BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.
BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjopen.bmj.com).
If you have any questions on BMJ Open’s open peer review process please email [email protected]
http://bmjopen.bmj.com/ on September 26, 2021 by guest. Protected copyright. BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open
Functional visual fields: A cross-sectional study to determine which visual field paradigms best reflect difficulty with mobility function
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2017-018831
Article Type: Research
Date Submitted by the Author: 25-Jul-2017
Complete List of Authors: Subhi, Hikmat; Anglia Ruskin University, Vision and Hearing Sciences Latham, Keziah; Anglia Ruskin University, Vision and Hearing Sciences Myint, Joy; University of Hertfordshire School of Life and Medical Sciences Crossland, Michael; Moorfields Eye Hospital NHS Foundation Trust, Optometry
Primary Subject Ophthalmology Heading:
Secondary Subject Heading: Ophthalmology
Low vision, Visual fields, Perimetry, Mobility function, Visual impairment, http://bmjopen.bmj.com/ Keywords: Self-reported function
on September 26, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 1 of 39 BMJ Open
1 2 3 Title Functional visual fields: A cross sectional study to determine which 4 visual field paradigms best reflect difficulty with mobility function 5 Running head Functional visual fields 6 Authors *Subhi, Hikmat; Anglia Ruskin University, Department of Vision and 7 Hearing Sciences and Vision and Eye Research Unit 8 9 Latham, Keziah; Anglia Ruskin University, Department of Vision and 10 Hearing Sciences and Vision and Eye Research Unit 11 12 Myint, Joy; University of Hertfordshire, Life and Medical Sciences, 13 Postgraduate Medicine 14 15 For peer review only 16 Crossland, Michael D.; Moorfields Eye Hospital, London 17 18 *Corresponding author: [email protected] 19 Word count 3746 20 Key words Visual fields, self reported function, mobility function, functional vision, 21 visual impairment 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 2 of 39
1 2 3 Abstract 4 5 6 Objectives 7 8 9 To develop an appropriate method of assessing visual field (VF) loss which reflects its 10 11 functional consequences, this study aims to determine which method(s) of assessing visual 12 13 fields best reflect mobility difficulty. 14 15 For peer review only 16 Setting 17 18 19 20 This cross sectional observational study took place within a single primary care setting. 21 22 Participants attended a single session at a University Eye Clinic, with data collected by a 23 24 single researcher (HS), a qualified optometrist. 25 26 27 Participants 28 29 30 50 adult participants with peripheral field impairment were recruited for this study. Individuals 31
32 with conditions not primarily affecting peripheral visual function, such as macular http://bmjopen.bmj.com/ 33 34 degeneration, were excluded from the study. 35 36 37 Primary and secondary outcome measures 38 39
40 on September 26, 2021 by guest. Protected copyright. Participants undertook three custom and one standard binocular VF tests assessing visual 41 42 43 field to 60 degrees, and also integrated monocular threshold 24 2 visual fields (IVF). Primary 44 45 VF outcomes were average mean threshold, percentage of stimuli seen, and VF area. VF 46 47 outcomes were compared to self reported mobility function assessed with the Independent 48 49 Mobility Questionnaire, and time taken and patient acceptability were also considered. 50 51 Receiver Operating Characteristic (ROC) curves determined which tests best predicted 52 53 difficulty with mobility tasks. 54 55 56 Results 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 3 of 39 BMJ Open
1 2 3 Greater VF loss was associated with greater self reported mobility difficulty with all field 4 5 paradigms (R2 0.38 0.48, all p<0.001). All four binocular tests were better than the IVF at 6 7 predicting difficulty with at least three mobility tasks in ROC analysis. Mean duration of the 8 9 tests ranged from 1min 26sec (±9sec) for kinetic assessment to 9min 23sec (±24 sec) for 10 11 IVF. 12 13 14 Conclusions 15 For peer review only 16 17 The binocular VF tests extending to 60 deg eccentricity all relate similarly to self reported 18 19 mobility function, and slightly better than integrated monocular VFs. A kinetic assessment of 20 21 VF area is quicker than and as effective at predicting mobility function as static threshold 22 23 24 assessment. 25 26 27 28 29 30 Strengths and limitations of this study 31
