- Email: [email protected]
Visual Impairment and Its Impact on Health-related Quality of Life in Adolescents
Visual Impairment and Its Impact on Health-related Quality of Life in Adolescents HWEE-BEE WONG, DAVID MACHIN, SAY-BENG TAN, TIEN-YIN WONG, AND SEANG-MEI SAW ● PURPOSE:
To determine the impact of visual impairment on health-related quality of life (QoL) measures in adolescents. ● DESIGN: School-based, cross-sectional study. ● METHODS: Adolescents aged 11 to 18 years from the Singapore Cohort Study of the Risk Factors for Myopia were analyzed. QoL scores were determined using parallel child-self and parent proxy-report of PedsQL 4.0 Generic Core Scales. Refractive error was measured using the table-mounted autorefractor (model RK5 Canon Inc, Ltd, Tochigiken, Japan) and habitual distance logarithm of the minimal angle of resolution (logMAR) visual acuity charts were used. ● RESULTS: Data on 1,249 adolescents and 948 parents were analyzed. The prevalence of better eye presenting visual impairment > 0.3 logMAR was 5.7%. The mean (standard deviation) total, physical, and psychosocial health scores of all adolescents were 83.6 (11.8), 89.9 (11.8), and 80.3 (13.7). Healthy adolescents with visual impairment reported statistically but not clinically lower total (ⴚ3.8; 95% confidence interval [CI], ⴚ7.1 to ⴚ0.5; P ⴝ .03), psychosocial (ⴚ4.2; 95% CI, ⴚ8.1 to ⴚ0.3; P ⴝ .03), and school functioning scores (ⴚ5.5, 95% CI, ⴚ10.2 to ⴚ0.9; P ⴝ .02) than those with normal vision. However, no significant difference was observed in the parent proxy-reported scores between the two groups. Differences in total scores between high (1.9; 95% CI, ⴚ0.6 to 4.4) and low-myopes (0.2; 95% CI, ⴚ1.3 to 1.6) compared with nonmyopes were not significant. Comparable scores were also reported by hyperopes, astigmatism, and their counterparts, as well as their Accepted for publication Sep 18, 2008. From the Health Services Research and Evaluation Division, Ministry of Health (H.-B.W.), Singapore, Republic of Singapore; the Clinical Trials and Epidemiology Research Unit (H.-B.W., D.M., S.-B.T.), Singapore, Republic of Singapore; the Department of Community, Occupational and Family Medicine, National University of Singapore (H.-B.W., T.-Y.W., S.-M.S.), Singapore, Republic of Singapore; the Division of Clinical Trials and Epidemiological Sciences, National Cancer Centre (D.M.), Singapore, Republic of Singapore; the Medical Statistics Group, School of Health and Related Sciences, University of Sheffield (D.M.), Sheffield, United Kingdom; the Singapore Clinical Research Institute (S.-B.T.), Singapore, Republic of Singapore; the Duke-National University of Singapore Graduate Medical School (S.-B.T.), Singapore, Republic of Singapore; the Centre for Eye Research Australia, University of Melbourne (T.-Y.W.), Melbourne, Victoria, Australia; and the Singapore Eye Research Institute, Singapore National Eye Centre (T.-Y.W., S.M.S.), Singapore, Republic of Singapore. Inquiries to Hwee-Bee Wong, Health Services Research and Evaluation Division, Ministry of Health, 16 College Road, Singapore 169854, Republic of Singapore; e-mail: [email protected] 0002-9394/09/$36.00 doi:10.1016/j.ajo.2008.09.025
©
2009 BY
parents. Concordance between child and parent proxyreport was < 0.07. ● CONCLUSION: Healthy adolescents with visual impairment experienced statistically though not clinically impaired health related QoL, but refractive errors did not appear to have an impact on QoL. (Am J Ophthalmol 2009;147: 505–511. © 2009 by Elsevier Inc. All rights reserved.)
I
N CHILDREN AND ADOLESCENTS, UNDER-CORRECTED
refractive error is the main cause of visual impairment.1–3 There have been previous studies on the impact of visual impairment and refractive error on quality of life (QoL). A study of 1,361 Taiwanese adults at least 65 years of age suggested that lower general health QoL score, as measured by the Short-Form Health Survey (SF36) questionnaire, was associated with visual impairment (67.5 vs 71.2).4 Lower QoL scores across all domains of QoL measured by a validated instrument in the population of Indian adults 40 years of age or older were also reported.5 Similarly, in 112 myopic adults aged 18 to 65 years from the United Kingdom, persons with higher severity of myopia reported a lower QoL score, as assessed by the Vision-Related QoL Core Measure, as compared with those with moderate and low grades of myopia.6 However, there are currently no data on the impact of visual impairment or refractive error on health-related QoL in children and adolescents. The primary aim of this study is to evaluate the relationship between better eye presenting visual impairment, refractive errors, and QoL measures in a cohort of healthy adolescents. We also investigated whether parent-proxy’s and child’s reports of QoL measures were correlated.
METHODS ● STUDY POPULATION:
This was a cross-sectional study using data from the Singapore Cohort Study of the Risk Factors for Myopia (SCORM), described previously in other reports. In brief, all children aged 7 to 9 years of age from three schools located in the Southeastern, Western, and Northern parts of Singapore were invited to join the study in November 3, 1999 and May 16, 2001. Children with any serious medical or eye disorder, such as congenital cataract, were excluded. The methodology has been previously described. Of the 1979 children initially enrolled
ELSEVIER INC. ALL
RIGHTS RESERVED.
505
in SCORM, adolescents who attended the eye examination during 2005 to 2006 school visits and their parentproxy were invited to participate in this study.7,8 ● INSTRUMENTS: Although a number of instruments are available to evaluate QoL in adult patients with eye conditions, none has been developed specifically for children or adolescent populations for ophthalmic conditions. Thus, we used a generic instrument, the Pediatric Quality of Life Inventory Version 4.0 Generic Core Scales (PedsQL 4.0), which has been used for children/adolescents ranging from ages 13 to 18 years.9 The 23-item PedsQL 4.0 measures the core physical, mental, and social health dimensions as delineated by the World Health Organization, as well as role (school) functioning.10 This validated instrument had been used in the measurement of QoL of both healthy children and children with different diseases although not in children with ophthalmic conditions.11–16 The PedsQL 4.0 comprises parallel child selfreport and parent-proxy report formats for age ranges of 5 to 7 years, 8 to 12 years, and 13 to 18 years. The items for each of the reports are essentially identical, differing only in developmentally appropriate language and the use of first- or third-person tense. Instructions ask about the difficulty of performing each item during the past month (eg, “It is hard for me to run”).9 Responses are made on a five-point Likert scale. Items are reverse-scored and linearly transformed to a 0 to 100 score. A total and two subscales: physical and psychosocial health summary scores can be derived from the reversed item scores, with higher scores indicating better QoL. As the majority of SCORM subjects are Chinese, the parallel parent-proxy report was translated into Mandarin by two bilingual speakers and backward translated into English by another two bilingual speakers. A pilot testing of the translated Mandarin report was also carried out following recommended guidelines.17 Both English and Mandarin versions of the reports were sent to the parentproxy for their completion at home. During the school visit, the English version of the child self-report was completed by the adolescents. Current medical conditions such as asthma and migraine were reported using a self-administered questionnaire. The height and weight were also measured and the adolescents were classified as obese, overweight, or normal weight based on their age and gender using the recommended international cut-off points for body mass index by Cole and associates.18 The sociodemographic factors including the child’s date of birth, gender, ethnicity, and father’s highest educational levels were derived using parent-administered questionnaires.
