Free
Clinical and Epidemiologic Research  |   June 2014
Development of a Chinese Version of the Ocular Comfort Index
Author Affiliations & Notes
  • Cecilia Chao
    The School of Optometry and Vision Science, University of New South Wales, Sydney, New South Wales, Australia
  • Blanka Golebiowski
    The School of Optometry and Vision Science, University of New South Wales, Sydney, New South Wales, Australia
  • Yu Cui
    Brien Holden Vision Institute, University of New South Wales, Sydney, New South Wales, Australia
  • Fiona Stapleton
    The School of Optometry and Vision Science, University of New South Wales, Sydney, New South Wales, Australia
  • Correspondence: Fiona Stapleton, School of Optometry and Vision Science, University of New South Wales, Sydney, NSW 2052, Australia; f.stapleton@unsw.edu.au
Investigative Ophthalmology & Visual Science June 2014, Vol.55, 3562-3571. doi:10.1167/iovs.14-14276
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Cecilia Chao, Blanka Golebiowski, Yu Cui, Fiona Stapleton; Development of a Chinese Version of the Ocular Comfort Index. Invest. Ophthalmol. Vis. Sci. 2014;55(6):3562-3571. doi: 10.1167/iovs.14-14276.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: Dry eye is common in Asian populations, but there are limited validated instruments available to assess ocular symptoms within these populations. This study aimed to develop and assess the psychometric properties of a Chinese version of the ocular comfort index (OCI-C).

Methods.: The OCI was translated and back translated by 32 bilingual volunteers and a preliminary OCI-C was produced for analysis; 165 bilingual volunteers completed both the OCI and the preliminary OCI-C. The item scores for preliminary OCI-C were compared with the OCI using Cronbach's α. Repeatability of the total score of OCI-C was evaluated in a subgroup of 20 participants after a week by calculating the coefficient of repeatability (CoR) and intraclass correlation coefficient (ICC) for the OCI-C score; 322 subjects, including 81 dry eye subjects, completed the OCI-C to verify the sensitivity in dry eye diagnosis. Rasch analysis was used to estimate interval measures from ordinal Likert score and assess psychometric property of item-fit, category function, targeting of items to subjects, and person separation reliability.

Results.: Cronbach's α for all items was above 0.85. The CoR was ±5.84 and the ICC was 0.96 (95% confidence interval 0.91–0.98). The total score of OCI-C was able to differentiate dry eye (40.0) with non–dry eye (30.2) (P < 0.0001). In the Rasch analysis model, the OCI-C showed robust psychometric properties for item-fit and category calibration: person-separation reliability: 2.47 and item reliability: 8.42.

Conclusions.: The psychometric properties of the OCI-C are comparable with the OCI and the instrument is appropriate as a diagnostic tool and to determine the severity of dry eye in this population.

Introduction
The Dry Eye WorkShop 1 has defined dry eye as “a multifactorial disease of the tears and ocular surface that results in symptoms of discomfort, visual disturbance, and tear film instability.” Dry eye is more commonly reported in females, 2,3 in an aging population, 3,4 in contact lens wearers, 5 in those who have undergone refractive surgery, 615 and is associated with autoimmune disease, and use of certain medications. 4 Associations were found between symptom results from validated questionnaires and diagnoses of dry eye. 1621 The subjective symptomatology questionnaire is an important tool for dry eye disease diagnosis used in conjunction with clinical tests. Approximately 83% of practitioners apply questionnaires as part of the diagnosis of dry eye disease. 22 Hence, a validated questionnaire allows practitioners to diagnose and determine the severity of dry eye to manage and monitor the condition. 
The Ocular Comfort Index (OCI), which enables a quick assessment of ocular comfort using simple English, was developed by Johnson and Murphy in 2007. 23 The OCI was designed using items from the literature and from focus groups and uses a continuous interval scale (0–100) as developed by Rasch analysis. It has been used in monitoring the effect of topical treatment on dry eye symptoms 23,24 and in assessing contact lens-induced dry eye. 25,26 The questionnaire contains 12 items with six typical symptoms (dryness, grittiness, stinging, tiredness, pain, and itchiness), and the subquestion for each symptom probes the frequency and intensity of the impact of dry eye and each item is scored on a 0 to 6 scale: 0 (no symptoms) to 6 (most frequent/most severe in the past week). The total score is presented as a linear continuous interval scale validated by Rasch analysis. 23 As a result, lower scores indicate a better ocular comfort, whereas higher scores represent poorer ocular comfort. Good reliability of the OCI has been reported. 23  
In Chinese populations, the frequency of dry eye is higher than in Caucasians, with more than one in three individuals reporting dry eye. 5,2729 However, limited valid questionnaires exist in Chinese, except for a translated version of the ocular surface disease index (OSDI); however, no validation data have been published. 30 Given this lack of validated instruments, this study aims to develop and characterize the OCI in Chinese. 
Methods
Overview
The overview of this study is presented in Figure 1. Each sequential step is described in detail below. For preliminary item reliability and sensitivity, analysis was performed using SPSS 19.0 statistical software (IBM, Chicago, IL, USA) and Rasch analysis was performed using Winsteps (Version 3.78; Winsteps, Chicago, IL, USA, rating scale model). 
Figure 1
 
The framework of the methodology of the development of the OCI-C.
Figure 1
 
The framework of the methodology of the development of the OCI-C.
Forward Translation
The OCI instrument was translated by bilingual volunteers fluent in both English and Chinese. Three bilingual non–eye care practitioners, including one qualified translator, independently translated the OCI, including instruction, item content, and grading options. In addition, seven bilingual optometrists and ophthalmologists, and five bilingual non–eye professionals forward translated the six descriptors of dry eye symptoms in the OCI into Chinese. They were told to place emphasis on the concepts expressed. 
Synthesis of Translations
Since the translations of instruction and grading options were in agreement among the three translators, a template of the instrument was constructed. There was less agreement for certain descriptors, specifically stinging and grittiness. Consequently, six descriptors in Chinese of stinging and grittiness, three for each symptom, were provided in the provisional translation for the backward translation process. 
Backward Translations
All descriptors in Chinese were reviewed by a further group of 10 bilingual optometrists and ophthalmologists, and 10 bilingual non–eye professionals, who were masked from the original OCI, for backward translation independently. 
Translation Quality Evaluation
The forward and backward translations were then given to an Australian bilingual optometrist to evaluate the quality of the translation to ensure appropriate wording in Chinese (Fig. 2). 31 The bilingual optometrist agreed that grittiness and stinging were difficult to interpret in Chinese and three alternative translations for these two descriptors were proposed and formed the basis of the 20 items of preliminary version OCI in Chinese (preliminary OCI-C). 
Figure 2
 
The items in the original OCI are presented in blue boxes. Translated Chinese terms are the results of the translation process and were used on the preliminary OCI-C.
Figure 2
 