32 http://bmjopen.bmj.com/ 33 Strengths of this study include: 34 35 36 o Rigorous and novel visual field testing on a relative large cohort of people 37 38 with visual field loss 39
40 o Consideration of the patient’s perspective of visual field assessment on September 26, 2021 by guest. Protected copyright. 41 42 o Use of Rasch analysed data from a validated questionnaire to reflect a broad 43 44 range of mobility function 45 46 o Different visual field paradigms compared to functional vision, which has not 47 48 previously been reported 49 50 51 52 Potential limitations include: 53 54 55 o Single, rather than repeated, measures of visual field function are used 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 4 of 39
1 2 3 Copyright 4 5 6 The Corresponding Author has the right to grant on behalf of all authors and does grant on 7 8 behalf of all authors, an exclusive licence (or non exclusive for government employees) on a 9 10 worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be 11 12 published in BMJ editions and any other BMJPGL products and sublicences such use and 13 14 exploit all subsidiary rights, as set out in our licence. 15 For peer review only 16 17 Funding statement 18 19 20 21 This work was supported by a College of Optometrists’ Postgraduate Research Scholarship. 22 23 24 Competing interests 25 26 27 The authors report no conflicts of interest and have no proprietary interest in any of the 28 29 materials mentioned in this article. 30 31
32 Results included in this manuscript were presented in part at Vision 2017, The Hague, The http://bmjopen.bmj.com/ 33 34 Netherlands (June 2017). 35 36 37 Author contribution 38 39
40 on September 26, 2021 by guest. Protected copyright. All authors (HS, KL, JM and MC) made substantial contributions to the design of the work. 41 42 43 HS acquired the data for the study. All authors contributed to the analysis and interpretation 44 45 of the data, drafting the manuscript, and critical revision of the article. All authors give final 46 47 approval for the manuscript to be published and agree to be accountable for all aspects of 48 49 the work. 50 51 52 Transparency declaration 53 54 55 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 5 of 39 BMJ Open
1 2 3 I affirm that the manuscript is an honest, accurate, and transparent account of the study 4 5 being reported; that no important aspects of the study have been omitted; and that any 6 7 discrepancies from the study as planned (and, if relevant, registered) have been explained. 8 9 10 Data sharing statement 11 12 13 Data are available from FigShare, doi to be confirmed. 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 6 of 39
1 2 3 Introduction 4 5 6 1 Visual field loss is an extremely important dimension of visual impairment. There are 7 8 2 significant consequences of visual field loss for functional ability1 6, including increased 9 10 3 difficulty with mobility7 12, and potentially also increased risk of falling11,13 18. However, 11 12 4 currently available conventional visual fields tests are designed to detect and monitor the 13 14 5 progression of disease, and are not optimised for reflecting the functional consequences of 15 For peer review only 16 6 visual field loss. Although binocular visual fields are known to be important in individuals with 17 18 7 visual impairment, particularly visual fields beyond 30 degrees19, conventional visual field 19 20 8 tests only measure monocular visual fields of the central 30 degrees or less20 23. 21 22 23 24 24 9 Visual field assessments are demanding procedures and many people dislike performing
25 24,25 26 10 visual field tests . An ideal functional field test would take into account the acceptability of
27 26 28 11 tests to a patient, on the basis that an acceptable test and duration will provide better data . 29 30 12 Patients’ opinion of vision testing is largely unreported, although it has been suggested that 31
32 13 this is due to difficulty objectively quantifying subjective, or “human factors”, of field http://bmjopen.bmj.com/ 33 24,27 34 14 assessment . 35 36 37 15 One aspect of an ideal functional visual field test that is not clear is what test strategies are 38 39 16 most appropriate to use. Of the studies that have previously assessed the visual field