FIGURE. Bland-Altman plot. Agreement between child selfreported and parent-proxy reported total scale scores.
protocol. After instillation of proparacaine 0.5%, cycloplegia was induced in each eye by administering 3 drops of cyclopentolate 1% solution at 5-minute intervals. Thirty minutes after the last drop of cyclopentolate solution was instilled, an autokeratorefractometer (model RK5; Canon Inc, Ltd, Tochigiken, Japan) was used to obtain the average of 5 consecutive measurements (all readings ⬍ 0.25 diopters [D] apart). The machines were calibrated at the beginning of the study. The measurements were made by staff that were masked to the contents of the instruments. ● DEFINITIONS:
Visual impairment was defined as presenting VA of ⬎ 0.3 logMAR in the better-seeing eye, according to the United States of America driving requirement.3,4,19 Day-to-day vision is driven primarily by VA in the better eye. The spherical equivalent (SE) was calculated as sphere power ⫹ 0.5 negative cylinder power. Adolescents were classified based on refractive groups defined by the SE of the worst eye: high-myope (SE ⱕ ⫺6.0 D), low-myope (⫺6.0 D ⬍ SE ⱕ ⫺0.5 D), and nonmyope (SE ⬎ ⫺0.5 D). The definition of myopia based on myopia in the worst or any eye has been adopted by several studies including the refractive error study in children (RESC) studies.1–3 Hyperope was defined as SE ⬎ ⫹1.0 D and astigmatism as at least ⫺0.5 cylinder in either eye. ● STATISTICAL ANALYSIS:
The principal study endpoint was the total score, and for this domain, the differences in means of those with better eye presenting visual impairment (BEVI) ⬎ 0.3 logMAR BEVI present or absent were compared and the associated 95% confidence intervals (CIs) and P values were presented. Scale scores were computed as the sum of the items divided by the number of items answered. If more than 50% of the items in the scale are missing, the scale score were not computed. Multivariable adjusted analysis of these differences con-
● VISUAL ACUITY AND REFRACTIVE ERROR MEASUREMENTS: Measurements of habitual distance logarithm of
the minimal angle of resolution (logMAR) visual acuity (VA) in each eye were conducted according to a standard 506
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009
TABLE 1. Characteristics of Adolescents Presence or Absence of Better Eye Presenting Visual Impairment ⬎ 0.3 Logarithm of the Minimal Angle of Resolution BEVI Child
Better eye visual acuity (logMAR) Mean (SD) Median (range) Age in yrs (%), mean (SD) 11 12 13 14 15 ⱖ 16 Gender, n (%) Female Male Ethnicity, n (%) Chinese Malay Indian Others Father’s highest educational level, n (%) No formal education Primary Secondary Preuniversity/diploma Tertiary/university Medical problem, n (%) No Yes Asthma Diabetes mellitus Inflammatory bowel disease Migraine Obese Overweight Others Use of corrective device, n (%) No Yes Spectacles only Contact lens only Spectacles and contact lens
Present (n ⫽ 71)
Absent (n ⫽ 1,178)
0.45 (0.02) 0.4 (0.30 to 1.00) 13.8 (1.3) 12 (7) 8 (4) 13 (4) 28 (9) 7 (4) 3 (4)
0.01 (0.01) 0 (⫺0.60 to 0.28) 13.8 (1.4) 158 (93) 199 (96) 288 (96) 296 (91) 158 (96) 79 (96)
P value
.81
.64 38 (6) 33 (5)
597 (94) 581 (95)
48 (5) 12 (6) 4 (5) 7 (1)
830 (95) 201 (94) 83 (95) 64 (90)
2 (4) 26 (8) 30 (6) 5 (3) 8 (5)
46 (96) 309 (92) 490 (94) 185 (97) 146 (95)
50 (6) 21 (6) 5 0 0 4 4 8 2
845 (94) 333 (94) 81 2 3 20 56 180 30
9 (2) 62 (7) 58 0 4
352 (98) 826 (93) 778 3 45
.53
.15
.81
⬍.01
BEVI ⫽ better eye presenting visual impairment; logMAR ⫽ logarithm of the minimal angle of resolution; SD ⫽ standard deviation; yrs ⫽ years.
trolling for age, gender, ethnicity, father’s highest educational level, medical conditions, and use of corrective device were also performed. The variables included in the model were selected based on the subject-matter knowledge. Comparison of other domains and adjusted comparisons were primarily hypothesis generating; the associated P values only regarded as indicative of possible differences with no adjustment for multiple testing being made.20 Analyses were repeated for healthy adolescents without VOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
any medical problems. The concordance between child and parent-proxy was determined through the intra-class correlation (ICC) coefficient. The agreement of total scores reported by the child and parent-proxy was also assessed using Bland-Altman plot (Figure). These plot the difference in total scores between the parent-proxy and child against the mean value of the two scores. The plot allows the mean of these difference, d, to be visually displayed and illustrates the scatter around that mean. The QOL
OF
ADOLESCENTS
507
TABLE 2. Self-reported Total and Summary Scores Among Adolescents by Presence or Absence of Better Eye Presenting Visual Impairment ⬎ 0.3 Logarithm of the Minimal Angle of Resolution BEVI Present
N Total Mean (SD) Interquartile range Range Physical health Mean (SD) Interquartile range Range Psychosocial health Mean (SD) Interquartile range Range Emotional functioning Mean (SD) Interquartile range Range Social functioning Mean (SD) Interquartile range Range School functioning Mean (SD) Interquartile range Range
71
Absent
Difference (Yes – No) 95% Confidence Interval
P value
1178
81.6 (14.1) 73 to 92 43 to 100
83.7 (11.6) 76 to 92 30 to 100
⫺2.2 (⫺5.0 to 0.6)
.13
87.7 (14.4) 81 to 100 41 to 100
90.0 (11.6) 84 to 100 25 to 100
⫺2.3 (⫺5.1 to 0.5)
.11
78.3 (15.3) 67 to 90 45 to 100
80.4 (13.6) 72 to 92 32 to 100
⫺2.1 (⫺5.4 to 1.2)
.21
73.2 (20.5) 60 to 90 20 to 100
75.8 (18.2) 60 to 90 0 to 100
⫺2.6 (⫺7.0 to 1.8)
.25
87.3 (15.3) 80 to 100 45 to 100
87.4 (15.0) 80 to 100 15 to 100
⫺0.1 (⫺3.7 to 3.5)
.97
74.3 (17.6) 60 to 90 30 to 100
78.0 (16.6) 65 to 90 0 to 100
⫺3.6 (⫺7.6 to 0.4)
.08
Multivariable Analysis: Healthy Subjects Without Medical Problems and Adjusted for Age, Gender, Ethnicity, Father’s Highest Educational Level, and Use of Corrective Device N 50 843 Total Adjusted mean 79.90 83.7 ⫺3.8 (⫺7.1 to ⫺0.5) .03 Physical health Adjusted mean 87.1 90.0 ⫺2.9 (⫺6.1 to 0.4) .08 Psychosocial health Adjusted mean 76.1 80.3 ⫺4.2 (⫺8.1 to ⫺0.3) .03 Emotional functioning Adjusted mean 70.3 74.7 ⫺4.4 (⫺9.6 to 0.7) .09 Social functioning Adjusted mean 86.6 89.3 ⫺2.7 (⫺6.9 to 1.5) .21 School functioning Adjusted mean 71.4 77.0 ⫺5.5 (⫺10.2 to ⫺0.9) .02 BEVI ⫽ better eye presenting visual impairment; logMAR ⫽ logarithm of the minimal angle of resolution; SD ⫽ standard deviation.