The items in the original OCI are presented in blue boxes. Translated Chinese terms are the results of the translation process and were used on the preliminary OCI-C.
Subjects
Participants who were 18 years or older and proficient in understanding English and Chinese in writing and speaking (bilingual group), were invited to complete both OCI and the preliminary OCI-C online between April and July 2012. Another group of participants, who speak and write in Chinese (Chinese group), were also invited to fill in the OCI-C only online between July 2012 and February 2013. All eligible participants gave written consent before the study. This study followed the tenets of the Declaration of Helsinki and ethics approval was obtained from the University of New South Wales Human Research Ethics Advisory Panel. To understand the risk of dry eye and to determine dry eye status, screening questions, including gender, age, contact lens history, and the Women's Health Questionnaire (WHQ), 32 were delivered before completion of the OCIs. The WHQ is used to identify moderate to severe dry eye in the general population, 26 and was used to differentiate dry eye from non–dry eye subjects in this study. 
Optimization of Descriptors and Reliability of the OCI-C
A total of 165 subjects completed both OCI and the preliminary OCI-C. The optimization of descriptors, which determined the most appropriate descriptor related to grittiness and stinging in Chinese, were conducted to form the OCI-C. The item scores for the preliminary OCI-C were compared with the OCI by establishing the interitem correlation coefficient. 33,34 A Bland and Altman plot includes bias (mean difference), coefficient of repeatability (CoR = 1.96 × within-subjects SD) and limits of agreement (LoA = bias ± CoR). 35 Ladder plots were used to provide additional information about the distribution between the items' scores of the OCI and the preliminary OCI-C. 
The reliability of the OCI-C from the bilingual participants was assessed by the comparison between the OCI-C items and the OCI using Cronbach's α coefficient. A coefficient of 0.7 is considered acceptable and more than 0.8 indicates good reliability. 34,36 To evaluate the construct validity of the OCI-C, items' scores and the total score of the OCI-C calculated using the OCI-C calculator (developed from the Rasch model in this study) were compared with the OCI items' scores and total score calculated using the OCI calculator using the paired t-test. 33 Additionally, Bland and Altman analysis compared the difference between the two questionnaires, and floor (score of 0)/ceiling (score of 100) effects for total score were checked. 
Repeatability of the OCI-C
To evaluate test–retest reliability, 37 a subgroup of 20 participants were asked to complete the OCI-C again after a week. The OCI-C was sent via e-mail to the participants who agreed to complete the questionnaire again. The questionnaire also was completed online. Coefficient of repeatability was calculated and Bland and Altman analysis also was used to examine repeatability. Sample variability was measured using intraclass correlation coefficient (ICC). Differences between items were assessed with a Wilcoxon Signed Ranks test. 
The Measurement of Sensitivity of the OCI-C
There were 157 additional Chinese participants who completed the OCI-C online, giving a total of 322 respondents, comprised of 81 dry eye, diagnosed using the WHQ, 32 and 241 non–dry eye participants, for the assessment of the sensitivity of the OCI-C in diagnosing dry eye. Also, the item scores and total score between full-time contact lens and non–contact lens groups was analyzed by using either the Student's t-test or Mann-Whitney U test, as appropriate. 34 Floor/ceiling effects for the total score also were checked. 
Rasch Analysis of the OCI-C and the OCI-C Calculator Development
Rasch analysis was applied to the 322 responses to assess overall fit of the model (demonstration of unidimensionality), rating responses (category functions as used in the questionnaire), item and person fit statistics (the degree of difference between the observed response and the expected response to each item from each participant), and person separation (an indicator of test reliability) based on the natural logarithm of log-odd unit (logits). To produce a more comparable questionnaire, a linear 0 to 100 scale (total score) was transformed from the raw score of the OCI-C items using a Rasch model in this population. 
Results
Optimization of Descriptors
A total of 165 bilingual participants, with a mean age of 27 ± 8 years (range, 19–90), including 79 males, 38 contact lens wearers, and 29 participants with dry eye, were eligible to complete both the OCI and the preliminary OCI-C. 
For determination of grittiness in Chinese, GC1 (rough eyelid sensation) was selected because of the highest interitem correlation found in both frequency and intensity of grittiness with these items from the OCI, with the narrowest range of LoA. Further, the smallest differences in the scores in both frequency and intensity between grittiness in English and GC1 were found comparing with other descriptors of grittiness (Table 1). 
Table 1
 
The Optimization of Descriptors
Table 1
 
The Optimization of Descriptors
English Items Mean ± SD Chinese Items Bias LoA Mean ± SD Interitem Correlation
Frequency of grittiness 2.2 ± 0.2 FGC1 0.01 −1.4, 1.5 2.1 ± 0.2 0.84
FGC2 0.23 −1.4, 1.9 1.9 ± 0.2 0.78
FGC3 0.05 −2.1, 2.2 2.1 ± 0.2 0.63
Intensity of grittiness 2.2 ± 0.2 IGC1 −0.01 −1.4, 1.5 2.2 ± 0.2 0.83
IGC2 0.18 −1.8, 2.2 2.0 ± 0.2 0.67
IGC3 0 −2.2, 2.2 2.1 ± 0.2 0.60
Frequency of stinging 2.0 ± 0.2 FSC1 0.18 −2.1, 2.5 1.9 ± 0.2 0.55
FSC2 0.31 −2.1, 2.7 1.7 ± 0.2 0.54
FSC3 0.40 −1.7, 2.5 1.6 ± 0.1 0.59
Intensity of stinging 2.1 ± 0.2 ISC1 0.13 −2.3, 2.5 2.0 ± 0.2 0.58
ISC2 0.29 −2.1, 2.7 1.8 ± 0.2 0.55
ISC3 0.40 −1.9, 2.7 1.7 ± 0.2 0.53
In terms of stinging, even though the bias of SC1 (sharp pain) with stinging was the closest to 0, the interitem correlation between the frequency and the intensity of stinging in Chinese were not consistent. However, the differences in the scores of frequency and intensity between stinging and SC1 were the smallest comparing with other descriptors of stinging in Chinese (Table 1). Therefore, SC1 was selected from the preliminary OCI-C to form the next iteration of the questionnaire (OCI-C). 
The Reliability of the OCI-C
The reliability of the scores of all items in the OCI-C was satisfactory with the Cronbach's α range from 0.71 to 0.96 (Table 2). The reliability indices were also supported by both ladder plot and Bland and Altman plots (data not shown). The LoA was less than ±2, except for stinging (Table 2). The mean scores of the OCI and OCI-C were generally consistent. However, there were higher scores in frequency and intensity of tiredness and frequency of itching in the OCI than in OCI-C, and higher score of frequency of pain in the OCI-C than in the OCI (Table 2). 
Table 2
 
The Overall Results of Reliability
Table 2
 
The Overall Results of Reliability
English Items Mean Score of Items ± SD Chinese Items Mean Score of Items ± SD Cronbach's α P Value Bias LoA
Frequency of dryness 2.8 ± 0.2 FDC 2.9 ± 0.3 0.96 NS −0.06 −1.3, 1.2
Intensity of dryness 2.8 ± 0.2 IDC 2.8 ± 0.2 0.92 NS 0.01 −1.4, 1.5
Frequency of grittiness 2.2 ± 0.2 FGC1 2.1 ± 0.2 0.91 NS 0.01 −1.4, 1.5
Intensity of grittiness 2.2 ± 0.2 IGC1 2.2 ± 0.2 0.91 NS −0.01 −1.4, 1.5
Frequency of stinging 2.0 ± 0.2 FSC1 1.9 ± 0.2 0.71 NS 0.18 −2.1, 2.5
Intensity of stinging 2.1 ± 0.2 ISC1 2.0 ± 0.2 0.73 NS 0.13 −2.3, 2.5
Frequency of tiredness 3.9 ± 0.2 FTC 3.8 ± 0.2 0.95 0.02 0.13 −1.2, 1.5
Intensity of tiredness 3.9 ± 0.3 ITC 3.7 ± 0.2 0.92 0.002 0.20 −1.4, 1.8
Frequency of pain 1.8 ± 0.2 FPC 1.9 ± 0.2 0.92 0.01 −0.13 −1.4, 1.1
Intensity of pain 1.9 ± 0.2 IPC 1.9 ± 0.2 0.95 NS −0.07 −1.1, 1.0
Frequency of itching 2.8 ± 0.2 FIC 2.7 ± 0.2 0.94 0.01 0.15 −1.3, 1.6
Intensity of itching 2.8 ± 0.2 IIC 2.7 ± 0.2 0.93 NS 0.09 −1.3, 1.5
For the floor/ceiling effect in the total score of the OCI and the OCI-C, 3% (5/165) in the OCI and 3.6% (6/165) in OCI-C were floor responses. In addition, 0.6% (1/165) in both the OCI and the OCI-C were ceiling responses. The mean difference between two total scores was 2.4, indicating that the total score of the OCI-C calculated using the OCI-C calculator (30.9 ± 13.1) is lower than the OCI using the OCI calculator (33.4 ± 9.4) (P < 0.0001) with the inclusion of floor/celling data. The CoR was ±12.7 with the LoA between 10.3 and −15.2 in a 0 to 100 scale (Fig. 3). 
Figure 3
 
