angle of resolution (logMAR) chart and by-letter scoring. Contrast sensitivity was measured using a Pelli-Robson chart at a working distance of 3 m, because little change is found in Pelli-Robson CS in cataract subjects when the chart is used at the recommended 1 m. 11 Early cataract preferentially affects CS at the higher spatial frequencies 26; assuming that the most important spatial frequency of letters is about 2 cycles/letter, the change in working distance increased the most important frequency of the letters from about 0.71 c/deg to 2.1 c/deg. Disability glare was measured using the Berkeley glare test 27 and was calculated as the difference in the number of letters read correctly on the low contrast chart with and without the medium setting glare source. 25, 27 Anisometropia was calculated as the difference between the mean spherical refractive correction (sphere plus half the cylinder) of the two eyes. Stereoacuity was tested with the RANDOT circles test, and subjects who failed to see depth in any of the panels were assigned a stereoacuity of 600 sec arc. 25
Performance-based measures of functional vision
Functional vision was assessed using performance-based tests that attempted to simulate real world tasks with which cataract patients have problems. These tasks have been identified in the production of cataract-specific quality-of-life questionnaires 8, 19, 28 and include reading, seeing people’s faces, mobility orientation, and driving. Given the lack of a standardized performance-based driving task and the difficulties in developing one, 29 driving performance was not measured in this study.
Four functional vision tasks were assessed:
Face identity and face expression recognition.
These tests have been used successfully in earlier studies 11, 30, 31 and more detailed information regarding the tests can be found in those reports.
Reading speed and word acuity was measured by having subjects read aloud three Bailey-Lovie word charts at 40 cm using the optimal distance (4 m) refractive correction and +2.25 DS working distance lens. 32 Luminance was 106 cd/m2 and Weber contrast was 69%, which is reduced compared with earlier studies (392 cd/m2)11, 31 to represent more closely the levels found in the home. Reading speed for each print size was calculated in words per minute and was averaged across the three charts. Optimal reading speed was determined as the mean of the two peak speeds. Reading speed for 1 M print (8 point) was also determined (this was taken to represent the size typically encountered in newspapers). The percentage of subjects able to read 0.4 M (3 point) was determined as this is the size of print typically found on medicine bottles.
Face recognition was assessed using the method developed by Bullimore et al. 33 Black and white photographs of four male and four female faces were arranged in a letter chart format. There were five faces per line with each line decreasing in size by 0.15 log units. For each person pictured, there were four different facial expressions; happy, sad, angry, and afraid or surprised, giving a total of 32 photographs. The angular size of the faces was expressed in terms of the equivalent viewing distance (EVD), the distance at which a real face would subtend the same angle that the photograph subtended. Credit was given (0.03 log units) for each correct answer, and two threshold scores were obtained: for correct recognition of identity and for correct recognition of expression. During testing subjects could refer to a panel with large photographs of the eight characters in neutral facial expressions.
Because of the possible restrictions on the field of view caused by trial case lenses, the mobility orientation and obstacle avoidance tasks were performed with the patient’s own spectacles unless these differed significantly from the optimal correction (>1.00 DS or DC in both eyes). This was not the case for any subject.
The detailed locomotor changes required for obstacle avoidance were assessed by asking subjects to walk over obstacles of different heights. Subjects walked a straight path in a dimly lit environment (approximately 1 lux) and stepped over one of two low-contrast foam obstacles (7 and 27 cm high). The path length was adjusted for each subject to ensure that the lead leg over the obstacle was always the right leg. Four infra-red emitting diodes were placed on the right toe, heel, greater trochanter, and left toe and sampled at 100 Hz to monitor limb trajectory. Measures that previous work 30 has indicated reflect important obstacle avoidance strategies were assessed using the OPTOTRAK motion analysis system (Northern Digital, Canada). The percentage of times the obstacle was hit was also recorded. Ten trials for each obstacle height and 10 control trials were randomized. The experimental paradigm has been used previously for young and elderly subjects with and without visual deficits. 30
Mobility orientation performance was assessed by recording the time needed to travel two 16.0-m paths that included a variety of foam obstacles of different size, contrast, and shapes, both on and above the ground. The number of “mistakes” made by the subject, where a mistake was defined as contact with an obstacle, stopping, straying outside the pathway, and avoidance strategies when none were required, was also recorded. The illumination for the two pathways was approximately 1 lux , to simulate twilight, and one of the two paths included appropriately placed glare sources. Participants were given 3 min to adjust to the dim light conditions before the start of the trial and were not allowed to see the actual travel path before testing. They were then instructed to walk as quickly as possible through the pathway, avoiding all obstacles, while remaining within the marked boundary of the travel path. A previous study using a course with illumination in the photopic range had shown no change in performance with simulated cataract and therefore this level of illumination was not used. 31