limits of agreement are calculated by d ⫾ 1.96sd where sd is standard deviation (SD) of the difference. If the child and parent-proxy tend to disagree but without a consistent pattern of one rating higher than the other, the mean will be near zero but the agreement limits will be wide.21 All probabilities quoted are two-sided, with a significance level of .05. The statistical analyses were carried out using SAS version 9.1 (SAS Institute Inc, Cary, North Carolina, USA). 508
AMERICAN JOURNAL
RESULTS OF THE 1,554 ADOLESCENTS INVITED FOR A FOLLOW-UP EYE
examination during the 2005 to 2006 school visits, 1,250 (80.4%) attended the examination although one was aged 20 years and so is not considered further in this study. All these adolescents and the parent-proxy of 948 (61%: 72% by mother and 86% in English) completed the QoL reports. The Cronbach alphas were all ⬎ 0.70 for the OF
OPHTHALMOLOGY
MARCH 2009
Mandarin parent-proxy report (total, 0.92; physical, 0.84; psychosocial, 0.90; emotional, 0.86; social, 0.81; school, 0.77). The parent-proxy of younger (P ⬍ .01), female (P ⬍ .01), and Chinese (compared with Malay, P ⬍ .01) adolescents were more likely to complete the parent-proxy reports. The adolescents’ mean age was 13.8 years (SD, 1.4 years). Fifty-one percent were female, with the racial composition being 70% Chinese, 17% Malay, 7% Indian, and 6% others. Three hundred and fifty-four (28%) had medical conditions such as overweight (188), asthma (87), obesity (60), and migraine (24). The BEVI was identified for 71 adolescents and the prevalence rate was 5.7% (95% CI, 4.5 to 7.1). One hundred and ten (8.8%; 95% CI, 7.4 to 10.5) had high-myopia, 784 (62.8%; 95% CI, 60.1 to 65.4) low-myopia, and 355 (28.4%; 95% CI, 26.0 to 31.0) had nonmyopia. Based on either of the eyes, 81 (6.5%; 95% CI, 5.2 to 8.0) out of 1,249 had hyperopia and 972 (77.8%; 95% CI, 75.4 to 80.0) had astigmatism. The age, gender, ethnicity, father’s highest educational levels, and medical conditions of adolescents with BEVI present or absent were comparable (Table 1). The mean (SD) total, physical, and psychosocial health scores of all adolescents were 83.6 (11.8), 89.9 (11.8), and 80.3 (13.7), respectively, while 83.1 (13.0), 87.7 (12.9), and 80.7 (14.7) were reported by their parent-proxy. The adolescents with BEVI reported lower total (⫺2.2; 95% CI, ⫺5.0 to 0.6) and other summary scores (Table 2), but the differences were statistically insignificant even after adjusting for age, gender, ethnicity, father’s highest educational levels, medical conditions, and use of corrective device of adolescents. When only the 895 healthy adolescents without any medical problem were considered, the total (⫺3.8; 95% CI, ⫺7.1 to ⫺0.5; P ⫽ .03), psychosocial (⫺4.2; 95% CI, ⫺8.1 to ⫺0.3; P ⫽ .03), and school functioning scores (⫺5.5; 95% CI, ⫺10.2 to ⫺0.9; P ⫽ .02) of the adolescents with BEVI present were significantly lower. The parent-proxy of adolescents with BEVI present reported slightly lower mean [SD] total (82.4 [11.0] vs 83.1 [13.1]; P ⫽ .70), physical (85.9 [11.6] vs 87.7 [13.0]; P ⫽ .35), and psychosocial (80.5 [12.6] vs 80.7 [14.9]; P ⫽ .93) scores. A lack of a statistically significant difference was observed between the parent-proxy reported mean scores of all scales even after adjustment. Similarly, no significant difference was observed in the parent-proxy reported scores for healthy adolescents only. The results were consistent when the English and Mandarin parent-proxy reports were analyzed separately. The mean (SD) total score reported by the high, low, and nonmyopes were 85.3 (12.1), 83.5 (11.8), and 83.4 (11.8). The differences reported by high (1.9; 95% CI, ⫺0.6 to 4.4) and low-myopes (0.2; 95% CI, ⫺1.3 to 1.6) compared with nonmyopes were not significant and the results remain unaltered after adjustment. A lack of a statistically significant difference was also observed beVOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
tween the self-reported mean scores by healthy adolescents. However, the parent-proxy of high-myopes reported significant lower total (⫺4.4; 95% CI, ⫺8.4 to ⫺0.4; P ⫽ .03), psychosocial (⫺5.5; 95% CI, ⫺10.0 to ⫺1.0; P ⫽ .02), and social (⫺7.1; 95% CI, ⫺12.0 to ⫺2.1; P ⬍ .01) scores than parent-proxy of nonmyopes after adjustment. Slightly greater differences in these scores and school functioning were observed when only healthy adolescents were considered (total, ⫺5.6; 95% CI, ⫺10.3 to ⫺0.9; P ⫽ .02; psychosocial, ⫺7.0; 95% CI, ⫺12.4 to ⫺1.7; P ⬍ .01; social, ⫺7.2; 95% CI, ⫺13.0 to ⫺1.4; P ⫽ .01; school, ⫺7.5; 95% CI, ⫺13.5 to ⫺1.4; P ⫽ .02). Lower social score was also reported by parent-proxy of high-myopes when compared with low-myopes (all, ⫺4.7; 95% CI, ⫺8.3 to ⫺1.0; P ⫽ .01; healthy only, ⫺4.5; 95% CI, ⫺8.6 to ⫺0.5; P ⫽ .03). Parent-proxy of low-myopes only reported lower score than nonmyopes in school functioning (⫺4.6; 95% CI, ⫺9.2 to ⫺0.1; P ⫽ .047). No significant difference was observed between hyperopes and nonhyperopes in total (0.8; 95% CI, ⫺1.8 to 3.5; P ⫽ .54) and other self-reported mean scores (physical, P ⫽ .98; psychosocial, P ⫽ .42). Adolescents with astigmatism had self-reported (total, P ⫽ .92; physical, P ⫽ .42; psychosocial, P ⫽ .81) scores that were comparable to their counterparts. Except for higher emotional scores reported by parent-proxy of hyperopes (5.6; 95% CI, 0.6 to 10.6; P ⫽ .03), nonsignificant different results were found for other parent-proxy reported scores and when only healthy adolescents were analyzed. Most of the effect size concordance between child self-report and parent-proxy report in BEVI and each refractive group were small; all ICC values ⬍ 0.07 (data not shown). The Figure shows a comparison of total scores reported by child and parent-proxy. The mean difference was 0.8 (95% CI, ⫺0.03 to 1.63). The limits of agreement ranged from ⫺25.2 to 26.8, corresponding to two points on the Likert scales. Twenty-nine (3.1%) child-parent dyads lie above the upper limits (ie, the child rated at least one-point higher than their parent-proxy) while 26 (2.7%) child-parent dyads had child rated at least one point lower. The majority (71%) of the child-parent dyads had a difference within one point on the Likert scale (25 points).