The difference between the total score of OCI-C and the OCI plotted against their mean. The dashed line represents a bias of −2.4. The yellow and green lines represent the limits of agreement of +10.3 to −15.2.
Figure 3
 
The difference between the total score of OCI-C and the OCI plotted against their mean. The dashed line represents a bias of −2.4. The yellow and green lines represent the limits of agreement of +10.3 to −15.2.
The Repeatability of the OCI-C
A subgroup of 20 participants, including four males, three contact lens wearers, and five subjects with dry eye, repeated the OCI-C a week later. There were no significant differences in the intra-item and total scores between the two repeats of the OCI-C (all P > 0.05). The 95% confidence interval for two-way random effects ICC of the two OCI-C repeats was 0.91 to 0.98. The median (interquartile range [IQR]) of items, total score, and CoR are reported in Table 3. The total scores for the two repeats for each participant are shown in Figure 4. The mean difference in total score of the OCI-C between two repeats was 0.5 of 100. 
Figure 4
 
The difference between the total score of OCI-C repeat measurements 1 and 2 plotted against their mean. The dashed line represents a bias of −0.5. The yellow and green lines represent the limits of agreement of +6.4 to −5.3.
Figure 4
 
The difference between the total score of OCI-C repeat measurements 1 and 2 plotted against their mean. The dashed line represents a bias of −0.5. The yellow and green lines represent the limits of agreement of +6.4 to −5.3.
Table 3
 
The Repeatability of the OCI-C
Table 3
 
The Repeatability of the OCI-C
Items Median of the Score for the First Repeat (IQR) Median of the Score for the Second Repeat (IQR) CoR
FDC 3.0 (2.0) 2.8 (1.4) 1.72
IDC 2.0 (2.0) 2.9 (1.4) 1.80
FGC1 2.0 (1.8) 1.5 (0.9) 2.02
IGC1 2.0 (2.0) 1.8 (1.1) 2.26
FSC1 1.0 (0.0) 1.0 (0.8) 1.19
ISC1 1.0 (0.8) 1.8 (1.1) 2.67
FTC 4.0 (2.5) 3.6 (1.8) 2.25
ITC 4.0 (1.8) 3.7 (1.8) 2.06
FPC 1.0 (0.0) 1.5 (1.0) 1.91
IPC 1.0 (0.0) 1.9 (1.3) 2.58
FIC 2.0 (2.0) 2.5 (1.4) 2.33
IIC 2.0 (3.0) 2.7 (1.3) 1.85
Total score 27.9 (9.7) 28.1 (11.5) 5.84
Sensitivity of the OCI-C
There were 322 responses, including 145 males, 56 contact lens wearers, and 81 subjects with dry eye. The mean age of the subjects was 26 ± 8 years. 
Table 4 indicates that the scores of all items were significantly higher in dry eye than non–dry eye subjects (P < 0.001) except for the Intensity of Itching in Chinese (P = 0.72). Dry eye subjects who were not contact lens wearers reported higher scores in dryness and tiredness (P < 0.05). However, there were no differences in the scores of the items between contact lens and non–contact lens wearers in the non–dry eye group. 
Table 4
 
Median (IQR) Scores of the 12 Questions Between Dry Eye and Non–Dry Eye
Table 4
 
Median (IQR) Scores of the 12 Questions Between Dry Eye and Non–Dry Eye
Dry Eye, n = 81 Non–Dry Eye, n = 241 Dry Eye vs. Non–Dry Eye
All CL, n = 22 NCL, n = 59 P Value All CL, n = 34 NCL, n = 207 P Value P Value
FDC 4.2 (1.4) 3.6 (1.3) 4.4 (1.4) 0.03 2.7 (1.4) 2.9 (1.3) 2.6 (1.4) NS <0.0001
IDC 4.1 (1.3) 3.5 (1.2) 4.2 (1.2) 0.02 2.8 (1.4) 2.9 (1.2) 2.7 (1.3) NS <0.0001
FGC1 3.3 (1.4) 3.5 (1.4) 3.2 (1.4) NS 2.1 (1.3) 2.3 (1.1) 2.0 (1.3) NS <0.0001
IGC1 3.3 (1.4) 3.5 (1.4) 3.2 (1.4) NS 2.1 (1.3) 2.4 (1.2) 2.1 (1.3) NS <0.0001
FSC1 2.7 (1.5) 3.2 (1.7) 2.5 (1.5) NS 2.0 (1.3) 1.9 (1.1) 2.1 (1.4) NS 0.004
ISC1 2.8 (1.5) 3.1 (1.6) 2.6 (1.5) NS 2.2 (1.5) 2.0 (1.2) 2.3 (1.5) NS <0.0001
FTC 4.5 (1.5) 3.9 (1.6) 4.6 (1.4) NS 3.5 (1.5) 3.5 (1.3) 3.5 (1.6) NS <0.0001
ITC 4.3 (1.4) 3.7 (1.3) 4.5 (1.3) 0.04 3.5 (1.4) 3.3 (1.1) 3.5 (1.4) NS <0.0001
FPC 2.7 (1.5) 3.1 (1.5) 2.6 (1.5) NS 2.1 (1.4) 2.3 (1.4) 2.1 (1.4) NS <0.0001
IPC 2.7 (1.5) 3.0 (1.7) 2.6 (1.5) NS 2.1 (1.4) 2.1 (1.3) 2.3 (1.4) NS 0.002
FIC 3.2 (1.7) 3.3 (1.8) 3.2 (1.6) NS 2.7 (1.5) 2.7 (1.4) 2.7 (1.5) NS 0.01
IIC 3.1 (1.6) 3.3 (1.8) 2.6 (1.5) NS 2.8 (1.5) 2.9 (1.6) 2.7 (1.4) NS 0.72
A significant difference was found in the total score of the OCI-C between dry eye and non–dry eye subjects (P < 0.001). The median of the total score (IQR) in the dry eye group was 40.8 (13.9) and the median of the total score (IQR) in the non–dry eye group was 30.2 (13.2). The calculation above included 10 (3.1%) floor responses and 1 (0.3%) ceiling response. There were no significant differences in total score between contact lens and non–contact lens wearers and score was not affected by sex (Table 5). There was no association between age and the total score of the OCI-C overall or within either the dry eye or non–dry eye population. 
Table 5
 
Comparison of Mean Total Scores Between Dry Eye and Non–Dry Eye Subjects
Table 5
 
Comparison of Mean Total Scores Between Dry Eye and Non–Dry Eye Subjects
n Dry Eye, n = 81 Non–Dry Eye, n = 241
CL, n = 22 NCL, n = 59 CL, n = 34 NCL, n = 207
Total score ± SD 40.0 ± 23.5 41.0 ± 15.0 30.0 ± 14.0 27.0 ± 21.0
Sex (n) Male (36) Female (45) Male (109) Female (132)
Total score ± SD 43.1 ± 14.7 39.0 ± 13.2 30.2 ± 13.7 30.2 ± 12.9
Rasch Analysis of the OCI-C
Preliminary Analysis.
The fit statistics of the items based on the exploratory analysis indicated that all items fitted with a single underlying construct (fit statistics lies in 0.7–1.3), which holistically describes the ocular comfort in the Chinese subjects recruited in the present study (Table 6; Fig. 5). 
Figure 5
 
Whole fit statistics layout with infit versus outfit scales as ZSTDs. Red box represents ±2 ZSTD.
Figure 5
 