40 on September 26, 2021 by guest. Protected copyright. 41 17 binocularly, strategies including kinetic28 32, suprathreshold strategies including the Esterman 42 43 18 visual field test10,32 37, and threshold tests23,38,39, have been used. 44 45 46 19 The closest current visual field tests to a ‘gold standard’ for assessing functional loss are the 47 48 20 Esterman test40 and the integrated visual field (IVF)41. The Esterman test, the only standard 49 50 21 binocular field test available, is used to determine the extent of visual fields in UK drivers42, 51 52 22 although it was not originally designed for this purpose39. Although some studies suggest the 53 54 23 Esterman test is a good predictor of visual function35,43,44, its limitations are well documented 55 56 24 and include clustering of scores at the higher end of scoring2,10,28,36,37,43,45,46, and inability to 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 7 of 39 BMJ Open
1 2 3 25 differentiate effectively between people with different levels of field loss37. As an alternative, 4 5 26 Crabb and Viswanathan41 constructed integrated visual fields from monocular field plots 6 7 27 using software that estimates sensitivities at points of overlapping locations using a “best 8 9 28 location” algorithm20,37. While integrated visual fields provide a rapid estimate of a patient’s 10 11 29 binocular field without extra perimetric examination20,22,47, they do assume that previous 12 13 30 visual field results are available which is not always the case in low vision clinics. Binocular 14 15 31 function is derivedFor rather peer than measured, review and the plots used usuallyonly only assess the central 16 17 32 30 deg of visual field. 18 19 20 33 The purpose of this study is to develop an optimal binocular visual field test to reflect the 21 22 34 functional consequences of visual field loss by assessing threshold, suprathreshold and 23 24 35 kinetic binocular visual field paradigms examining the visual field to 60 deg eccentricity, and 25 26 existing standard Esterman and IVF fields. Results are compared to self reported difficulty 27 36 28 with mobility to determine which visual field assessment paradigms best reflect functional 29 37 30 31 38 difficulty. The visual field testing experience of individuals with visual impairment is also
32 http://bmjopen.bmj.com/ 33 39 considered to help devise optimal strategies for functional field assessment. 34 35 36 40 Methods 37 38 39 41 Participants with self reported peripheral visual field loss were recruited for this cross
40 on September 26, 2021 by guest. Protected copyright. 41 42 sectional observational study through advertisement of the study with local voluntary agency 42 43 43 Cam Sight, and the charities Retinitis Pigmentosa Fighting Blindness and the International 44 45 44 Glaucoma Association. Individuals with conditions not primarily affecting peripheral visual 46 47 45 function, such as macular degeneration, were excluded from the study, along with those 48 49 46 under 18 years old and those unable to perform verbal evaluations in English. Ethical 50 51 47 approval was granted by Anglia Ruskin University Faculty of Science and Technology 52 53 48 Research Ethics committee. The tenets of the Declaration of Helsinki were upheld. All 54 55 49 participants gave informed consent after the nature of the study was explained. Participants 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 8 of 39
1 2 3 50 attended a single session at the University Eye Clinic, Anglia Ruskin University, with data 4 5 51 collected between 3/7/2015 and 4/3/2016 by a single researcher (HS), a qualified 6 7 52 optometrist. 8 9 10 53 Part 1 of the Independent Mobility Questionnaire (IMQ)28,48 was used to assess self 11 12 54 perceived ability in mobility. Participants were asked to report if they encountered difficulty in 13 14 55 each of 35 mobility situations (Table 1) to obtain a binary response (Yes/No), and then to 15 For peer review only 16 56 rate the level the difficulty they experienced on a 5 point Likert scale. Respondents could 17 18 57 indicate that a mobility task was not applicable to them, which was scored as missing data. 19 20 21 22 How difficult is it for you to undertake the following tasks independently/without 23 assistance (but with the aid of any mobility aids as required): 19. Adjusting to lighting changed during the 24 1. Walking in familiar areas 25 day: indoor to outdoor 20. Adjusting to lighting changed during the 26 2. Walking in unfamiliar areas 27 day: outdoor to indoor 21. Adjusting to lighting changed at night: 28 3. Moving about: at home 29 indoor to streetlights 22. Adjusting to lighting changed at night: 30 4. Moving about: at work 31 streetlights to indoor