DISCUSSION THIS STUDY IS UNIQUE IN THAT IT EVALUATES THE IMPACT
of presenting visual impairment and refractive errors in an Asian population of adolescent school pupils. The prevalence of presenting visual impairment of ⬎ 0.3 logMAR in the better eye amongst 1,249 adolescents with a mean age of 13.8 years was 5.7%. Adolescents with impairment reported lower total, psychosocial, and school scores, as measured by the PedsQL 4.0 scale. Thus, presenting visual QOL
OF
ADOLESCENTS
509
impairment ⬎ 0.3 logMAR was significantly associated with a decrement in their overall well-being, although these differences are below the minimal clinically meaningful differences (total, 4.36; psychosocial, 5.30; school, 9.12) suggested by Varni and associates.11 The most common cause of low presenting visual impairment in our study, and in others, is under-corrected refractive errors, but we found no association between refractive error per se and QoL. Nevertheless, presenting visual impairment may affect school learning, outdoor activity, and social life. This may compromise an adolescent’s school achievement and thus a lower school functioning score is reported. Our study also showed a deleterious effect on psychosocial functioning of adolescents with visual impairment. However, much lower PedsQL scores have been reported by children with cancer (total, 72.2; physical, 71.8; psychosocial, 72.6) and chronically ill children (total, 77.2; physical, 77.4; psychosocial 77.1).9,12 Overall, we found similar QoL scores for adolescents with and without refractive errors. Unlike cataract and glaucoma, which usually require surgical intervention, refractive errors are corrected using spectacles or contact lenses. Thus, most people with refractive error have good vision and show no reduction in their QoL. Although an adult study suggested that high-myopic patients had poorer vision-related QoL compared with moderate and mildmyopic patients, this result was not observed in our study.6 A possible explanation is that most complications associated with high-myopia, such as myopic macular degeneration, macular holes, retinal breaks, and tears, occur during the later years of life. In our study, there was a low prevalence of complications secondary to high myopia. The concordance in QoL measures between adolescents with BEVI or refractive errors and their parent-proxy were small. The result of a Bland-Altman plot also showed that there was considerable difference between scores reported by the child and their parent-proxy (ie, difference ranges within two points on the Likert scale). The lack of correlation between the parent-proxy and child’s report may be attributable to differences in perception of the teenager’s QOL, especially by parents who are working full-time. Hence, parent-proxy report should be included to complement child-self report in pediatric population with BEVI or refractive errors. Unlike the various instruments for the measurement of vision-related QoL in adults, such as Vision-Related QoL
Core Measure and NEI-VFQ-25, PedsQL, as a generic instrument, may not be sensitive enough to detect subtle effects attributable to the ocular conditions. Nevertheless, this well validated (Cronbach alphas reported were all ⬎ 0.709,12–16) and widely used instrument that allows for comparisons of QoL in the pediatric population across a variety of medical conditions such as asthma, cancer, diabetes, epilepsy, inflammatory bowel disease, headaches, obesity, pediatric rheumatology, cardiology, orthopedics, and psychiatrics warranted its appropriateness for the purpose of this study. Furthermore, with its parallel parentproxy report, we were able to investigate the relationship between child and parent-proxy ratings of their QoL. Although the multiplicity of potential endpoints arises with PedsQL, as with other QoL instruments, we have focused on the total score as the principal endpoint and regarded the other scores as secondary to this. In addition, multiple statistical testing can create problems in the interpretation of the results and may inflate the number of statistically significant differences found. In our case, there was a paucity of statistically significant differences (using the conventional P ⬍ .05), so this is likely to have little impact on the conclusions we draw. A limitation of our study is that the best-corrected VA was not assessed and thus visual impairment attributable to uncorrected refractive error (differences of habitual and best-corrected VA) cannot be determined. Other causes of visual impairment such as congenital glaucoma and cataract were also not assessed; however, these conditions are likely to be rare. Our study was also limited by the differences found in child’s age, gender, and ethnicity between the respondents and nonrespondent of the parent-proxy report. This study is the first Asian study to investigate the impact of BEVI and refractive errors on QoL using the PedQoL instrument in the pediatric population, and to evaluate the concordance between child self-report and parent-proxy report of QoL with a well validated measure. Other strengths of the study include a high participation rate of 80% for the child self-report and 61% for the parent-proxy report. In summary, we show in this population of healthy adolescents that BEVI was statistically though not clinically associated with impaired total, psychosocial, and school functioning QoL scores. Thus, there is a need to understand and address visual impairment at a young age.
THIS STUDY WAS SUPPORTED BY THE NATIONAL MEDICAL RESEARCH COUNCIL (NMRC/0975/2005), SINGAPORE, REPUBLIC OF Singapore and a Grant from the Singapore Children Society (RNO/059/06), Singapore, Republic of Singapore. The authors indicate no financial conflict of interest. Involved in design and conduct of study (H.-B.W., S.-M.S., D.M., S.-B.T.); data collection and management (H.-B.W.); obtaining funding and providing resources (H.-B.W., S.M.S.); analysis and interpretation of data (H.-B.W., S.M.S., D.M, S.-B.T.); and manuscript preparation, critical revision, and final approval of manuscript (H.-B.W., S.M.S., D.M., S.-B.T., T.-Y.W.). Approval was granted by the Institutional Review Board, Singapore Eye Research Institute, and the study protocol adhered to the tenets of the Declaration of Helsinki. The adolescents signed an assent form and their parent-proxy’s written informed consent was obtained.
510
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009
REFERENCES 1. Pokharel GP, Negrel AD, Munoz SR, Ellwein LB. Refractive error study in children: results from Mechi Zone, Nepal. Am J Ophthalmol 2000;129:436 – 444. 2. Zhao J, Pan X, Sui R, Munoz SR, Sperduto RD, Ellwein LB. Refractive error study in children: results from Shunyi District, China. Am J Ophthalmol 2000;129:427– 435. 3. Maul E, Barroso S, Munoz SR, Sperduto RD, Ellwein LB. Refractive error study in children: results from La Florida, Chile. Am J Ophthalmol 2000;129:445– 454. 4. Tsai SY, Chi LY, Cheng CY, Hsu WM, Liu JH, Chou P. The impact of visual impairment and use of eye services on health-related quality of life among the elderly in Taiwan: the Shihpai Eye Study. Qual Life Res 2004;13:1415–1424. 5. Nirmalan PK, Tielsch JM, Katz J, et al. Relationship between vision impairment and eye disease to vision-specific quality of life and function in rural India: the Aravind Comprehensive Eye Survey. Invest Ophthalmol Vis Sci 2005;46:2308 – 2312. 6. Rose K, Harper R, Tromans C, et al. Quality of life in myopia. Br J Ophthalmol 2000;84:1031–1034. 7. Saw SM, Hong CY, Chia KS, Stone R, Tan D. Nearwork and myopia in young children. Lancet 2001;357:390. 8. Saw SM, Chua WH, Hong CY, et al. Nearwork in early onset myopia. Invest Ophthalmol Vis Sci 2002;43:332–339. 9. Varni JW, Seid M, Rode CA. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care 2001;39:800 – 812. 10. World Health Organization. Constitution of the World Health Organization, basic document. Geneva, Switzerland: World Health Organization, 1948:1. 11. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr 2003;3:329 –341.
VOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
12. Varni JW, Burwinkle TM, Katz ER, Meeske K, Dickinson P. The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer 2002;94:2090 –2106. 13. Powers SW, Patton SR, Hommel KA, Hershey AD. Quality of life in childhood migraines: clinical impact and comparison to other chronic illnesses. Pediatrics 2003;112:e1– e5. 14. Upton P, Eiser C, Cheung I, et al. Measurement properties of the United Kingdom-English version of the Pediatric Quality of Life Inventory 4.0 (PedsQL) generic core scales. Health Qual Life Outcomes 2005;3:22. 15. Felder-Puig R, Frey E, Proksch K, Varni JW, Gadner H, Topf R. Validation of the German version of the Pediatric Quality of Life Inventory (PedsQL) in childhood cancer patients off treatment and children with epilepsy. Qual Life Res 2004; 13:223–234. 16. Bastiaansen D, Koot HM, Bongers IL, Varni JW, Verhulst FC. Measuring quality of life in children referred for psychiatric problems: psychometric properties of the PedsQL 4.0 generic core scales. Qual Life Res 2004;13:489 – 495. 17. Varni JW. Linguistic validation of the PedsQLTM – a Quality of Life Questionnaire 2002. Available at http:// www.pedsql.org/PedsQL-Linguistic-Validation-Guidelines.doc. Accessed: September 3, 2005. 18. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320:1240 –1243. 19. Fishbaugh J. Look who’s driving now–visual standards for driver licensing in the United States. Insight 1995;20:11–20. 20. Perneger TV. What’s wrong with Bonferroni adjustments. BMJ 1998;316:1236 –1238. 21. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310.
QOL
OF
ADOLESCENTS
511
Biosketch Hwee-Bee Wong, MSc, graduated from University of Malaya Malaysia with a BSc in Statistics and a MSc in Statistics from National University of Singapore, Singapore. She is currently pursuing her part-time PhD at Department of Community, Occupational and Family Medicine, Yong Loo Lin School of Medicine at National University of Singapore and working as a senior biostatistician in the Ministry of Health Singapore. Dr Wong’s area of research interest are epidemiology of pediatrics myopia and quality of life.