Whole fit statistics layout with infit versus outfit scales as ZSTDs. Red box represents ±2 ZSTD.
Table 6
 
Fit Statistics for the 12 Items in 322 Subjects
Table 6
 
Fit Statistics for the 12 Items in 322 Subjects
Items in Chinese Infit Outfit
MNSQ ZSTD MNSQ ZSTD
1. FDC - In the past week, how often did your eyes feel dry? 1.19 2.3 1.15 1.8
2. IDC - When your eyes felt dry, typically, how intense was the dryness? 0.88 −1.5 0.9 −1.2
3. FGC1 - In the past week, how often did your eyes feel gritty? 0.88 −1.5 0.80 −2.2
4. IGC1 - When your eyes felt gritty, typically, how intense was the grittiness? 0.80 −2.5 0.75 −2.8
5. FSC1 - In the past week, how often did your eyes feel stingy? 0.97 −0.4 0.82 −1.8
6. ISC1 - When your eyes stung, typically, how intense was the stinging? 1.12 1.4 0.99 −0.1
7. FTC - In the past week, how often did your eyes feel tired? 1.12 1.5 1.12 1.5
8. ITC - When your eyes felt tired, typically, how intense was the tiredness? 0.95 −0.7 0.98 −0.3
9. FPC - In the past week, how often did your eyes feel painful? 0.93 −0.8 0.79 −2.2
10. IPC - When your eyes felt painful, typically, how intense was the pain? 0.96 −0.5 0.85 −1.5
11. FIC - In the past week, how often did your eyes feel itchy? 1.19 2.3 1.11 1.3
12. IIC - When your eyes felt itchy, typically, how intense was the itching? 1.18 2.1 1.24 2.6
Sixteen (5%) of 322 of the subjects had more than 3 SD units at z scores (ZSTD) persons infit from the expected value (Fig. 6). These 5% person responses increased the tendency of idiosyncratic response patterns. Therefore, the person responses of the participants were reviewed and found that they tended to have greater differences between item scores. Generally it is recommended such responses are removed for instrument optimization 38 ; however, the authors did not find a significant improvement in both item fit and item mapping when those responses were excluded (Table 7). Therefore, the outfit responses were included. 
Figure 6
 
Person information-weight fit (Person infit) expressed as ZSTD versus person measurement (n = 322).
Figure 6
 
Person information-weight fit (Person infit) expressed as ZSTD versus person measurement (n = 322).
Table 7
 
Estimates of Item Difficulties From the Primary (322 Participants) and Secondary (305 Participants) Rasch Analysis for the 12 Items in Chinese OCI-C and Persons
Table 7
 
Estimates of Item Difficulties From the Primary (322 Participants) and Secondary (305 Participants) Rasch Analysis for the 12 Items in Chinese OCI-C and Persons
Items in Chinese 322 Participants 305 Participants
Difficulty, Logits SEs Difficulty, Logits SEs
1. FDC −0.33 0.06 −0.33 0.06
2. IDC −0.36 0.06 −0.38 0.06
3. FGC1 0.41 0.06 0.41 0.06
4. IGC1 0.35 0.06 0.34 0.06
5. FSC1 0.61 0.06 0.67 0.07
6. ISC1 0.43 0.06 0.47 0.06
7. FTC −1.01 0.06 −1.08 0.06
8. ITC 0.95 0.06 −1.00 0.06
9. FPC 0.52 0.06 0.59 0.07
10. IPC 0.52 0.06 0.58 0.07
11. FIC −0.08 0.06 −0.12 0.06
12. IIC −0.11 0.06 −0.16 0.06
Category Structure.
The rating category measures are unimodal and sequentially ordered, indicating the subjects can differentiate categories and endorse preference for the categories as designed (Fig. 7A). The advances were relatively small between categories 5 to 7 (Table 8), so collapse of categories 5 to 7 can be attempted to achieve better person separation and item fit. However, the collapse of categories did not deliver higher person separation, reliability, and construct validity so the original category structure was retained. The targeting of OCI-C items was skewed toward subjects with potential symptoms (Fig. 7B). 
Figure 7
 
Probability curve showing seven-category rating scale (Fig. 7A) and the expected response (Fig. 7B). Intersection of the dashed vertical lines with x-axis represents measures relative to item difficulty.
Figure 7
 
Probability curve showing seven-category rating scale (Fig. 7A) and the expected response (Fig. 7B). Intersection of the dashed vertical lines with x-axis represents measures relative to item difficulty.
Table 8
 
Category Diagnostics for the 12 Items in OCI-C
Table 8
 
Category Diagnostics for the 12 Items in OCI-C
Score Frequency Infit MNSQ Outfit MNSQ OBSVD AVRGE Sample Expect ANDRICH Threshold Score Measure
0 2462 0.83 0.88 −2.71 −2.63 None (−3.20)
1 1388 0.82 0.68 −1.85 −1.85 −1.74 −1.84
2 1187 1.05 0.81 −0.97 −1.20 −1.37 −0.98
3 980 0.97 0.91 −0.54 −0.61 −0.71 −0.17
4 568 1.22 1.30 −0.15 −0.04 0.22 0.78
5 204 1.40 1.43 0.28 0.54 1.26 1.96
6 65 2.14 1.93 0.42 1.12 2.33 (3.64)
Item Calibration.
The fit statistics of most OCI-C items (11 of 12 items) performed well. Most of the items were in the zone of ±2 ZSTD or within 5% of ±2 ZSTD (Fig. 5). 23 The fit statistics of IGC1 were relatively low, which indicates the item is relatively predictable (item has less variation than expected) and may not be sensitive enough to modify variables, even though it is considered usable. The nature of the OCI-C questions was unidimensional, as the principal domain of the unrotated factor analysis was associated with each item in the range of 0.69 to 0.81. The estimates of item difficulty between with and without the 16 unfitted responses gave similar values, with less than 0.07 logits. 
Person and Item Estimates.
Figure 8 maps the person ability/item difficulty in a 0 to 100 scale (logit × 9.4881 + 35.9753, the mean of the item difficulties: 48.3529, the user-scaled value of logit: 8.5310). Also, the map indicates that the instrument was relatively easy for this population. Item separation was 8.42 and person separation was 2.47, which indicated that the instrument is sufficiently sensitive. 
Figure 8
 
Person ability/item difficulty map for the OCI-C. The person (represented to the left of the vertical line) and item (represented to the right side of the vertical line) appear in descending order of ability and difficulty on a common scale, which represents the logit value (0–100) of ocular comfort, as transformed from Rasch analysis. Those responses with higher score of ocular discomfort (more able people) and those items with lower scores (harder items) are on the top of the scale, whereas the responses with lower scores of discomfort and easier items are at the bottom. Each “#” represents two participants and “.” represents one. The scale is in log units. M indicates mean; S, 1 SD from the mean; T, 2 SD from the mean; F, frequency; I, intensity; DC, dryness in Chinese; TC, tiredness in Chinese; PC, pain in Chinese; IC, itching in Chinese.
Figure 8
 