Self-reported functional vision was assessed using the ADVS instrument developed by Mangione et al. 7, 19 The ADVS determines perceived disability in 20 visual activities, and provides an overall perceived visual disability score and scores in five sub-categories of distance vision, near vision, glare disability, night driving, and daytime driving. The ADVS has been shown to be a reliable and valid measure of self-reported functional vision in patients with cataract. 7, 19
The subjective well-being aspect of quality of life 14 was assessed using the SRS. This is a test-retest version of the Psychosocial Impact of Assistive Devices questionnaire (PIADS; n = 157, Cronbach’s α= 0.96) 23 and was developed especially for the study. It is a 27-question instrument that asks subjects whether they feel competent, happy, adequate, confused, worried, embarrassed, capable, etc. These various psychological variables are scored on a scale from +3 (very much so) to −3 (just the opposite). For this report, a composite score from all questions is reported. The majority of questions scored +3 for a ‘positive’ outlook (e.g., scoring +3 [very much so] for “I feel happy”). In the original questionnaire, some questions scored −3 for a ‘positive’ outlook (e.g., scoring −3 [just the opposite] for “I feel worried”), and the signs of these scores were reversed before averaging. Therefore, the overall range was from −3 (very poor subjective well being) to +3 (excellent subjective well-being). The ADVS and SRS questionnaires were administered in person by one investigator.
The majority of the clinical and functional vision data assumed a normal distribution and parametric statistical analyses were used. Stereopsis and anisometropia results were log transformed to provide a normal distribution. ADVS and SRS data did not assume a normal distribution because of a ceiling effect in that many subjects scored maximum or near maximum scores, so nonparametric statistics were used with these data. As the kinesiology evaluations were completed by a subsample of the subjects, all the nonkinesiology analyses were repeated using this subsample to ensure that any conclusions reached did not depend upon which subjects were included. The analyses with the smaller subsample of subjects produced the same results as the full sample.
Test-retest control data
There was no significant difference between test and retest data from the control group for any of the tests, except for binocular VA (t24 = 3.06, p < 0.05). Given the number of t-tests and Wilcoxon signed rank tests performed, one significant difference would be expected by chance at the 5% level (5% of 24). The standardized effect size for the control group (n = 25) was 0.80 using a power of 80% and a two-tailed alpha of 5%. 34 Using these figures, the test-retest data of the control group would have been able to detect a significant change in monocular high contrast VA, for example, of 0.08 logMAR (0.8 of a line) or more. The mean difference between retest and test results from the control group and the 95% confidence limits of the differences (calculated as 1.96 × S.D.) are shown in Table 1 and give an indication of the repeatability of the tests.
Clinical vision data
The significance of any change from pre- to postoperation were assessed using two-tailed paired t-tests. Given the large numbers of t-tests used (46 for clinical and functional vision data), a Bonferroni adjustment of the p-value would have produced an overly stringent value. 34, 35 Instead, it should be noted that 2.3 (5% of 46) of the improvements from pre- to postsurgery could have been caused by chance at the 5% level. The significance of differences between first eye, second eye, and control group data were assessed using analyses of variance and post hoc Scheffé F-tests.
No significant changes occurred for any measurement in the nonoperated eyes of those undergoing surgery from test to retest, except for VA in the second eye group (p < 0.01). The slight reductions in mean VA and CS in the nonoperated eye of the first eye surgery group were not significant. Pre and postoperative clinical vision data from the first and second eye subject groups are shown in Table 2. Both groups showed a similar improvement in mean VA of about four lines, from approximately 0.50 logMAR (Snellen 6/18) to 0.10 logMAR (6/7.5) and were near age-matched control levels (mean −0.02 logMAR, ∼6/6). Mean CS improved by about 0.50 log units (over three steps on the Pelli-Robson chart), although postoperative values were about one step below the control mean. Some patients had reduced vision scores postoperatively, but no clinically obvious abnormality. One patient had posterior capsular remnants that the surgeon considered not significant to treat and another patient had significant debris in the tear film that could have reduced scores. Some subjects could not see any of the letters with the disability glare test preoperatively in the operated eye so that mean data are from a smaller sample (1st eye, n = 12; 2nd eye, n = 16). Not surprisingly, the subjects who could see the letters under glare conditions and whose data were included were those with the least dense cataracts [their mean preoperative VA from the operated eye was 0.34 ± 0.18 (Snellen 6/12) compared with 0.53 ± 0.33 (Snellen 6/18) for the whole sample], and improvements in disability glare after surgery were slight (see Table 2).