511.e1
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009
To determine the impact of visual impairment on health-related quality of life (QoL) measures in adolescents. ● DESIGN: School-based, cross-sectional study. ● METHODS: Adolescents aged 11 to 18 years from the Singapore Cohort Study of the Risk Factors for Myopia were analyzed. QoL scores were determined using parallel child-self and parent proxy-report of PedsQL 4.0 Generic Core Scales. Refractive error was measured using the table-mounted autorefractor (model RK5 Canon Inc, Ltd, Tochigiken, Japan) and habitual distance logarithm of the minimal angle of resolution (logMAR) visual acuity charts were used. ● RESULTS: Data on 1,249 adolescents and 948 parents were analyzed. The prevalence of better eye presenting visual impairment > 0.3 logMAR was 5.7%. The mean (standard deviation) total, physical, and psychosocial health scores of all adolescents were 83.6 (11.8), 89.9 (11.8), and 80.3 (13.7). Healthy adolescents with visual impairment reported statistically but not clinically lower total (ⴚ3.8; 95% confidence interval [CI], ⴚ7.1 to ⴚ0.5; P ⴝ .03), psychosocial (ⴚ4.2; 95% CI, ⴚ8.1 to ⴚ0.3; P ⴝ .03), and school functioning scores (ⴚ5.5, 95% CI, ⴚ10.2 to ⴚ0.9; P ⴝ .02) than those with normal vision. However, no significant difference was observed in the parent proxy-reported scores between the two groups. Differences in total scores between high (1.9; 95% CI, ⴚ0.6 to 4.4) and low-myopes (0.2; 95% CI, ⴚ1.3 to 1.6) compared with nonmyopes were not significant. Comparable scores were also reported by hyperopes, astigmatism, and their counterparts, as well as their Accepted for publication Sep 18, 2008. From the Health Services Research and Evaluation Division, Ministry of Health (H.-B.W.), Singapore, Republic of Singapore; the Clinical Trials and Epidemiology Research Unit (H.-B.W., D.M., S.-B.T.), Singapore, Republic of Singapore; the Department of Community, Occupational and Family Medicine, National University of Singapore (H.-B.W., T.-Y.W., S.-M.S.), Singapore, Republic of Singapore; the Division of Clinical Trials and Epidemiological Sciences, National Cancer Centre (D.M.), Singapore, Republic of Singapore; the Medical Statistics Group, School of Health and Related Sciences, University of Sheffield (D.M.), Sheffield, United Kingdom; the Singapore Clinical Research Institute (S.-B.T.), Singapore, Republic of Singapore; the Duke-National University of Singapore Graduate Medical School (S.-B.T.), Singapore, Republic of Singapore; the Centre for Eye Research Australia, University of Melbourne (T.-Y.W.), Melbourne, Victoria, Australia; and the Singapore Eye Research Institute, Singapore National Eye Centre (T.-Y.W., S.M.S.), Singapore, Republic of Singapore. Inquiries to Hwee-Bee Wong, Health Services Research and Evaluation Division, Ministry of Health, 16 College Road, Singapore 169854, Republic of Singapore; e-mail: [email protected] 0002-9394/09/$36.00 doi:10.1016/j.ajo.2008.09.025
©
2009 BY
parents. Concordance between child and parent proxyreport was < 0.07. ● CONCLUSION: Healthy adolescents with visual impairment experienced statistically though not clinically impaired health related QoL, but refractive errors did not appear to have an impact on QoL. (Am J Ophthalmol 2009;147: 505–511. © 2009 by Elsevier Inc. All rights reserved.)
I
N CHILDREN AND ADOLESCENTS, UNDER-CORRECTED
refractive error is the main cause of visual impairment.1–3 There have been previous studies on the impact of visual impairment and refractive error on quality of life (QoL). A study of 1,361 Taiwanese adults at least 65 years of age suggested that lower general health QoL score, as measured by the Short-Form Health Survey (SF36) questionnaire, was associated with visual impairment (67.5 vs 71.2).4 Lower QoL scores across all domains of QoL measured by a validated instrument in the population of Indian adults 40 years of age or older were also reported.5 Similarly, in 112 myopic adults aged 18 to 65 years from the United Kingdom, persons with higher severity of myopia reported a lower QoL score, as assessed by the Vision-Related QoL Core Measure, as compared with those with moderate and low grades of myopia.6 However, there are currently no data on the impact of visual impairment or refractive error on health-related QoL in children and adolescents. The primary aim of this study is to evaluate the relationship between better eye presenting visual impairment, refractive errors, and QoL measures in a cohort of healthy adolescents. We also investigated whether parent-proxy’s and child’s reports of QoL measures were correlated.
METHODS ● STUDY POPULATION:
This was a cross-sectional study using data from the Singapore Cohort Study of the Risk Factors for Myopia (SCORM), described previously in other reports. In brief, all children aged 7 to 9 years of age from three schools located in the Southeastern, Western, and Northern parts of Singapore were invited to join the study in November 3, 1999 and May 16, 2001. Children with any serious medical or eye disorder, such as congenital cataract, were excluded. The methodology has been previously described. Of the 1979 children initially enrolled
ELSEVIER INC. ALL
RIGHTS RESERVED.
505
in SCORM, adolescents who attended the eye examination during 2005 to 2006 school visits and their parentproxy were invited to participate in this study.7,8 ● INSTRUMENTS: Although a number of instruments are available to evaluate QoL in adult patients with eye conditions, none has been developed specifically for children or adolescent populations for ophthalmic conditions. Thus, we used a generic instrument, the Pediatric Quality of Life Inventory Version 4.0 Generic Core Scales (PedsQL 4.0), which has been used for children/adolescents ranging from ages 13 to 18 years.9 The 23-item PedsQL 4.0 measures the core physical, mental, and social health dimensions as delineated by the World Health Organization, as well as role (school) functioning.10 This validated instrument had been used in the measurement of QoL of both healthy children and children with different diseases although not in children with ophthalmic conditions.11–16 The PedsQL 4.0 comprises parallel child selfreport and parent-proxy report formats for age ranges of 5 to 7 years, 8 to 12 years, and 13 to 18 years. The items for each of the reports are essentially identical, differing only in developmentally appropriate language and the use of first- or third-person tense. Instructions ask about the difficulty of performing each item during the past month (eg, “It is hard for me to run”).9 Responses are made on a five-point Likert scale. Items are reverse-scored and linearly transformed to a 0 to 100 score. A total and two subscales: physical and psychosocial health summary scores can be derived from the reversed item scores, with higher scores indicating better QoL. As the majority of SCORM subjects are Chinese, the parallel parent-proxy report was translated into Mandarin by two bilingual speakers and backward translated into English by another two bilingual speakers. A pilot testing of the translated Mandarin report was also carried out following recommended guidelines.17 Both English and Mandarin versions of the reports were sent to the parentproxy for their completion at home. During the school visit, the English version of the child self-report was completed by the adolescents. Current medical conditions such as asthma and migraine were reported using a self-administered questionnaire. The height and weight were also measured and the adolescents were classified as obese, overweight, or normal weight based on their age and gender using the recommended international cut-off points for body mass index by Cole and associates.18 The sociodemographic factors including the child’s date of birth, gender, ethnicity, and father’s highest educational levels were derived using parent-administered questionnaires.