32 http://bmjopen.bmj.com/ 5. Moving about: in the classroom 23. Walking in dimly lit indoor areas 33 34 24. Being aware of another person’s 6. Moving about: in stores 35 presence 36 37 7. Moving about: outdoors 25. Avoiding bumping into: people 38 39 8. Moving about: in crowded situations 26. Avoiding bumping into: walls
40 27. Avoiding bumping into: head height on September 26, 2021 by guest. Protected copyright. 9. Walking at night 41 objects 42 28. Avoiding bumping into: shoulder height 10. Using public transportation 43 objects 44 29. Avoiding bumping into: waist height 45 11. Detecting ascending stairwells objects 46 30. Avoiding bumping into: knee height 47 12. Detecting descending stairwells objects 48 49 13. Walking up steps 31. Avoiding bumping into: low lying objects 50 32. Avoiding tripping over uneven travel 51 14. Walking down steps 52 surfaces 53 15. Stepping onto curbs 33. Moving around in social gatherings 54 55 16. Stepping off curbs 34. Finding restrooms in public places 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 9 of 39 BMJ Open
1 2 3 17. Walking through doorways 35. Seeing cars at intersections 4 5 6 18. Walking in high glare areas 7 58 8 9 10 59 Table 1. 35 mobility situations from Part 1 of the Independent Mobility Questionnaire 11 60 (IMQ)28,48. 12 13 61 Participants were also asked to report whether they had fallen in the last 12 months, defined 14 15 as an eventFor which results peer in a person comingreview to rest inadvertently only on the ground, floor or 16 62
17 49 18 63 other lower level . 19 20 21 64 High contrast distance visual acuity with habitual distance spectacle correction was 22 50 51 23 65 assessed binocularly and scored by letter using a 3m internally illuminated EDTRS chart . 24 25 66 If the largest letters could not be read at 3m, the chart was moved 50% closer to the 26 27 67 participant to 1.5m and 0.75m. Participants with acuity that was not measureable were 28 29 68 assigned an acuity of 3.00logMAR52. Contrast sensitivity was measured binocularly with 30 31 69 habitual distance spectacle correction using a Pelli Robson Chart53 at 1m scored on a by
32 http://bmjopen.bmj.com/ 33 70 letter basis54. Participants with no measurable CS function were assigned a score of 34 35 71 0.00logCS52. 36 37 38 72 Five visual field assessments were performed using the Octopus 900 Perimeter55 and the 39 56 40 73 Humphrey Field Analyser . Visual field test order was randomised, and regular breaks were on September 26, 2021 by guest. Protected copyright. 41 42 74 provided. The duration of all assessments was noted. The standard size III Goldmann white 43 44 75 stimulus was used throughout. Participants fixated on the standard orange central point 45 46 76 target (HFA) or green cross target (Octopus 900). When participants had difficulty seeing the 47 48 77 target or maintaining binocular fixation, a 6 deg ring target was utilised on the Octopus, and 49 50 78 an adaption to the fixation target on the HFA that slotted into the fixation target hole to 51 52 79 provide a black 2mm high contrast pericentral ring around the fixation spot were used. 53 54 55 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 10 of 39
1 2 3 80 The three tests conducted on the Octopus 900 Perimeter were as follows, using the 4 5 81 ‘binocular’ setting in all cases: 6 7 8 82 1. Binocular threshold 9 10 83 A custom test point pattern was used, assessing the binocular field to 60 deg from 11 12 84 fixation with 52 points spaced every 7.5 deg in the central 30 deg, and 36 points 13 14 85 spaced every 15 deg in the peripheral 30 60 deg. Full aperture trial lenses were used 15 For peer review only 16 86 in adult half eye trial frames with lens centration distances corrected for near for 17 18 87 assessment of the central 30 deg. The peripheral 30 60 deg field was assessed 19 20 88 without correction to minimise the possibility of lens and frame artefacts. The ‘low 21 22 89 vision’ test strategy was utilised: stimuli are presented using a 4 2 1 dB bracketing 23 24 90 test method starting at 0dB (4000asb) in order to arrive quickly at the expected 25 26 threshold level in subjects with impaired visual fields. The absolute thresholds 27 91 28 achieved at each test location were used to calculate the mean threshold38,41. 29 92 30 31 93
32 http://bmjopen.bmj.com/ 33 94 2. Binocular supra threshold 34 35 95 The same custom test pattern as for binocular threshold assessment was utilized. 36 37 96 Each point was assessed with a stimulus of 10dB intensity. The number of points 38 39 97 seen from the total of 88 was used to calculate a percentage score.