There was much less improvement in binocular clinical scores than those from the operated eye (Table 2). For example, binocular VA improved by about 0.10 logMAR (one line) after surgery compared with the 0.40 logMAR (four lines) improvement in the operated eye. This is because in patients with bilateral cataract, it is usual to operate on the eye with the worst vision first (39 of the 43 patients in this study had the eye with the poorer VA operated on), and binocular improvement relates to the improvement in the best monocular score. There were significant improvements for the first eye surgery group for binocular VA and CS (p < 0.02), but not for stereopsis or anisometropia (p > 0.05). Binocular VA improved by 0.13 logMAR (∼11/3 lines) and binocular CS by 0.18 log units. Log stereoacuity (mean of 2.02 log sec arc, 105 sec arc) and log anisometropia (mean of −0.07 log D, 0.85 D) remained poor after first eye surgery, and two patients (11%) had unrecordable stereoacuity postoperatively.
The second eye surgery group showed significant improvement for binocular VA, CS, stereoacuity and anisometropia (p < 0.003, Table 2). Preoperatively, five of the second eye surgery group (20%) had unrecordable stereoacuity and another four (16%) had the lowest recordable value of 400 sec arc. Postoperatively, all subjects had a stereoacuity of 140 sec arc or better, and the mean stereoacuity improved by 0.38 log units (2.4×) from 2.18 (152 sec arc) to 1.80 (63 sec arc) and was similar to normal levels (Scheffé, p > 0.05). Mean log anisometropia improved after second eye surgery by 0.33 log units (2.1×) from 0.00 (1.00 D) to −0.33 (0.47 D) and was similar to the age-matched control group (Scheffé, p > 0.05).
Performance-based measures of functional vision
Mean (±1 S.D.) data for the performance based measures of functional vision before and after surgery are shown in Table 3. The percentage hits during the obstacle avoidance measure were highly skewed; the majority scored zero, so nonparametric statistics were used. The number of subjects completing the Kinesiology aspects of the study was slightly smaller; some of the patients became too tired to complete this part of the study during either the pre- or postoperative session, and they either did not want to or were unable to return at a later date (a return visit was often impossible because of their imminent cataract surgery). There were improvements in more tasks after first eye surgery than after second eye surgery. There were significant improvements for the first eye subject group in face identity and face expression recognition, newspaper print reading speed, word acuity, the time and number of mistakes in both travel pathways, and the hip and toe velocities over the low and high obstacles (all t-test p-values < 0.003 except hip velocity over the low obstacle, p < 0.03). For example, mean face expression recognition improved from 1.11 log EVD (13 m) to 1.26 log EVD (18 m), so that patients could now recognize facial expressions an average of 0.15 log EVD (1.4×) further away. Optimal reading speed was at control levels preoperatively and did not change after surgery. Legge et al. 36 have previously shown that optimal reading speed is normal in cataract patients provided the text is made large enough. Mean newspaper print (1 M) reading speed improved from 1.23 log wpm (17 wpm) to 1.83 log wpm (68 wpm), so that patients could read newspaper print an average of 0.60 log wpm (about 4×) faster after surgery. The number of hits on the two travel pathways reduced by an average of 2.6 and 5.2 and the time taken was also reduced by 21 and 32%. In addition, subjects moved substantially more slowly when stepping over the high and low obstacles before first eye surgery, but after surgery they moved at near age-matched normal speeds. After first eye surgery, newspaper reading speed and hip and toe velocity over the two obstacles were returned to age-matched control levels (optimal reading speed was already at this level preoperatively). All other functional vision tasks were below control levels postoperatively in the first eye surgery group (Scheffé, p > 0.05).
The improvements in most of the performance-based tests of functional vision were smaller after second eye surgery than after first eye surgery (Table 3). Mean face expression recognition improved from 1.27 log EVD (19 m) to 1.35 log EVD (22 m) after second eye surgery, an improvement of 0.08 log EVD (1.2×). This compares with the improvement after first eye surgery of 0.15 log EVD (1.4×). There was no improvement in newspaper reading speed after second eye surgery, and word acuity improved by 0.15 logMAR (11/2 lines) after first eye surgery and only 0.06 logMAR (approximately 1/2 line) after second eye surgery. The number of patients who could read medicine bottle size print (0.4 M) after surgery increased by 59% after first eye surgery and 15% after second eye surgery. However, there were substantial improvements in mobility orientation after second eye surgery that were similar to the improvements after first eye surgery. The number of mistakes on the two travel pathways reduced by an average of 2.3 and 4.7 (compared with 2.6 and 5.2 after first eye surgery) and the time taken was also reduced by 25 and 44% (compared with 21% and 32% after first eye surgery). Finally, the improvements in toe clearance and maximum toe clearance when stepping over the smaller obstacle only occurred after second eye surgery, and first eye surgery made no difference to toe clearance measures. After second eye surgery, toe clearance levels fell to age-matched normal levels. All performance-based tasks were at control levels after second eye surgery except for face identity and expression recognition (Scheffé, p < 0.05).