FIGURE. Bland-Altman plot. Agreement between child selfreported and parent-proxy reported total scale scores.
protocol. After instillation of proparacaine 0.5%, cycloplegia was induced in each eye by administering 3 drops of cyclopentolate 1% solution at 5-minute intervals. Thirty minutes after the last drop of cyclopentolate solution was instilled, an autokeratorefractometer (model RK5; Canon Inc, Ltd, Tochigiken, Japan) was used to obtain the average of 5 consecutive measurements (all readings ⬍ 0.25 diopters [D] apart). The machines were calibrated at the beginning of the study. The measurements were made by staff that were masked to the contents of the instruments. ● DEFINITIONS:
Visual impairment was defined as presenting VA of ⬎ 0.3 logMAR in the better-seeing eye, according to the United States of America driving requirement.3,4,19 Day-to-day vision is driven primarily by VA in the better eye. The spherical equivalent (SE) was calculated as sphere power ⫹ 0.5 negative cylinder power. Adolescents were classified based on refractive groups defined by the SE of the worst eye: high-myope (SE ⱕ ⫺6.0 D), low-myope (⫺6.0 D ⬍ SE ⱕ ⫺0.5 D), and nonmyope (SE ⬎ ⫺0.5 D). The definition of myopia based on myopia in the worst or any eye has been adopted by several studies including the refractive error study in children (RESC) studies.1–3 Hyperope was defined as SE ⬎ ⫹1.0 D and astigmatism as at least ⫺0.5 cylinder in either eye. ● STATISTICAL ANALYSIS:
The principal study endpoint was the total score, and for this domain, the differences in means of those with better eye presenting visual impairment (BEVI) ⬎ 0.3 logMAR BEVI present or absent were compared and the associated 95% confidence intervals (CIs) and P values were presented. Scale scores were computed as the sum of the items divided by the number of items answered. If more than 50% of the items in the scale are missing, the scale score were not computed. Multivariable adjusted analysis of these differences con-
● VISUAL ACUITY AND REFRACTIVE ERROR MEASUREMENTS: Measurements of habitual distance logarithm of
the minimal angle of resolution (logMAR) visual acuity (VA) in each eye were conducted according to a standard 506
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009
TABLE 1. Characteristics of Adolescents Presence or Absence of Better Eye Presenting Visual Impairment ⬎ 0.3 Logarithm of the Minimal Angle of Resolution BEVI Child
Better eye visual acuity (logMAR) Mean (SD) Median (range) Age in yrs (%), mean (SD) 11 12 13 14 15 ⱖ 16 Gender, n (%) Female Male Ethnicity, n (%) Chinese Malay Indian Others Father’s highest educational level, n (%) No formal education Primary Secondary Preuniversity/diploma Tertiary/university Medical problem, n (%) No Yes Asthma Diabetes mellitus Inflammatory bowel disease Migraine Obese Overweight Others Use of corrective device, n (%) No Yes Spectacles only Contact lens only Spectacles and contact lens
Present (n ⫽ 71)
Absent (n ⫽ 1,178)
0.45 (0.02) 0.4 (0.30 to 1.00) 13.8 (1.3) 12 (7) 8 (4) 13 (4) 28 (9) 7 (4) 3 (4)
0.01 (0.01) 0 (⫺0.60 to 0.28) 13.8 (1.4) 158 (93) 199 (96) 288 (96) 296 (91) 158 (96) 79 (96)
P value
.81
.64 38 (6) 33 (5)
597 (94) 581 (95)
48 (5) 12 (6) 4 (5) 7 (1)
830 (95) 201 (94) 83 (95) 64 (90)
2 (4) 26 (8) 30 (6) 5 (3) 8 (5)
46 (96) 309 (92) 490 (94) 185 (97) 146 (95)
50 (6) 21 (6) 5 0 0 4 4 8 2
845 (94) 333 (94) 81 2 3 20 56 180 30
9 (2) 62 (7) 58 0 4
352 (98) 826 (93) 778 3 45
.53
.15
.81
⬍.01
BEVI ⫽ better eye presenting visual impairment; logMAR ⫽ logarithm of the minimal angle of resolution; SD ⫽ standard deviation; yrs ⫽ years.
trolling for age, gender, ethnicity, father’s highest educational level, medical conditions, and use of corrective device were also performed. The variables included in the model were selected based on the subject-matter knowledge. Comparison of other domains and adjusted comparisons were primarily hypothesis generating; the associated P values only regarded as indicative of possible differences with no adjustment for multiple testing being made.20 Analyses were repeated for healthy adolescents without VOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
any medical problems. The concordance between child and parent-proxy was determined through the intra-class correlation (ICC) coefficient. The agreement of total scores reported by the child and parent-proxy was also assessed using Bland-Altman plot (Figure). These plot the difference in total scores between the parent-proxy and child against the mean value of the two scores. The plot allows the mean of these difference, d, to be visually displayed and illustrates the scatter around that mean. The QOL
OF
ADOLESCENTS
507
TABLE 2. Self-reported Total and Summary Scores Among Adolescents by Presence or Absence of Better Eye Presenting Visual Impairment ⬎ 0.3 Logarithm of the Minimal Angle of Resolution BEVI Present
N Total Mean (SD) Interquartile range Range Physical health Mean (SD) Interquartile range Range Psychosocial health Mean (SD) Interquartile range Range Emotional functioning Mean (SD) Interquartile range Range Social functioning Mean (SD) Interquartile range Range School functioning Mean (SD) Interquartile range Range
71
Absent
Difference (Yes – No) 95% Confidence Interval
P value
1178
81.6 (14.1) 73 to 92 43 to 100
83.7 (11.6) 76 to 92 30 to 100
⫺2.2 (⫺5.0 to 0.6)
.13
87.7 (14.4) 81 to 100 41 to 100
90.0 (11.6) 84 to 100 25 to 100
⫺2.3 (⫺5.1 to 0.5)
.11
78.3 (15.3) 67 to 90 45 to 100
80.4 (13.6) 72 to 92 32 to 100
⫺2.1 (⫺5.4 to 1.2)
.21
73.2 (20.5) 60 to 90 20 to 100
75.8 (18.2) 60 to 90 0 to 100
⫺2.6 (⫺7.0 to 1.8)
.25
87.3 (15.3) 80 to 100 45 to 100
87.4 (15.0) 80 to 100 15 to 100
⫺0.1 (⫺3.7 to 3.5)
.97
74.3 (17.6) 60 to 90 30 to 100
78.0 (16.6) 65 to 90 0 to 100
⫺3.6 (⫺7.6 to 0.4)
.08
Multivariable Analysis: Healthy Subjects Without Medical Problems and Adjusted for Age, Gender, Ethnicity, Father’s Highest Educational Level, and Use of Corrective Device N 50 843 Total Adjusted mean 79.90 83.7 ⫺3.8 (⫺7.1 to ⫺0.5) .03 Physical health Adjusted mean 87.1 90.0 ⫺2.9 (⫺6.1 to 0.4) .08 Psychosocial health Adjusted mean 76.1 80.3 ⫺4.2 (⫺8.1 to ⫺0.3) .03 Emotional functioning Adjusted mean 70.3 74.7 ⫺4.4 (⫺9.6 to 0.7) .09 Social functioning Adjusted mean 86.6 89.3 ⫺2.7 (⫺6.9 to 1.5) .21 School functioning Adjusted mean 71.4 77.0 ⫺5.5 (⫺10.2 to ⫺0.9) .02 BEVI ⫽ better eye presenting visual impairment; logMAR ⫽ logarithm of the minimal angle of resolution; SD ⫽ standard deviation.