Person ability/item difficulty map for the OCI-C. The person (represented to the left of the vertical line) and item (represented to the right side of the vertical line) appear in descending order of ability and difficulty on a common scale, which represents the logit value (0–100) of ocular comfort, as transformed from Rasch analysis. Those responses with higher score of ocular discomfort (more able people) and those items with lower scores (harder items) are on the top of the scale, whereas the responses with lower scores of discomfort and easier items are at the bottom. Each “#” represents two participants and “.” represents one. The scale is in log units. M indicates mean; S, 1 SD from the mean; T, 2 SD from the mean; F, frequency; I, intensity; DC, dryness in Chinese; TC, tiredness in Chinese; PC, pain in Chinese; IC, itching in Chinese.
Discussion
This article reports for the first time translation and validation of a Chinese version of the OCI questionnaire. Our approach has involved both Rasch analysis for instrument validation and conventional analysis to determine repeatability and sensitivity of the new instrument in this population. The new OCI-C appears to be comparable in psychometric properties to the original version and generalizable across Chinese-speaking populations. We believe this is the first validated Chinese language instrument for the diagnosis and treatment of dry eye disease. 
The OCI was selected as the instrument of choice, as the English version has undergone Rasch analysis; it has been shown to be reliable and repeatable and has been used in a range of dry eye studies in English-speaking populations. 24,26 Although the ocular surface disease index (OSDI) also uses a continuous internal scale to determine dry eye severity, it is difficult to obtain a linear relationship between OSDI assessment and degree of dry eye due to an inability to assess the difficulty of categories and items. 23 Johnson and Murphy 23 suggested that the response of “not applicable” is missing so as to achieve the maximum possible values. The reliability is relatively lower than the OCI, with the mean of 0.82 of ICC. 39 For floor and ceiling responses, approximately 10% of OSDI responses with mild-moderate dry eye scored 0, 39 whereas most of the responses in the OCI occupy the middle of the scale, with 3% responses scoring 0 only and no ceiling responses. 39 An ordinal ranking was applied to the data in a previous study to estimate the severity of dry eye. 23 Consequently, the OCI was chosen as the basis of the instrument development in this study. 
This instrument can be applied generally to a Chinese-speaking population as the strength of this approach is that this instrument was translated and back translated by 17 bilingual eye care professionals (four Chinese second generation in Australia, three from Hong Kong, two from Malaysia/Singapore, four from Taiwan, and four from China) and 20 nonprofessionals (two Chinese second generation in Australia, five from Hong Kong, two from Malaysia/Singapore, five from Taiwan, and six from China). Additionally, all translations were then evaluated by a masked Australian optometrist and the preliminary OCI-C was completed by 165 bilingual participants from Australia, Hong Kong, Macau, Taiwan, China, Singapore, and Malaysia. To achieve a higher correlation with the OCI, two-item descriptors were optimized via interitem correlation, Bland and Altman plot, and using ladder plots. The OCI items grittiness and stinging were difficult to interpret in Chinese and therefore multiple translation alternatives were evaluated in the preliminary OCI-C. “Rough eyelid sensation” and “sharp pain,” respectively, were the descriptors selected for use in the OCI-C because the scores between the original OCI and the preliminary OCI-C were the most consistent. The reliability of the OCI-C was found to be acceptable to excellent (Cronbach's α coefficient of 0.71–0.96). 34,36  
Differences in item scores between the OCI and the OCI-C, namely frequency of tiredness, pain, and itching (FTC, FPC, and FIC) and intensity of tiredness (ITC) were found in this study. However, the differences in item scores were small (less than 0.25 [of 5]) between the English and Chinese versions. The total score of the OCI (33.4) was higher than the OCI-C (30.9) in 165 bilingual participants. This difference may have occurred because of the use of different calculators, based on the individual Rasch model and that the models differ from population to population. To assess the difference between the total score of the OCI and the OCI-C, the authors have compared two total scores calculated by using the original OCI calculator only and found that the difference in the score is 0.7 of 100. The authors have further compared the differences in the groups' total scores between two calculators. Significant differences in total score only can be found in non–dry eye participants, with a higher score (2.4 units) using the OCI calculator (Fig. 3). The differences are likely due to the wider spread of the OCI-C calculator model. 
The repeatability in this study was superior to the original OCI; with the ICC value in this study was closer to 1. Findings from the original OCI questionnaire showed that the 95% confidence intervals for the two-way random-effects ICC was 0.81 to 0.91 in total score, 23 whereas for the new instrument it was 0.91 to 0.98. Further, the CoR of the OCI-C total score (5.8) was smaller than in the OCI (13.1) in a 0 to 100 scale. The repeatability of the OCI-C was also evaluated using a Bland and Altman plot. The difference was 0.5 units in a 0 to 100 scale between two repeats (Fig. 4). 
As expected, item score and total score of the OCI-C were able to differentiate between dry eye and non–dry eye groups except for the intensity of itching. The finding for itching may reflect that itch is more likely to be due to allergy rather than ocular dryness. 40,41  
Higher scores of dryness and tiredness in Chinese were reported from non–contact lens wearers who had dry eye compared with contact lens wearers who had dry eye. Also, no difference in total score was found between non–dry eye contact lens and non–lens wearers, which contrasts with a previous report. 5 This study defined contact lens wear as full-time wear of more than 6 hours per day, 5 days per week, whereas part-time or occasional contact lens wearers were grouped as non–lens wearers. Therefore, the authors believe that this may be because those intolerant contact lens wearers (who may be occasional contact lens wearers) may have been classified as non–contact lens wearers and, consequently, significantly higher item scores and slightly higher total score of symptoms were found in non–contact lens wearers who had dry eye. Further investigation is required to assess the OCI-C in a contact lens population with consideration of lens wear tolerance. 
The percentage of person responses with more than ±3 ZSTD was similar to that reported in a previous study (4.9%). 3 Unlike described in the original OCI development, 3 these responses were not removed, because the values of fit statistics and item difficulty of 12 items were similar irrespective of whether the misfitting respondents were included or not. 
In the category structure analysis, the frequency of the score was highest at category 1 (score 0) and lowest at category 7 (score 6), which is similar to the previous findings. 23 Since the frequency of the responses was small in categories 5 to 7, collapse of categories was attempted but this did not result in higher person separation. A similar finding was reported for the OCI, where the collapse of categories reduced person separation and item fit. 38  
The item IGC1 (intensity of grittiness in Chinese–rough eyelid sensation) was included in the OCI-C even though it performed poorly in the fit statistics. Figure 7 shows that most of the OCI-C items fit the Rasch model well except for IGC1. However, the item associations between the principal domain and items, including IGC1, were higher than the OCI (range from 0.63–0.79). 23 It is argued that IGC1 is therefore unlikely to corrupt the measurement and was hence included. 
The person-item map (Fig. 8) indicates good targeting between item difficulty and person ability. Also, the person separation in this study was beyond the acceptable value of more than 2. Therefore, this instrument fits well with this target population. 
The limitation of this study is that the relationship with clinical indicators of dry eye or other available dry eye questionnaires was not evaluated. 1619,23 Johnson and Murphy 23 reported an association between the OCI and log10 tear breakup time. Also, the OCI is associated with corneal nerve tortuosity. 26 Future studies should determine associations between the OCI-C and dry eye signs. 