40 on September 26, 2021 by guest. Protected copyright. 41 98 42 43 99 3. Binocular kinetic 44 45 100 A white III 4e (10dB) target was presented kinetically at an angular velocity of 5 46 47 101 deg/sec. Vectors were presented from 60 deg eccentricity and moved centrally in 12 48 49 102 meridians spaced every 30 degrees. This assessment was completed without 50 51 103 refractive correction. The solid angle (deg2) subtended by the isopter was determined 52 53 104 automatically using the Eye Suite software54. The perimeter corrected results for 54 55 105 reaction time bias. 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 11 of 39 BMJ Open
1 2 3 106 The two tests conducted on the HFA were as follows: 4 5 6 107 4. Integrated monocular threshold fields (IVF) 7 8 108 The central 24 2 threshold test with SITA Fast strategy was used to assess 9 10 109 monocular visual fields. Full aperture trial lenses provided refractive correction. 11 12 110 Integrated visual field scores were calculated using the best location 13 14 111 algorithm20,22,23,36,37,41,46,57 59. 15 For peer review only 16 112 17 18 113 5. Esterman 19 20 114 The binocular Esterman visual field test examines 120 test points out to 80 deg 21 22 115 eccentricity using a suprathreshold paradigm with a stimulus intensity of 10dB. This 23 24 116 assessment was completed without refractive correction. The number of points seen 25 26 was used to calculate the percentage Esterman Efficiency score. 27 117 28 29 30 118 Participants were asked to rank the field tests in order from 1 to 5 based on how acceptable 31
32 119 they felt the tests were, and also to rank the perceived usefulness of results presented. http://bmjopen.bmj.com/ 33 34 120 Participants were also invited to make further comments relating to acceptability and output 35 36 121 of the tests, and qualitative themes were evaluated from this data. 37 38 39 122 Analysis
40 on September 26, 2021 by guest. Protected copyright. 41 42 123 Interval data was derived from the ordinal IMQ responses by the use of Rasch analysis60. 43 44 124 Person measures were derived using all 35 items, which have been found to constitute a 45 46 125 unidimensional scale in people with peripheral field loss due to RP28 and glaucoma48. Higher 47 48 126 person measures indicate greater perceived ability. 49 50 51 127 Mann Whitney U tests compared visual function in participants who had reported a fall in the 52 53 128 previous 12 months and those who had not. Continuous clinical function variables were 54 55 129 compared to self reported mobility function in non parametric 2 tailed Spearman’s rho 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 12 of 39
1 2 3 130 bivariate correlations. Due to the multiple number of comparisons performed (7) a more 4 5 131 stringent significance level is more appropriate for these tests, as suggested by the 6 7 132 Bonferroni correction61. A corrected significance level of p=0.007 was used. 8 9 10 133 For the binary responses to the 35 mobility tasks of the IMQ, difficulty was compared to 11 12 134 different visual field test scores using receiver operating characteristic (ROC) analysis to 13 14 135 evaluate how effective the tests were at selecting participants with perceived mobility 15 For peer review only 16 136 difficulty (sensitivity), and without perceived mobility difficulty (specificity). Sensitivity and 17 18 137 specificity were determined for all possible cut off values for the visual field scores and 19 20 138 plotted as ROC curves. An area under the ROC curve of 1 indicates a perfect diagnostic 21 22 139 procedure, whereas 0.5 indicates a poor procedure. A statistical technique appropriate 23 24 140 where two measures are applied to the same set of participants62 was used to compare 25 26 areas under the ROC curves and establish if any visual field test was statistically significantly 27 141 28 better at predicting perceived difficulty. 29 142 30 31
32 143 Results http://bmjopen.bmj.com/ 33 34 35 144 Table 2 shows the characteristics of the 50 participants who were recruited to the study: all 36 37 145 completed the study. Sample size is consistent with previous similar studies 1,28,33,41. Person 38 39 146 measures for the IMQ are shown. Other parameters derived from the Rasch analysis include