limits of agreement are calculated by d ⫾ 1.96sd where sd is standard deviation (SD) of the difference. If the child and parent-proxy tend to disagree but without a consistent pattern of one rating higher than the other, the mean will be near zero but the agreement limits will be wide.21 All probabilities quoted are two-sided, with a significance level of .05. The statistical analyses were carried out using SAS version 9.1 (SAS Institute Inc, Cary, North Carolina, USA). 508
AMERICAN JOURNAL
RESULTS OF THE 1,554 ADOLESCENTS INVITED FOR A FOLLOW-UP EYE
examination during the 2005 to 2006 school visits, 1,250 (80.4%) attended the examination although one was aged 20 years and so is not considered further in this study. All these adolescents and the parent-proxy of 948 (61%: 72% by mother and 86% in English) completed the QoL reports. The Cronbach alphas were all ⬎ 0.70 for the OF
OPHTHALMOLOGY
MARCH 2009
Mandarin parent-proxy report (total, 0.92; physical, 0.84; psychosocial, 0.90; emotional, 0.86; social, 0.81; school, 0.77). The parent-proxy of younger (P ⬍ .01), female (P ⬍ .01), and Chinese (compared with Malay, P ⬍ .01) adolescents were more likely to complete the parent-proxy reports. The adolescents’ mean age was 13.8 years (SD, 1.4 years). Fifty-one percent were female, with the racial composition being 70% Chinese, 17% Malay, 7% Indian, and 6% others. Three hundred and fifty-four (28%) had medical conditions such as overweight (188), asthma (87), obesity (60), and migraine (24). The BEVI was identified for 71 adolescents and the prevalence rate was 5.7% (95% CI, 4.5 to 7.1). One hundred and ten (8.8%; 95% CI, 7.4 to 10.5) had high-myopia, 784 (62.8%; 95% CI, 60.1 to 65.4) low-myopia, and 355 (28.4%; 95% CI, 26.0 to 31.0) had nonmyopia. Based on either of the eyes, 81 (6.5%; 95% CI, 5.2 to 8.0) out of 1,249 had hyperopia and 972 (77.8%; 95% CI, 75.4 to 80.0) had astigmatism. The age, gender, ethnicity, father’s highest educational levels, and medical conditions of adolescents with BEVI present or absent were comparable (Table 1). The mean (SD) total, physical, and psychosocial health scores of all adolescents were 83.6 (11.8), 89.9 (11.8), and 80.3 (13.7), respectively, while 83.1 (13.0), 87.7 (12.9), and 80.7 (14.7) were reported by their parent-proxy. The adolescents with BEVI reported lower total (⫺2.2; 95% CI, ⫺5.0 to 0.6) and other summary scores (Table 2), but the differences were statistically insignificant even after adjusting for age, gender, ethnicity, father’s highest educational levels, medical conditions, and use of corrective device of adolescents. When only the 895 healthy adolescents without any medical problem were considered, the total (⫺3.8; 95% CI, ⫺7.1 to ⫺0.5; P ⫽ .03), psychosocial (⫺4.2; 95% CI, ⫺8.1 to ⫺0.3; P ⫽ .03), and school functioning scores (⫺5.5; 95% CI, ⫺10.2 to ⫺0.9; P ⫽ .02) of the adolescents with BEVI present were significantly lower. The parent-proxy of adolescents with BEVI present reported slightly lower mean [SD] total (82.4 [11.0] vs 83.1 [13.1]; P ⫽ .70), physical (85.9 [11.6] vs 87.7 [13.0]; P ⫽ .35), and psychosocial (80.5 [12.6] vs 80.7 [14.9]; P ⫽ .93) scores. A lack of a statistically significant difference was observed between the parent-proxy reported mean scores of all scales even after adjustment. Similarly, no significant difference was observed in the parent-proxy reported scores for healthy adolescents only. The results were consistent when the English and Mandarin parent-proxy reports were analyzed separately. The mean (SD) total score reported by the high, low, and nonmyopes were 85.3 (12.1), 83.5 (11.8), and 83.4 (11.8). The differences reported by high (1.9; 95% CI, ⫺0.6 to 4.4) and low-myopes (0.2; 95% CI, ⫺1.3 to 1.6) compared with nonmyopes were not significant and the results remain unaltered after adjustment. A lack of a statistically significant difference was also observed beVOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
tween the self-reported mean scores by healthy adolescents. However, the parent-proxy of high-myopes reported significant lower total (⫺4.4; 95% CI, ⫺8.4 to ⫺0.4; P ⫽ .03), psychosocial (⫺5.5; 95% CI, ⫺10.0 to ⫺1.0; P ⫽ .02), and social (⫺7.1; 95% CI, ⫺12.0 to ⫺2.1; P ⬍ .01) scores than parent-proxy of nonmyopes after adjustment. Slightly greater differences in these scores and school functioning were observed when only healthy adolescents were considered (total, ⫺5.6; 95% CI, ⫺10.3 to ⫺0.9; P ⫽ .02; psychosocial, ⫺7.0; 95% CI, ⫺12.4 to ⫺1.7; P ⬍ .01; social, ⫺7.2; 95% CI, ⫺13.0 to ⫺1.4; P ⫽ .01; school, ⫺7.5; 95% CI, ⫺13.5 to ⫺1.4; P ⫽ .02). Lower social score was also reported by parent-proxy of high-myopes when compared with low-myopes (all, ⫺4.7; 95% CI, ⫺8.3 to ⫺1.0; P ⫽ .01; healthy only, ⫺4.5; 95% CI, ⫺8.6 to ⫺0.5; P ⫽ .03). Parent-proxy of low-myopes only reported lower score than nonmyopes in school functioning (⫺4.6; 95% CI, ⫺9.2 to ⫺0.1; P ⫽ .047). No significant difference was observed between hyperopes and nonhyperopes in total (0.8; 95% CI, ⫺1.8 to 3.5; P ⫽ .54) and other self-reported mean scores (physical, P ⫽ .98; psychosocial, P ⫽ .42). Adolescents with astigmatism had self-reported (total, P ⫽ .92; physical, P ⫽ .42; psychosocial, P ⫽ .81) scores that were comparable to their counterparts. Except for higher emotional scores reported by parent-proxy of hyperopes (5.6; 95% CI, 0.6 to 10.6; P ⫽ .03), nonsignificant different results were found for other parent-proxy reported scores and when only healthy adolescents were analyzed. Most of the effect size concordance between child self-report and parent-proxy report in BEVI and each refractive group were small; all ICC values ⬍ 0.07 (data not shown). The Figure shows a comparison of total scores reported by child and parent-proxy. The mean difference was 0.8 (95% CI, ⫺0.03 to 1.63). The limits of agreement ranged from ⫺25.2 to 26.8, corresponding to two points on the Likert scales. Twenty-nine (3.1%) child-parent dyads lie above the upper limits (ie, the child rated at least one-point higher than their parent-proxy) while 26 (2.7%) child-parent dyads had child rated at least one point lower. The majority (71%) of the child-parent dyads had a difference within one point on the Likert scale (25 points).