It is unlikely that the use of the habitual refractive correction (patient’s own distance spectacles if worn) rather than the optimal refractive correction made the improvement in mobility orientation after surgery substantially greater than would otherwise have been the case. For example, the mean (±1 S.D.) difference between the preoperative optimal and habitual refractive correction for the right eye was −0.41 D ± 0.75 and −0.07 D ± 0.55 for the first and second eye groups, respectively. Postoperatively, these values were −0.19 D ± 0.77 and −0.20 D ± 0.47 for the first and second eye groups respectively. The negative value indicates that the mean optimal refractive correction was more negative than the mean habitual correction and probably reflects the effect of uncorrected nuclear cataract-induced myopia in some of these patients. 37 The pertinent mean value for the control group was +0.29 D ± 0.56 and the positive mean value reflects the typical age-related hyperopic shift which is found when patients with nuclear cataract are excluded.
Quality of life assessments
The significance of any change from pre- to postoperation was assessed using the Wilcoxon signed rank test for ADVS and SRS data. Given the large numbers of Wilcoxon signed rank tests used (22, which includes the assessment of percentage of hits over the obstacles shown in Table 3), a Bonferroni adjustment of the p-value would again have produced an overly stringent value. 34, 35 Instead, it should be noted that 1.1 (5% of 22) of the improvements from pre- to postsurgery could have been caused by chance at the 5% level. The significance of differences between first eye, second eye, and control group data were assessed using the Kruskal-Wallis test. The median and range scores for the overall and five subcategory ADVS scores of distance vision, near vision, glare disability, night driving, and daytime driving are shown in Table 4. The overall median ADVS score improved more after first eye surgery (median improved by 10; z17 = −3.72, p < 0.0003) than after second eye surgery (median improved by 6; z24 = −3.46, p < 0.0006), and there were considerably larger improvements in the median score for most of the subcategories (Table 4). All the subcategory ADVS scores improved after surgery except day driving after second eye surgery (p > 0.06). All ADVS scores were statistically similar to age-matched normal levels postoperatively except for night driving after first eye surgery (p > 0.05). Some subjects scored a maximum 100 score on the ADVS, so the data are truncated. Table 5 shows the percentage of maximum scores (scores of 100) found for the overall and subcategory ADVS scores.
The median and range of the composite SRS scores are shown in Table 6. The improvement after second eye surgery was significant (z = −2.56, p < 0.02) but the improvement after first eye surgery was not (z = −0.95, p > 0.10).The Kruskall-Wallis test indicated that there were significant differences between the three groups preoperatively (H = 14.51, p < 0.0008) that disappeared after surgery (H = 3.84, p > 0.10).
Clinical tests as predictors of functional vision improvement because of surgery
Regression models were used to determine if any of the clinical tests predicted improvement in functional vision and quality-of-life indicators after surgery. Because of the relatively large number of tests compared with the number of subjects, the number of clinical test predictors used in the analyses was kept to a minimum. Only monocular measurements of VA and CS and log stereopsis were used. The other clinical tests were not used in these analyses based on the following rationales: disability glare was excluded because it would have reduced the sample size even further; anisometropia was excluded because it is unlikely to influence functional vision; and binocular measurements were excluded because they are rarely used in a clinical situation. Regression models of clinical test predictors against the change in the various functional vision tasks after surgery (i.e., postoperative score − preoperative score) were examined. Some of the changes in functional vision scores were highly correlated (the four mobility measures and the 10 obstacle avoidance measures) and principal components varimax factor analyses were used to summarize and reduce the data. Two factors were found to be significant from the mobility data: a ‘mistakes’ factor and a ‘time taken’ factor. Four factors emerged from the obstacle avoidance data and could be identified as ‘velocity,’ ‘toe clearance,’ ‘hits over the low obstacle,’ and ‘hits over the high obstacle’ factors. The improvements in VA from the worst and best eyes were forced into the regression model as the first and second steps, respectively. These measurements form the traditional assessment of vision in cataract patients and other tests should only be adopted if they provide additional information about improvement in functional vision. Any additional step that could enter the regression model would then be providing significant extra information beyond VA. The regression analyses are shown in Table 7.
The test-retest control data indicate that there were no significant learning effects, and all retest-test mean values were very close to zero (Table 1). This suggests that any improvements after surgery were caused by removal of the cataract rather than any learning effects (although the test-retest data were from healthy control subjects rather than subjects with cataracts). The 95% confidence limits of retest-test difference data are similar to data published previously 24 and indicate that the tests used were providing repeatable data.