The psychometric properties of the OCI-C are robust and comparable with the original English version. It offers superior reliability and repeatability compared with the OCI. In addition, good person and item separation were reported from Rasch analysis. The OCI-C is sufficiently sensitive to identify dry eye sufferers and it is expected that the OCI-C can be used in Chinese-speaking populations to assess ocular discomfort. Copies of the questionnaire (see Supplementary Material) and the instruction for calculation of total OCI-C score are available from the corresponding author or at http://www.optometry.unsw.edu.au/research/oci-c-symptoms-questionnaire-chinese (in the public domain). 
Supplementary Materials
Acknowledgments
The authors thank Yvonne Wu for the assistance with translation quality evaluation, Shi Zhou and Jung Wang for assistance with data collection in China, and the help of Kholoud Bokhary for the Rasch analysis. 
Supported by the University of New South Wales (all authors) and an International Postgraduate Research Scholarship (CC). The authors alone are responsible for the content and writing of the paper. 
Disclosure: C. Chao, None; B. Golebiowski, None; Y. Cui, None; F. Stapleton, None 
References
DEWS. The definition and classification of dry eye disease: report of the definition and classification subcommittee of the International Dry Eye Workshop (2007). Ocul Surf . 2007; 5: 75–92. [CrossRef] [PubMed]
Schaumberg DA Uchino M Christen WG Semba RD Buring JE Li JZ. Patient reported differences in dry eye disease between men and women: impact, management, and patient satisfaction. PLoS One . 2013; 8: e76121. [CrossRef] [PubMed]
Basak SK Pal PP Basak S Bandyopadhyay A Choudhury S Sar S. Prevalence of dry eye diseases in hospital-based population in West Bengal, Eastern India. J Indian Med Assoc . 2012; 110: 789–794. [PubMed]
Chia EM Mitchell P Rochtchina E Lee AJ Maroun R Wang JJ. Prevalence and associations of dry eye syndrome in an older population: The Blue Mountains Eye Study. Clin Exp Ophthalmol . 2003; 31: 229–232. [CrossRef]
Shimmura S Shimazaki J Tsubota K. Results of a population-based questionnaire on the symptoms and lifestyles associated with dry eye. Cornea . 1999; 18: 408–411. [CrossRef] [PubMed]
Salomao MQ Ambrosio R Jr Wilson SE. Dry eye associated with laser in situ keratomileusis: mechanical microkeratome versus femtosecond laser. J Cataract Refract Surg . 2009; 35: 1756–1760. [CrossRef] [PubMed]
Quinto GG Camacho W Behrens A. Postrefractive surgery dry eye. Curr Opin Ophthalmol . 2008; 19: 335–341. [CrossRef] [PubMed]
Ang RT Dartt DA Tsubota K. Dry eye after refractive surgery. Curr Opin Ophthalmol . 2001; 12: 318–322. [CrossRef] [PubMed]
Hovanesian JA Shah SS Maloney RK. Symptoms of dry eye and recurrent erosion syndrome after refractive surgery. J Cataract Refract Surg . 2001; 27: 577–584. [CrossRef] [PubMed]
Solomon KD Holzer MP Sandoval HP Refractive surgery survey 2001. J Cataract Refract Surg . 2002; 28: 346–355. [CrossRef] [PubMed]
Solomon KD Fernandez de Castro LE Sandoval HP Bartholomew LR Vroman DT. Refractive surgery survey 2003. J Cataract Refract Surg . 2004; 30: 1556–1569. [CrossRef] [PubMed]
Sutton GL Kim P. Laser in situ keratomileusis in 2010—a review. Clin Exp Ophthalmol . 2010; 38: 192–210. [CrossRef]
Miller AE McCulley JP Bowman RW Cavanagh HD Wang XH. Patient satisfaction after LASIK for myopia. CLAO J . 2001; 27: 84–88. [PubMed]
Vroman DT Sandoval HP Fernandez de Castro LE Kasper TJ Holzer MP Solomon KD. Effect of hinge location on corneal sensation and dry eye after laser in situ keratomileusis for myopia. J Cataract Refract Surg . 2005; 31: 1881–1887. [CrossRef] [PubMed]
Mian SI Shtein RM Nelson A Musch DC. Effect of hinge position on corneal sensation and dry eye after laser in situ keratomileusis using a femtosecond laser. J Cataract Refract Surg . 2007; 33: 1190–1194. [CrossRef] [PubMed]
Nichols KK Nichols JJ Lynn Mitchell G. The relation between tear film tests in patients with dry eye disease. Ophthalmic Physiol Opt . 2003; 23: 553–560. [CrossRef] [PubMed]
Pflugfelder SC Tseng SC Sanabria O Evaluation of subjective assessments and objective diagnostic tests for diagnosing tear-film disorders known to cause ocular irritation. Cornea . 1998; 17: 38–56. [CrossRef] [PubMed]
Afonso AA Monroy D Stern ME Feuer WJ Tseng SC Pflugfelder SC. Correlation of tear fluorescein clearance and Schirmer test scores with ocular irritation symptoms. Ophthalmology . 1999; 106: 803–810. [CrossRef] [PubMed]
Macri A Pflugfelder S. Correlation of the Schirmer 1 and fluorescein clearance tests with the severity of corneal epithelial and eyelid disease. Arch Ophthalmol . 2000; 118: 1632–1638. [CrossRef] [PubMed]
Chalmers R Long B Dillehay S Begley C. Improving contact-lens related dryness symptoms with silicone hydrogel lenses. Optom Vis Sci . 2008; 85: 778–784. [CrossRef] [PubMed]
Begley CG Caffery B Chalmers RL Mitchell GL. Use of the dry eye questionnaire to measure symptoms of ocular irritation in patients with aqueous tear deficient dry eye. Cornea . 2002; 21: 664–670. [CrossRef] [PubMed]
Nichols KK Nichols JJ Zadnik K. Frequency of dry eye diagnostic test procedures used in various modes of ophthalmic practice. Cornea . 2000; 19: 477–482. [CrossRef] [PubMed]
Johnson ME Murphy PJ. Measurement of ocular surface irritation on a linear interval scale with the Ocular Comfort Index. Invest Ophthalmol Vis Sci . 2007; 48: 4451–4458. [CrossRef] [PubMed]
Liew SH Nichols KK Klamerus KJ Li JZ Zhang M Foulks GN. Tofacitinib (CP-690,550), a Janus kinase inhibitor for dry eye disease: results from a phase 1/2 trial. Ophthalmology . 2012; 119: 1328–1335. [CrossRef] [PubMed]
Evans KS North RV Purslow C. Tear ferning in contact lens wearers. Ophthalmic Physiol Opt . 2009; 29: 199–204. [CrossRef] [PubMed]
Jalbert I Chao C Stapleton F Ashby B Golebiowski B. Associations between corneal nerve density, corneal sensitivity, and tear substance P in lens wearers (E-Abstract 120722). Optom Vis Sci . 2012; 89.
Lin PY Tsai SY Cheng CY Liu JH Chou P Hsu WM. Prevalence of dry eye among an elderly Chinese population in Taiwan: the Shihpai eye study. Ophthalmology . 2003; 110: 1096–1101. [CrossRef] [PubMed]
Yazdani C McLaughlin T Smeeding JE Walt J. Prevalence of treated dry eye disease in a managed care population. Clin Ther . 2001; 23: 1672–1682. [CrossRef] [PubMed]
Doughty MJ Fonn D Richter D Simpson T Caffery B Gordon K. A patient questionnaire approach to estimating the prevalence of dry eye symptoms in patients presenting to optometric practices across Canada. Optom Vis Sci . 1997; 74: 624–631. [CrossRef] [PubMed]
Lam DKT Wong VWY Chow VWS Chi SCC. Epidemiology of dry eye syndrome in Hong Kong: a cross-sectional population-based study. Hong Kong J Ophthalmol . 2011; 15: 58–62.
Bullinger M Alonso J Apolone G Translating health status questionnaires and evaluating their quality: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol . 1998; 51: 913–923. [CrossRef] [PubMed]
Schaumberg DA Sullivan DA Buring JE Dana R. Prevalence of dry eye syndrome among US women. Am J Ophthalmol . 2003; 136: 318–326. [CrossRef] [PubMed]
Lin H Quan C Guo C Zhou C Wang Y Bao B. Translation and validation of the Chinese version of psychosocial impact of dental aesthetics questionnaire. Eur J Orthod . 2013; 35: 354–360. [CrossRef] [PubMed]
Wang CW Chan CLW Jin HY. Psychometric properties of the Chinese version of the 25-item National Eye Institute Vision Function Questionnaire. Optom Vis Sci . 2008; 85: 1091–1099. [CrossRef] [PubMed]
Bland JM Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet . 1986; 1: 307–310. [CrossRef] [PubMed]
Streiner DL Norman GR. Health Measurement Scales: A Practical Guide to Their Development and Use. 3rd ed. New York: Oxford University Press; 2003.
Fleiss JL. Reliability of measurements. In: The Design and Analysis Clinical Experiments . New York: John Wiley & Sons; 1986: 2–31.
Johnson ME Murphy PJ. Measurement of ocular surface irritation on a linear interval scale with the ocular comfort index. Invest Ophthalmol Vis Sci . 2007; 48: 4451–4458. [CrossRef] [PubMed]
Schiffman KK Christianson D Jacobsen G Hirsch JD Reis BL. Reliability and validity of the Ocular Surface Disease Index. Arch Ophthalmol . 2000; 118: 615–621. [CrossRef] [PubMed]
Hom MM Nguyen AL Bielory L. Allergic conjunctivitis and dry eye syndrome. Ann Allergy Asthma Immunol . 2012; 108: 163–166. [CrossRef] [PubMed]
Fujishima H Toda I Shimazaki J Tsubota K. Allergic conjunctivitis and dry eye. Br J Ophthalmol . 1996; 80: 994–997. [CrossRef] [PubMed]
Figure 1
 