40 on September 26, 2021 by guest. Protected copyright. 41 147 person separation of 3.43 (reliability 0.92), indicating that individuals can be reliably ordered 42 43 148 by the instrument in terms of their level of perceived ability. Item separation is 2.95 (reliability 44 45 149 0.90), slightly less than the minimum ideal value of 3, and indicating the instrument might not 46 47 150 be able to reliably order items in terms of their difficulty. Targeting ( 1.23±1.64 logits) is close 48 49 151 to the ideal of within ±1 logits of the mean item difficulty7. In terms of item fit, as a 50 51 152 representation of how well the questions fitted a unidimensional construct, there were three 52 53 153 mis fitting items with fits in the range 1.5 to 2.0 meansquare. These fits do not diminish the 54 55 154 validity of the measures and can be considered acceptable63, and so all items are 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 13 of 39 BMJ Open
1 2 3 155 considered in the analysis. Item fits are also comparable to other Rasch analyses of this 4 5 156 instrument, with three items with slight mis fit also found for samples with RP28,64 and 6 7 157 glaucoma48. 8 9 10 158 11 12 13 Demographic variables 14 Gender (n) 29 Males, 21 Females 15 Age (years) For peer review only 16 17 Median (25% IQ 75% IQ) 64(55 71) 18 Min max 24 84 19 Ocular diagnosis (n) 20 RP 14 21 Glaucoma 23 22 Retinal detachments/tears 4 23 Other 9 24 Registration status (n) 25 Registered severely sight 18 26 impaired 27 Registered sight impaired 8 28 Not registered 24 29 Use of mobility aids (n) 30 White cane or guide dog 23 31 No mobility aids used 27
32 http://bmjopen.bmj.com/ Have you been shown your visual field 33 results before? (n) 34 Yes 36 35 36 No 14 37 Have you fallen in the previous 12mos? 38 Yes 28 39 No 22 Clinical function variables Mean (±std) Median (25% Range
40 on September 26, 2021 by guest. Protected copyright. 41 IQ-75% IQ)) 42 Binocular distance visual acuity 0.28(±0.08) 0.09( 0.06 0.50) 0.28 3.00 43 (logMAR) 44 Binocular contrast sensitivity 1.51(±0.07) 1.65(1.30 1.95) 0.00 1.95 45 (logCS units) 46 Binocular visual field variables 47 Binocular threshold (dB) 10.87(±1.19) 10.14(2.13 0.27 25.60 48 19.40) 49 Binocular suprathreshold (%) 54.48(±5.09) 58.53(18.8 2.27 98.86 50 93.18) 51 Binocular kinetic solid angle 5966.77(±541.19) 7355.7(1783.80 64.20 52 (deg2) 9566.70) 10320.50 53 Esterman (%) 59.43(±4.81) 67.08(33.33 0.00 54 90.83) 100.00 55 Integrated monocular threshold 15.69(±1.52) 15.17(4.88 0.90 31.96 56 (dB) 26.48) 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 14 of 39
1 2 3 Independent Mobility Questionnaire 1.23(±0.23) 1.26( 2.29 ( 5.92 1.84 4 Person Measures 0.09)) 5 6 159 Table 2. Descriptive statistics of the variables assessed (n=50). The mean ± standard 7 160 deviation, and the median (interquartile range) are given for the clinical visual function 8 161 variables. *Number of comorbid conditions from a list of 12 common medical conditions 9 162 representing general health status65. 10 11 163 12 13
14 2 15 Clinical functionFor variables peer reviewIMQ score (Ronly) Have you fallen in the 16 previous 12mos? (U) 17 Distance visual acuity (logMAR) 0.31, p≤0.001 288.0, p=0.701 18 Contrast sensitivity (logCS units) 0.33, p≤0.001 302.5, p=0.913 19 Binocular threshold (dB) 0.47, p≤0.001 236.0, p=0.157 20 Binocular suprathreshold (%) 0.47, p≤0.001 235.0, p=0.161 21 Binocular kinetic solid angle (deg2) 0.48, p≤0.001 236.0, p=0.159 22 Esterman (%) 0.46, p≤0.001 209.0, p=0.053 23 IVF (dB) 0.38, p≤0.001 235.0, p=0.149 24 25 164 Table 3. Relationship between the variables assessed, and self reported mobility function 26 165 and falls history. Mann Whitney U values are provided for falls data and Spearman’s rho 27 166 bivariate correlations are provided for the IMQ score. 28 29 167 Table 3 shows the relationships between the parameters assessed and the outcome 30 31 168 measures of self reported mobility function, and fall history. Greater visual field loss is 32 http://bmjopen.bmj.com/ 33 169 significantly associated with greater self reported difficulty regardless of the method of field 34 35 36 170 assessment, although the relationship is a little weaker for the IVF score. 56% of the sample 37 38 171 reported falling at least once in the previous 12 months, but none of the clinical function 39 172 variables significantly associate with falls history (Table 3).