Clinical vision scores
The preoperative mean of 6/18 is indicative of the early nature of the cataracts being extracted. The earlier the cataracts are extracted, the smaller the improvements in all aspects of vision are likely to be. 5 This level of preoperative mean VA is similar to comparable recent studies from the U.S. and one from the UK (6/18 to 6/24) 1, 6, 7, 10, 12 but better than earlier U.S. studies (e.g., 6/304) and others from the UK (e.g., 6/60). 16 The improvements in VA and CS were as expected from previous studies. 11, 15, 16 Binocular VA and CS improved more after first eye surgery than after second eye surgery. However, two aspects of clinical binocular vision, anisometropia and stereoacuity, did not significantly improve after first eye surgery but did improve after second eye surgery (Table 2). Indeed postoperative levels of stereoacuity and anisometropia returned to age-matched normal levels after second eye surgery. Laidlaw and Harrad 16 reported similar improvements in stereoacuity and anisometropia after second eye cataract surgery, which were associated with corresponding improvements in symptoms. The improvements in stereoacuity may be important in improving mobility orientation; these results are discussed later. Anisometropia is common in cataract patients because of myopia caused by nuclear cataract increasing at different rates, 38 and can produce complaints of asthenopia. 38 Correcting the anisometropia can produce further symptoms because of spectacle-induced aniseikonia and anisophoria. 38
Performance-based tests of functional vision
There were significant improvements in the performance-based tests of functional vision after first eye surgery, and these improvements were generally larger than after second eye surgery (Table 3). These results support the improvements in self-reported functional vision after both first and second eye surgery found in this and other studies. 1, 6, 7, 10, 11 They also support the findings of Applegate et al. 4 of significant improvements after cataract surgery in other physical performance-based measures. First eye surgery particularly improves reading speed of relatively small text such as newspapers and medicine bottles and there was relatively little improvement in these tests after second eye surgery. Significant improvement in face and expression recognition, mobility orientation, and obstacle avoidance were also found after first eye surgery, and these results support and extend our earlier findings. 11
The most important performance-based test improvements after second eye surgery seem to be those for mobility orientation, walking speed, and obstacle avoidance (Table 3). The results suggest that both first and second eye surgery may prevent trips and subsequent falls from occurring and increase walking speed. Mobility orientation did not return to age-matched normal levels after first eye surgery, but it did after second eye surgery. The major obstacle avoidance strategies used were reducing speed when moving over an obstacle and increasing toe clearance to avoid tripping. Speeds increased after both first and second eye surgery, but toe clearance levels remained high after first eye surgery and only improved to control levels after second eye surgery (Table 3). The majority of falls are caused by trips, 39 and serious injury cause by falls is a significant problem to aging adults and is a significant economic burden to society. 20–22 The 20 to 32% reduction in travel time should also be considered. The consequences of a long travel time become pertinent when placed in the context of situations such as crossing a road. A slow walking speed (attributable in part to visual impairment) has been shown to make it difficult for older pedestrians to cross the street 40 and elderly people have the highest rate of pedestrian death. 41
There were relatively small improvements in ADVS scores after cataract surgery because scores in several subcategories were already at or near the maximum, particularly in the second eye subject group (Tables 4 and 5). The preoperative ADVS scores of both groups in this study were appreciably higher than those reported by Mangione et al. 7 This reflects the early nature of the cataracts being removed in this study and the exclusion of subjects with coexisting eye disease. The greatest improvements after both first and second eye surgery were for the night-driving subcategory. The improvement in perceived ability to drive at night after second eye surgery is particular noteworthy given that night driving after first eye surgery was the only subcategory not to reach age-matched normal levels postoperatively. A recent study found that 46% of cataract patients report problems in estimating distance while driving and suggested that second eye surgery is particularly helpful in improving these symptoms. 42 The relatively small improvement in perceived near vision after first eye surgery is surprising given the considerable improvements in reading speeds for small text. However, the ADVS near vision score is calculated from the answers from nine questions, only three of which relate to reading small text such as newspapers, medicine bottles, and ingredients on cans. The other questions relate to the ability to see much larger text or objects (e.g., the perceived ability to play cards), which may not be impaired given that optimal reading speed (i.e., reading speed of relatively large text) was not affected by cataracts (Table 3). The ADVS findings are similar to recent reports showing improvements in perceived visual disability after second eye surgery that were smaller than after first eye surgery. 6, 10, 11 Although the improvements in median ADVS score were relatively small, there were large improvements in the percentage of patients who were completely happy with their vision after second eye surgery (i.e., gave an ADVS score of 100, Table 5). For example, only one patient (6%) was perfectly happy with all aspects of their vision after first eye surgery, yet nine patients (36%) were perfectly happy with their vision after second eye surgery.