The framework of the methodology of the development of the OCI-C.
Figure 1
 
The framework of the methodology of the development of the OCI-C.
Figure 2
 
The items in the original OCI are presented in blue boxes. Translated Chinese terms are the results of the translation process and were used on the preliminary OCI-C.
Figure 2
 
The items in the original OCI are presented in blue boxes. Translated Chinese terms are the results of the translation process and were used on the preliminary OCI-C.
Figure 3
 
The difference between the total score of OCI-C and the OCI plotted against their mean. The dashed line represents a bias of −2.4. The yellow and green lines represent the limits of agreement of +10.3 to −15.2.
Figure 3
 
The difference between the total score of OCI-C and the OCI plotted against their mean. The dashed line represents a bias of −2.4. The yellow and green lines represent the limits of agreement of +10.3 to −15.2.
Figure 4
 
The difference between the total score of OCI-C repeat measurements 1 and 2 plotted against their mean. The dashed line represents a bias of −0.5. The yellow and green lines represent the limits of agreement of +6.4 to −5.3.
Figure 4
 
The difference between the total score of OCI-C repeat measurements 1 and 2 plotted against their mean. The dashed line represents a bias of −0.5. The yellow and green lines represent the limits of agreement of +6.4 to −5.3.
Figure 5
 
Whole fit statistics layout with infit versus outfit scales as ZSTDs. Red box represents ±2 ZSTD.
Figure 5
 
Whole fit statistics layout with infit versus outfit scales as ZSTDs. Red box represents ±2 ZSTD.
Figure 6
 
Person information-weight fit (Person infit) expressed as ZSTD versus person measurement (n = 322).
Figure 6
 
Person information-weight fit (Person infit) expressed as ZSTD versus person measurement (n = 322).
Figure 7
 
Probability curve showing seven-category rating scale (Fig. 7A) and the expected response (Fig. 7B). Intersection of the dashed vertical lines with x-axis represents measures relative to item difficulty.
Figure 7
 
Probability curve showing seven-category rating scale (Fig. 7A) and the expected response (Fig. 7B). Intersection of the dashed vertical lines with x-axis represents measures relative to item difficulty.
Figure 8
 
Person ability/item difficulty map for the OCI-C. The person (represented to the left of the vertical line) and item (represented to the right side of the vertical line) appear in descending order of ability and difficulty on a common scale, which represents the logit value (0–100) of ocular comfort, as transformed from Rasch analysis. Those responses with higher score of ocular discomfort (more able people) and those items with lower scores (harder items) are on the top of the scale, whereas the responses with lower scores of discomfort and easier items are at the bottom. Each “#” represents two participants and “.” represents one. The scale is in log units. M indicates mean; S, 1 SD from the mean; T, 2 SD from the mean; F, frequency; I, intensity; DC, dryness in Chinese; TC, tiredness in Chinese; PC, pain in Chinese; IC, itching in Chinese.
Figure 8
 
Person ability/item difficulty map for the OCI-C. The person (represented to the left of the vertical line) and item (represented to the right side of the vertical line) appear in descending order of ability and difficulty on a common scale, which represents the logit value (0–100) of ocular comfort, as transformed from Rasch analysis. Those responses with higher score of ocular discomfort (more able people) and those items with lower scores (harder items) are on the top of the scale, whereas the responses with lower scores of discomfort and easier items are at the bottom. Each “#” represents two participants and “.” represents one. The scale is in log units. M indicates mean; S, 1 SD from the mean; T, 2 SD from the mean; F, frequency; I, intensity; DC, dryness in Chinese; TC, tiredness in Chinese; PC, pain in Chinese; IC, itching in Chinese.
Table 1
 
The Optimization of Descriptors
Table 1
 
The Optimization of Descriptors
English Items Mean ± SD Chinese Items Bias LoA Mean ± SD Interitem Correlation
Frequency of grittiness 2.2 ± 0.2 FGC1 0.01 −1.4, 1.5 2.1 ± 0.2 0.84
FGC2 0.23 −1.4, 1.9 1.9 ± 0.2 0.78
FGC3 0.05 −2.1, 2.2 2.1 ± 0.2 0.63
Intensity of grittiness 2.2 ± 0.2 IGC1 −0.01 −1.4, 1.5 2.2 ± 0.2 0.83
IGC2 0.18 −1.8, 2.2 2.0 ± 0.2 0.67
IGC3 0 −2.2, 2.2 2.1 ± 0.2 0.60
Frequency of stinging 2.0 ± 0.2 FSC1 0.18 −2.1, 2.5 1.9 ± 0.2 0.55
FSC2 0.31 −2.1, 2.7 1.7 ± 0.2 0.54
FSC3 0.40 −1.7, 2.5 1.6 ± 0.1 0.59
Intensity of stinging 2.1 ± 0.2 ISC1 0.13 −2.3, 2.5 2.0 ± 0.2 0.58
ISC2 0.29 −2.1, 2.7 1.8 ± 0.2 0.55
ISC3 0.40 −1.9, 2.7 1.7 ± 0.2 0.53
Table 2
 
The Overall Results of Reliability
Table 2
 
The Overall Results of Reliability
English Items Mean Score of Items ± SD Chinese Items Mean Score of Items ± SD Cronbach's α P Value Bias LoA
Frequency of dryness 2.8 ± 0.2 FDC 2.9 ± 0.3 0.96 NS −0.06 −1.3, 1.2
Intensity of dryness 2.8 ± 0.2 IDC 2.8 ± 0.2 0.92 NS 0.01 −1.4, 1.5
Frequency of grittiness 2.2 ± 0.2 FGC1 2.1 ± 0.2 0.91 NS 0.01 −1.4, 1.5
Intensity of grittiness 2.2 ± 0.2 IGC1 2.2 ± 0.2 0.91 NS −0.01 −1.4, 1.5
Frequency of stinging 2.0 ± 0.2 FSC1 1.9 ± 0.2 0.71 NS 0.18 −2.1, 2.5
Intensity of stinging 2.1 ± 0.2 ISC1 2.0 ± 0.2 0.73 NS 0.13 −2.3, 2.5
Frequency of tiredness 3.9 ± 0.2 FTC 3.8 ± 0.2 0.95 0.02 0.13 −1.2, 1.5
Intensity of tiredness 3.9 ± 0.3 ITC 3.7 ± 0.2 0.92 0.002 0.20 −1.4, 1.8
Frequency of pain 1.8 ± 0.2 FPC 1.9 ± 0.2 0.92 0.01 −0.13 −1.4, 1.1
Intensity of pain 1.9 ± 0.2 IPC 1.9 ± 0.2 0.95 NS −0.07 −1.1, 1.0
Frequency of itching 2.8 ± 0.2 FIC 2.7 ± 0.2 0.94 0.01 0.15 −1.3, 1.6
Intensity of itching 2.8 ± 0.2 IIC 2.7 ± 0.2 0.93 NS 0.09 −1.3, 1.5
Table 3
 