DISCUSSION THIS STUDY IS UNIQUE IN THAT IT EVALUATES THE IMPACT
of presenting visual impairment and refractive errors in an Asian population of adolescent school pupils. The prevalence of presenting visual impairment of ⬎ 0.3 logMAR in the better eye amongst 1,249 adolescents with a mean age of 13.8 years was 5.7%. Adolescents with impairment reported lower total, psychosocial, and school scores, as measured by the PedsQL 4.0 scale. Thus, presenting visual QOL
OF
ADOLESCENTS
509
impairment ⬎ 0.3 logMAR was significantly associated with a decrement in their overall well-being, although these differences are below the minimal clinically meaningful differences (total, 4.36; psychosocial, 5.30; school, 9.12) suggested by Varni and associates.11 The most common cause of low presenting visual impairment in our study, and in others, is under-corrected refractive errors, but we found no association between refractive error per se and QoL. Nevertheless, presenting visual impairment may affect school learning, outdoor activity, and social life. This may compromise an adolescent’s school achievement and thus a lower school functioning score is reported. Our study also showed a deleterious effect on psychosocial functioning of adolescents with visual impairment. However, much lower PedsQL scores have been reported by children with cancer (total, 72.2; physical, 71.8; psychosocial, 72.6) and chronically ill children (total, 77.2; physical, 77.4; psychosocial 77.1).9,12 Overall, we found similar QoL scores for adolescents with and without refractive errors. Unlike cataract and glaucoma, which usually require surgical intervention, refractive errors are corrected using spectacles or contact lenses. Thus, most people with refractive error have good vision and show no reduction in their QoL. Although an adult study suggested that high-myopic patients had poorer vision-related QoL compared with moderate and mildmyopic patients, this result was not observed in our study.6 A possible explanation is that most complications associated with high-myopia, such as myopic macular degeneration, macular holes, retinal breaks, and tears, occur during the later years of life. In our study, there was a low prevalence of complications secondary to high myopia. The concordance in QoL measures between adolescents with BEVI or refractive errors and their parent-proxy were small. The result of a Bland-Altman plot also showed that there was considerable difference between scores reported by the child and their parent-proxy (ie, difference ranges within two points on the Likert scale). The lack of correlation between the parent-proxy and child’s report may be attributable to differences in perception of the teenager’s QOL, especially by parents who are working full-time. Hence, parent-proxy report should be included to complement child-self report in pediatric population with BEVI or refractive errors. Unlike the various instruments for the measurement of vision-related QoL in adults, such as Vision-Related QoL
Core Measure and NEI-VFQ-25, PedsQL, as a generic instrument, may not be sensitive enough to detect subtle effects attributable to the ocular conditions. Nevertheless, this well validated (Cronbach alphas reported were all ⬎ 0.709,12–16) and widely used instrument that allows for comparisons of QoL in the pediatric population across a variety of medical conditions such as asthma, cancer, diabetes, epilepsy, inflammatory bowel disease, headaches, obesity, pediatric rheumatology, cardiology, orthopedics, and psychiatrics warranted its appropriateness for the purpose of this study. Furthermore, with its parallel parentproxy report, we were able to investigate the relationship between child and parent-proxy ratings of their QoL. Although the multiplicity of potential endpoints arises with PedsQL, as with other QoL instruments, we have focused on the total score as the principal endpoint and regarded the other scores as secondary to this. In addition, multiple statistical testing can create problems in the interpretation of the results and may inflate the number of statistically significant differences found. In our case, there was a paucity of statistically significant differences (using the conventional P ⬍ .05), so this is likely to have little impact on the conclusions we draw. A limitation of our study is that the best-corrected VA was not assessed and thus visual impairment attributable to uncorrected refractive error (differences of habitual and best-corrected VA) cannot be determined. Other causes of visual impairment such as congenital glaucoma and cataract were also not assessed; however, these conditions are likely to be rare. Our study was also limited by the differences found in child’s age, gender, and ethnicity between the respondents and nonrespondent of the parent-proxy report. This study is the first Asian study to investigate the impact of BEVI and refractive errors on QoL using the PedQoL instrument in the pediatric population, and to evaluate the concordance between child self-report and parent-proxy report of QoL with a well validated measure. Other strengths of the study include a high participation rate of 80% for the child self-report and 61% for the parent-proxy report. In summary, we show in this population of healthy adolescents that BEVI was statistically though not clinically associated with impaired total, psychosocial, and school functioning QoL scores. Thus, there is a need to understand and address visual impairment at a young age.
THIS STUDY WAS SUPPORTED BY THE NATIONAL MEDICAL RESEARCH COUNCIL (NMRC/0975/2005), SINGAPORE, REPUBLIC OF Singapore and a Grant from the Singapore Children Society (RNO/059/06), Singapore, Republic of Singapore. The authors indicate no financial conflict of interest. Involved in design and conduct of study (H.-B.W., S.-M.S., D.M., S.-B.T.); data collection and management (H.-B.W.); obtaining funding and providing resources (H.-B.W., S.M.S.); analysis and interpretation of data (H.-B.W., S.M.S., D.M, S.-B.T.); and manuscript preparation, critical revision, and final approval of manuscript (H.-B.W., S.M.S., D.M., S.-B.T., T.-Y.W.). Approval was granted by the Institutional Review Board, Singapore Eye Research Institute, and the study protocol adhered to the tenets of the Declaration of Helsinki. The adolescents signed an assent form and their parent-proxy’s written informed consent was obtained.
510
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009
REFERENCES 1. Pokharel GP, Negrel AD, Munoz SR, Ellwein LB. Refractive error study in children: results from Mechi Zone, Nepal. Am J Ophthalmol 2000;129:436 – 444. 2. Zhao J, Pan X, Sui R, Munoz SR, Sperduto RD, Ellwein LB. Refractive error study in children: results from Shunyi District, China. Am J Ophthalmol 2000;129:427– 435. 3. Maul E, Barroso S, Munoz SR, Sperduto RD, Ellwein LB. Refractive error study in children: results from La Florida, Chile. Am J Ophthalmol 2000;129:445– 454. 4. Tsai SY, Chi LY, Cheng CY, Hsu WM, Liu JH, Chou P. The impact of visual impairment and use of eye services on health-related quality of life among the elderly in Taiwan: the Shihpai Eye Study. Qual Life Res 2004;13:1415–1424. 5. Nirmalan PK, Tielsch JM, Katz J, et al. Relationship between vision impairment and eye disease to vision-specific quality of life and function in rural India: the Aravind Comprehensive Eye Survey. Invest Ophthalmol Vis Sci 2005;46:2308 – 2312. 6. Rose K, Harper R, Tromans C, et al. Quality of life in myopia. Br J Ophthalmol 2000;84:1031–1034. 7. Saw SM, Hong CY, Chia KS, Stone R, Tan D. Nearwork and myopia in young children. Lancet 2001;357:390. 8. Saw SM, Chua WH, Hong CY, et al. Nearwork in early onset myopia. Invest Ophthalmol Vis Sci 2002;43:332–339. 9. Varni JW, Seid M, Rode CA. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care 2001;39:800 – 812. 10. World Health Organization. Constitution of the World Health Organization, basic document. Geneva, Switzerland: World Health Organization, 1948:1. 11. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr 2003;3:329 –341.
VOL. 147, NO. 3
VISUAL IMPAIRMENT
AND
12. Varni JW, Burwinkle TM, Katz ER, Meeske K, Dickinson P. The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer 2002;94:2090 –2106. 13. Powers SW, Patton SR, Hommel KA, Hershey AD. Quality of life in childhood migraines: clinical impact and comparison to other chronic illnesses. Pediatrics 2003;112:e1– e5. 14. Upton P, Eiser C, Cheung I, et al. Measurement properties of the United Kingdom-English version of the Pediatric Quality of Life Inventory 4.0 (PedsQL) generic core scales. Health Qual Life Outcomes 2005;3:22. 15. Felder-Puig R, Frey E, Proksch K, Varni JW, Gadner H, Topf R. Validation of the German version of the Pediatric Quality of Life Inventory (PedsQL) in childhood cancer patients off treatment and children with epilepsy. Qual Life Res 2004; 13:223–234. 16. Bastiaansen D, Koot HM, Bongers IL, Varni JW, Verhulst FC. Measuring quality of life in children referred for psychiatric problems: psychometric properties of the PedsQL 4.0 generic core scales. Qual Life Res 2004;13:489 – 495. 17. Varni JW. Linguistic validation of the PedsQLTM – a Quality of Life Questionnaire 2002. Available at http:// www.pedsql.org/PedsQL-Linguistic-Validation-Guidelines.doc. Accessed: September 3, 2005. 18. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320:1240 –1243. 19. Fishbaugh J. Look who’s driving now–visual standards for driver licensing in the United States. Insight 1995;20:11–20. 20. Perneger TV. What’s wrong with Bonferroni adjustments. BMJ 1998;316:1236 –1238. 21. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310.
QOL
OF
ADOLESCENTS
511
Biosketch Hwee-Bee Wong, MSc, graduated from University of Malaya Malaysia with a BSc in Statistics and a MSc in Statistics from National University of Singapore, Singapore. She is currently pursuing her part-time PhD at Department of Community, Occupational and Family Medicine, Yong Loo Lin School of Medicine at National University of Singapore and working as a senior biostatistician in the Ministry of Health Singapore. Dr Wong’s area of research interest are epidemiology of pediatrics myopia and quality of life.
511.e1
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
MARCH 2009