The improvements in SRS score after surgery were small compared with the large variability in scores and not even statistically significant for the first eye surgery group. Because the SRS is not an unreliable or insensitive instrument (it has shown high internal consistency and was able to show significant differences between young spectacle- and contact lens-wearing groups), it is unlikely that the small improvements are attributable to the instrument. 23 It is more likely to indicate a relatively small effect of cataract surgery, particularly with early age-related cataract, on measures of subjective well-being. Damiano et al. 9 reported that a generic health status measure, the Sickness Impact Profile (SIP), which includes questions regarding subjective well-being, was much less sensitive to preoperative functional vision impairment and change in functional impairment after surgery than a functional vision measure (the VF-14). 8 Brenner et al. 5 showed that vision and quality of life changes (including mental health and life satisfaction measures that obtained information similar to that of the SRS) because of cataract surgery depended on the baseline visual function level. In particular, they found relatively small changes in the mental health measures for those patients with good preoperative vision. 5 Anecdotally speaking, several subjects in our study who demonstrated reduced SRS scores after surgery complained of personal trauma such as divorce, long-term illness, or death of a partner. It seems likely that factors such as these made significant impact on the SRS scores, although no formal assessment of these factors were made.
Clinical tests as predictors of functional vision improvement because of surgery
The large number of tests and small sample size limits these results. However, the results strongly indicate that the poorer VA (typically that from the operated eye) provides a poor assessment of the likelihood of improvement in functional vision attributable to cataract surgery (Table 7); this agrees with previous studies. 1, 6, 7, 9 The better VA provides useful information regarding the improvement after surgery of several functional vision tasks, including face expression recognition, word acuity, reading speed, and the self-reported functional vision problems (ADVS) (r2 values between 0.17 to 0.37;Table 7). Contrast sensitivity seems to provide some useful additional information beyond VA regarding mobility orientation and reading speed, which is in accordance with other studies with low vision patients and simulated cataract. 25, 31, 43, 44 In all these cases, the association between the clinical tests and the improvement in functional vision after surgery is highly statistically significant but to a modest degree. This is similar to findings from previous studies 6, 8, 12 and perhaps reflects the very different nature of clinical tests and functional vision tasks. There was no association between any of the preoperative clinical measures and the change in quality of life (SRS scores) after surgery (Table 7). The only scores to correlate with the change in SRS scores was the improvement after surgery of the ADVS scores (r2 = 0.24, p < 0.002). Similar findings have been reported previously 12 and possibly reflect the self-reported nature of both results. Another interesting finding was the predictive value of stereopsis to the improvement in some of the obstacle avoidance measures (the percentage of hits when stepping over the high obstacle and the height of toe clearance over the low obstacle). A simple linear regression plot of log stereopsis against the improvement in maximum toe clearance over the low obstacle after surgery is shown in Fig. 1. This relationship was similar when the change in log stereopsis after surgery was plotted against the improvement in maximum toe clearance (r2 = 0.17, p < 0.02; with the removal of one major outlier, r2 = 0.30, p < 0.001). Although the association is of a modest degree, other studies have shown an association between stereoacuity and disability from falls among elderly people, 21, 45 and this is worthy of further investigation.
The inclusion criteria limit the generalizability of the present study. The results apply to patients without systemic disease affecting mobility orientation. However, we would expect patients with such disease to gain some improvement in mobility orientation after cataract surgery because of improvements in vision. Certainly, such patients are likely to achieve similar improvements in reading speed, face recognition, and all clinical vision scores as found in this study. Any improvements in quality of life due to surgery may be influenced by their systemic condition. The results from the second eye surgery group are only generalizable to patients without comorbid eye disease who had previously had successful first eye surgery. These patients were intentionally recruited as those least likely to need surgery. 2 Despite the conservative nature of the subject inclusion criteria for the second eye surgery group and the relatively low number of subjects, the study supports the need for second eye cataract surgery. The study has shown that there are many statistically significant improvements in clinical and functional vision after second eye surgery. A question remains of whether these improvements are clinically significant. A reasonable approach to answering this question is to compare the improvements to those after first eye surgery, which is commonly accepted as providing clinically significant improvements. Second eye surgery provides approximately half the improvement of first eye surgery for binocular VA and CS, face expression recognition, word acuity, speed of movement over the obstacles, and overall self-reported functional vision. It provides similar improvement in mobility orientation and substantially more improvement in anisometropia, stereopsis, and some of the obstacle avoidance measures. Although the small sample size limits the power of the study, the results suggest that any elimination or rationing of second eye surgery could leave people with significant visual disability. The link between vision and falls may be particularly important. The majority of falls are caused by trips 39 and serious injury from falls is a significant problem to the aging adult and a significant economic burden. The improvement in obstacle avoidance strategies and reduction in the number of hits around the travel pathways suggests that both first and second eye surgery may prevent trips and subsequent falls from occurring. Walking speed also increases, which may help avoid accidents to elderly pedestrians. 40 Any economic benefit from rationing second eye cataract surgery could therefore be lost to the cost of care for monocular pseudophakic patients who may trip and fall or have a traffic accident because of a slow walking speed. Given the serious implications of falls in elderly people and the availability of optometric and medical interventions to improve vision in this age group, the possible link between binocular vision and falls needs to be investigated further.