The Repeatability of the OCI-C
Table 3
 
The Repeatability of the OCI-C
Items Median of the Score for the First Repeat (IQR) Median of the Score for the Second Repeat (IQR) CoR
FDC 3.0 (2.0) 2.8 (1.4) 1.72
IDC 2.0 (2.0) 2.9 (1.4) 1.80
FGC1 2.0 (1.8) 1.5 (0.9) 2.02
IGC1 2.0 (2.0) 1.8 (1.1) 2.26
FSC1 1.0 (0.0) 1.0 (0.8) 1.19
ISC1 1.0 (0.8) 1.8 (1.1) 2.67
FTC 4.0 (2.5) 3.6 (1.8) 2.25
ITC 4.0 (1.8) 3.7 (1.8) 2.06
FPC 1.0 (0.0) 1.5 (1.0) 1.91
IPC 1.0 (0.0) 1.9 (1.3) 2.58
FIC 2.0 (2.0) 2.5 (1.4) 2.33
IIC 2.0 (3.0) 2.7 (1.3) 1.85
Total score 27.9 (9.7) 28.1 (11.5) 5.84
Table 4
 
Median (IQR) Scores of the 12 Questions Between Dry Eye and Non–Dry Eye
Table 4
 
Median (IQR) Scores of the 12 Questions Between Dry Eye and Non–Dry Eye
Dry Eye, n = 81 Non–Dry Eye, n = 241 Dry Eye vs. Non–Dry Eye
All CL, n = 22 NCL, n = 59 P Value All CL, n = 34 NCL, n = 207 P Value P Value
FDC 4.2 (1.4) 3.6 (1.3) 4.4 (1.4) 0.03 2.7 (1.4) 2.9 (1.3) 2.6 (1.4) NS <0.0001
IDC 4.1 (1.3) 3.5 (1.2) 4.2 (1.2) 0.02 2.8 (1.4) 2.9 (1.2) 2.7 (1.3) NS <0.0001
FGC1 3.3 (1.4) 3.5 (1.4) 3.2 (1.4) NS 2.1 (1.3) 2.3 (1.1) 2.0 (1.3) NS <0.0001
IGC1 3.3 (1.4) 3.5 (1.4) 3.2 (1.4) NS 2.1 (1.3) 2.4 (1.2) 2.1 (1.3) NS <0.0001
FSC1 2.7 (1.5) 3.2 (1.7) 2.5 (1.5) NS 2.0 (1.3) 1.9 (1.1) 2.1 (1.4) NS 0.004
ISC1 2.8 (1.5) 3.1 (1.6) 2.6 (1.5) NS 2.2 (1.5) 2.0 (1.2) 2.3 (1.5) NS <0.0001
FTC 4.5 (1.5) 3.9 (1.6) 4.6 (1.4) NS 3.5 (1.5) 3.5 (1.3) 3.5 (1.6) NS <0.0001
ITC 4.3 (1.4) 3.7 (1.3) 4.5 (1.3) 0.04 3.5 (1.4) 3.3 (1.1) 3.5 (1.4) NS <0.0001
FPC 2.7 (1.5) 3.1 (1.5) 2.6 (1.5) NS 2.1 (1.4) 2.3 (1.4) 2.1 (1.4) NS <0.0001
IPC 2.7 (1.5) 3.0 (1.7) 2.6 (1.5) NS 2.1 (1.4) 2.1 (1.3) 2.3 (1.4) NS 0.002
FIC 3.2 (1.7) 3.3 (1.8) 3.2 (1.6) NS 2.7 (1.5) 2.7 (1.4) 2.7 (1.5) NS 0.01
IIC 3.1 (1.6) 3.3 (1.8) 2.6 (1.5) NS 2.8 (1.5) 2.9 (1.6) 2.7 (1.4) NS 0.72
Table 5
 
Comparison of Mean Total Scores Between Dry Eye and Non–Dry Eye Subjects
Table 5
 
Comparison of Mean Total Scores Between Dry Eye and Non–Dry Eye Subjects
n Dry Eye, n = 81 Non–Dry Eye, n = 241
CL, n = 22 NCL, n = 59 CL, n = 34 NCL, n = 207
Total score ± SD 40.0 ± 23.5 41.0 ± 15.0 30.0 ± 14.0 27.0 ± 21.0
Sex (n) Male (36) Female (45) Male (109) Female (132)
Total score ± SD 43.1 ± 14.7 39.0 ± 13.2 30.2 ± 13.7 30.2 ± 12.9
Table 6
 
Fit Statistics for the 12 Items in 322 Subjects
Table 6
 
Fit Statistics for the 12 Items in 322 Subjects
Items in Chinese Infit Outfit
MNSQ ZSTD MNSQ ZSTD
1. FDC - In the past week, how often did your eyes feel dry? 1.19 2.3 1.15 1.8
2. IDC - When your eyes felt dry, typically, how intense was the dryness? 0.88 −1.5 0.9 −1.2
3. FGC1 - In the past week, how often did your eyes feel gritty? 0.88 −1.5 0.80 −2.2
4. IGC1 - When your eyes felt gritty, typically, how intense was the grittiness? 0.80 −2.5 0.75 −2.8
5. FSC1 - In the past week, how often did your eyes feel stingy? 0.97 −0.4 0.82 −1.8
6. ISC1 - When your eyes stung, typically, how intense was the stinging? 1.12 1.4 0.99 −0.1
7. FTC - In the past week, how often did your eyes feel tired? 1.12 1.5 1.12 1.5
8. ITC - When your eyes felt tired, typically, how intense was the tiredness? 0.95 −0.7 0.98 −0.3
9. FPC - In the past week, how often did your eyes feel painful? 0.93 −0.8 0.79 −2.2
10. IPC - When your eyes felt painful, typically, how intense was the pain? 0.96 −0.5 0.85 −1.5
11. FIC - In the past week, how often did your eyes feel itchy? 1.19 2.3 1.11 1.3
12. IIC - When your eyes felt itchy, typically, how intense was the itching? 1.18 2.1 1.24 2.6
Table 7
 
Estimates of Item Difficulties From the Primary (322 Participants) and Secondary (305 Participants) Rasch Analysis for the 12 Items in Chinese OCI-C and Persons
Table 7
 
Estimates of Item Difficulties From the Primary (322 Participants) and Secondary (305 Participants) Rasch Analysis for the 12 Items in Chinese OCI-C and Persons
Items in Chinese 322 Participants 305 Participants
Difficulty, Logits SEs Difficulty, Logits SEs
1. FDC −0.33 0.06 −0.33 0.06
2. IDC −0.36 0.06 −0.38 0.06
3. FGC1 0.41 0.06 0.41 0.06
4. IGC1 0.35 0.06 0.34 0.06
5. FSC1 0.61 0.06 0.67 0.07
6. ISC1 0.43 0.06 0.47 0.06
7. FTC −1.01 0.06 −1.08 0.06
8. ITC 0.95 0.06 −1.00 0.06
9. FPC 0.52 0.06 0.59 0.07
10. IPC 0.52 0.06 0.58 0.07
11. FIC −0.08 0.06 −0.12 0.06
12. IIC −0.11 0.06 −0.16 0.06
Table 8
 
Category Diagnostics for the 12 Items in OCI-C
Table 8
 
Category Diagnostics for the 12 Items in OCI-C
Score Frequency Infit MNSQ Outfit MNSQ OBSVD AVRGE Sample Expect ANDRICH Threshold Score Measure
0 2462 0.83 0.88 −2.71 −2.63 None (−3.20)
1 1388 0.82 0.68 −1.85 −1.85 −1.74 −1.84
2 1187 1.05 0.81 −0.97 −1.20 −1.37 −0.98
3 980 0.97 0.91 −0.54 −0.61 −0.71 −0.17
4 568 1.22 1.30 −0.15 −0.04 0.22 0.78
5 204 1.40 1.43 0.28 0.54 1.26 1.96
6 65 2.14 1.93 0.42 1.12 2.33 (3.64)
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×