The study was supported by a grant from Health Canada, National Health Research and Development Program (Grant #6606–5351-402).
1. Javitt JC, Steinberg EP, Sharkey P, Schein OD, Tielsch JM, Diener M, Legro M, Sommer A. Cataract surgery in one eye or both. A billion dollar per year issue. Ophthalmology 1995; 102: 1583–92.
2. Claridge KG, Francis PJ, Bates AK. Should second eye cataract surgery be rationed? Eye 1995; 9: 47–9.
3. Cataract Management Guideline Panel, United States. Cataract in Adults: Management of Functional Impairment. Rockville (MD): Agency for Health Care Policy and Research, Public Health Service, United States Department of Health and Human Services; AHCPR Publication No. 93-0542, 1993.
4. Applegate WB, Miller ST, Elam JT, Freeman JM, Wood TO, Gettlefinger TC. Impact of cataract surgery with lens implantation on vision and physical function in elderly patients. JAMA 1987; 257: 1064–6.
5. Brenner MH, Curbow B, Javitt JC, Legro MW, Sommer A. Vision change and quality of life in the elderly. Response to cataract surgery and treatment of other chronic ocular conditions. Arch Ophthalmol 1993; 111: 680–5.
6. Javitt JC, Brenner MH, Curbow B, Legro MW, Street DA. 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: 686–91.
7. Mangione CM, Phillips RS, Lawrence MG, Seddon JM, Orav EJ, Goldman L. Improved visual function and attenuation of declines in health-related quality of life after cataract extraction. Arch Ophthalmol 1994; 112: 1419–25.
8. Steinberg EP, Tielsch JM, Schein OD, Javitt JC, Sharkey P, Cassard SD, Legro MW, Diener-West M, Bass EB, Damiano AM, Steinwachs DM, Sommer A. The VF-14. An index of functional impairment in patients with cataract. Arch Ophthalmol 1994; 112: 630–8.
9. Damiano AM, Steinberg EP, Cassard SD, Bass EB, Diener-West M, Legro MW, Tielsch J, Schein OD, Javitt J, Kolb M. Comparison of generic vs. disease-specific measures of functional impairment in patients with cataract. Med Care 1995; 33 (4 Suppl):AS120–30.
10. Desai P, Reidy A, Minassian DC, Vafidis G, Bolger J. Gains from cataract surgery: visual function and quality of life. Br J Ophthalmol 1996; 80: 868–73.
11. Elliott DB, Patla A, Bullimore MA. Improvements in clinical and functional vision and perceived visual disability after first and second eye cataract surgery. Br J Ophthalmol 1997; 81: 889–95.
12. Bass EB, Wills S, Scott IU, Javitt JC, Tielsch JM, Schein OD, Steinberg EP. Preference values for visual states in patients planning to undergo cataract surgery. Med Decis Making 1997; 17: 324–30.
13. Fletcher A, Gore S, Jones D, Fitzpatrick R, Spiegelhalter D, Cox D. Quality of life measures in health care. : II: Design, analysis, and interpretation. BMJ 1992; 305: 1145–8.
14. Muldoon MF, Barger SD, Flory JD, Manuck SB. What are quality of life measurements measuring? BMJ 1998; 316: 542–5.
15. Laidlaw DA, Harrad RA, Hopper CD, Whitaker A, Donovan JL, Brookes ST, Marsh GW, Peters TJ, Sparrow JM, Frankel SJ. Randomised trial of effectiveness of second eye cataract surgery. Lancet 1998; 352: 925–9.
16. Laidlaw A, Harrad R. Can second eye cataract extraction be justified? Eye 1993; 7: 680–6.
17. West SK, Rubin GS, Munoz B, Abraham D, Fried LP. Assessing functional status: correlation between performance on tasks conducted in a clinic setting and performance on the same task conducted at home. The Salisbury Eye Evaluation Project Team. J Gerontol A Biol Sci Med Sci 1997; 52: M209–17.
18. Guralnik JM, Winograd CH. Physical performance measures in the assessment of older persons. Aging (Milano) 1994; 6: 303–5.
19. Mangione CM, Phillips RS, Seddon JM, Lawrence MG, Cook EF, Dailey R, Goldman L. Development of the ‘Activities of Daily Vision Scale.’ A measure of visual functional status. Med Care 1992; 30: 1111–26.
20. Felson DT, Anderson JJ, Hannan MT, Milton RC, Wilson PW, Kiel DP. Impaired vision and hip fracture. The Framingham Study. J Am Geriatr Soc 1989; 37: 495–500.
21. Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM. Risk factors for hip fracture in white women. Study of Osteoporotic Fractures Research Group. N Engl J Med 1995; 332: 767–73.
22. Ivers RQ, Cumming RG, Mitchell P, Attebo K. Visual impairment and falls in older adults: the Blue Mountains Eye Study. J Am Geriatr Soc 1998; 46: 58–64.
23. Day H, Jutai J. Measuring the psychosocial impact of assistive devices: the PIADS. Can J Rehabil 1996; 9: 159–68.
24. Elliott DB, Bullimore MA. Assessing the reliability, discriminative ability, and validity of disability glare tests. Invest Ophthalmol Vis Sci 1993; 34: 108–19.
25. Rubin GS, Roche KB, Prasada-Rao P, Fried LP. Visual impairment and disability in older adults. Optom Vis Sci 1994; 71: 750–60.
26. Elliott DB, Gilchrist J, Whitaker D. Contrast sensitivity and glare sensitivity changes with three types of cataract morphology: are these techniques necessary in a clinical evaluation of cataract? Ophthalmic Physiol Opt 1989; 9: 25–30.
27. Bailey IL, Bullimore MA. A new test for the evaluation of disability glare. Optom Vis Sci 1991; 68: 911–7.
28. Pesudovs K, Coster DJ. An instrument for assessment of subjective visual disability in cataract patients. Br J Ophthalmol 1998; 82: 617–24.
29. Owsley C. Vision and driving in the elderly. Optom Vis Sci 1994; 71: 727–35.
30. Patla AE, Prentice SD, Robinson C, Neufeld J. Visual control of locomotion: strategies for changing direction and for going over obstacles. J Exp Psychol Hum Percept Perform 1991; 17: 603–34.
31. Elliott DB, Bullimore MA, Patla AE, Whitaker D. Effect of a cataract simulation on clinical and real world vision. Br J Ophthalmol 1996; 80: 799–804.
32. Bullimore MA, Bailey IL. Reading and eye movements in age-related maculopathy. Optom Vis Sci 1995; 72: 125–38.
33. Bullimore MA, Bailey IL, Wacker RT. Face recognition in age-related maculopathy. Invest Ophthalmol Vis Sci 1991; 32: 2020–9.
34. Hulley SB, Cummins SR. Designing Clinical Research: An Epidemiologic Approach. Baltimore: Wilkins & Wilkins; 1988. p. 136–7.
35. Browner WS, Newman TB. Are all significant P values created equal? The analogy between diagnostic tests and clinical research. JAMA 1987; 257: 2459–63.
36. Legge GE, Rubin GS, Pelli DG, Schleske MM. Psychophysics of reading. : II. Low vision. Vision Res 1985; 25: 253–65.
37. Brown NA, Hill AR. Cataract: the relation between myopia and cataract morphology. Br J Ophthalmol 1987; 71: 405–14.
38. Bartlett JD. Anisometropia and aniseikonia. In: Amos JF, ed. Diagnosis and Management in Vision Care. Oxford: Butterworth-Heinemann; 1987. p. 173–202.
39. Prudham D, Evans JG. Factors associated with falls in the elderly: a community study. Age Ageing 1981; 10: 141–6.
40. Langlois JA, Keyl PM, Guralnik JM, Foley DJ, Marottoli RA, Wallace RB. Characteristics of older pedestrians who have difficulty crossing the street. Am J Public Health 1997; 87: 393–7.
41. Allard R. Excess mortality from traffic accidents among elderly pedestrians living in the inner city. Am J Public Health 1982; 72: 853–4.
42. Monestam E, Wachtmeister L. Impact of cataract surgery on car driving: a population based study in Sweden. Br J Ophthalmol 1997; 81: 16–22.
43. Elliott DB, Hurst MA, Weatherill J. Comparing clinical tests of visual function in cataract with the patient’s perceived visual disability. Eye 1990; 4: 712–7.
44. Leat SJ, Woo GC. The validity of current clinical tests of contrast sensitivity and their ability to predict reading speed in low vision. Eye 1997; 11: 893–9.
45. Nevitt MC, Cummings SR, Kidd S, Black D. Risk factors for recurrent nonsyncopal falls. A prospective study. JAMA 1989; 261: 2663–8.