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 15 of 39 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 173 50 51 Figure 1. Graphical representation of areas under ROC curves for the different visual field 52 174 53 175 tests. The questions to which the numbers on the x axis refer are given in Table 1. * 54 176 indicates a visual field assessment with a statistically significant greater area under the curve 55 177 (AUC) than an alternative assessment. All other comparisons were non significant. 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 16 of 39
1 2 3 178 The ability of the VF tests to discriminate between people with and without difficulty with 4 5 179 mobility tasks was reasonable (AUC>0.7) for at least one VF test for all except 2 of the 35 6 7 180 questions (Figure 3). Statistically significant differences in the areas under the ROC curves 8 9 181 were seen (indicated by * in Figure 1), in that binocular threshold and supra threshold 10 11 182 assessments were better than the IVF at predicting difficulty walking in familiar areas, 12 13 183 walking in unfamiliar areas, walking at home, walking in crowded areas, avoiding bumping 14 15 184 into knee heightFor objects, peer and finding public review toilets. The binocular only threshold assessment was 16 17 185 also better than the IVF at predicting difficulty avoiding bumping into people, while the 18 19 186 Esterman was found to better predict difficulty walking in high glare when compared with the 20 21 187 IVF. All three custom tests and the Esterman assessment were better than the IVF at 22 23 188 predicting difficulty avoiding bumping into waist height objects, and at predicting difficulty 24 25 189 avoiding bumping into low lying objects. 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47
48 190 49 50 191 Figure 2. Test durations of each of the five visual field assessments. 51 52 53 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 17 of 39 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 192 22 23 24 193 Figure 3. Participants’ ranking of visual field tests acceptability 25 26 27 194 The length of time taken to undertake each visual fields assessment is shown in Figure 2. 28 29 195 The quickest test was the kinetic which took a mean of 1min 26sec (±9sec), while the mean 30 31 196 duration of the longest test, the IVF, was 9min 23sec (±24 sec). 32 http://bmjopen.bmj.com/ 33 34 35 197 In terms of the participants’ view of how they found the tests to undertake, the most favoured 36 37 198 assessment was the kinetic (Figure 3), while the IVF was ranked the least favourite test by 38 39 199 over 60% of participants. In addition to ranking the tests, participants also made comments
40 on September 26, 2021 by guest. Protected copyright. 41 200 on the different test strategies, and common themes are considered here. Three participants 42 43 201 commented that shorter tests were preferred, the kinetic assessment in particular. 44 45 202 Participants found the kinetic assessment pleasant, less stressful, and encouraging, with 46 47 203 three participants commenting on being reassured by knowing that a light would be seen 48 49 204 eventually. Seven participants suggested that the kinetic assessment was more fun and 50 51 205 engaging than the static tests, and one remarked on the assessment’s novelty value. 52 53 206 However, six participants expressed concern that the test was too basic or too short, and 54 55 207 that the accuracy of results would be compromised by the test’s rapidity. Participants were 56 57 208 happy to conduct a longer test if they knew that results would be more beneficial to the 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 18 of 39
1 2 3 209 practitioner. Eleven participants indicated preference for binocular tests, which were reported 4 5 210 as more comfortable and less tiring than monocular assessments. Five participants preferred 6 7 211 the uniformly bright lights on the binocular suprathreshold assessment; however, four found 8 9 212 the bright lights at the start of the binocular threshold assessment encouraging. 10 11 12 213 All participants in the study reported previous experience of visual field testing, although 13 14 214 28% had not seen their visual field results before (Table 2). All perceived value in being 15 For peer review only 16 215 shown their visual fields. In terms of visual field outputs, 48% of participants preferred the 17 18 216 grey scale plots from the binocular threshold analysis, and individuals remarked on the 19 20 217 greater level of detail provided on a grey scale plot compared to others. The kinetic plot was 21 22 218 most favoured by 37% of participants, with preference expressed for plots that indicated 23 24 219 greater levels of residual field. Four participants commented on the usefulness of having 25 26 their peripheral field represented, with one participant suggesting that the monocular 27 220 28 threshold plots were an inaccurate depiction that overestimated the extent of their visual 29 221 30 31 222 field. The Esterman output was the least favoured by 38% of participants. Three participants
32 http://bmjopen.bmj.com/ 33 223 commented on the size of points on suprathreshold outputs, and expressed difficulty viewing 34 35 224 the results. 36 37 38 225 Discussion 39
40 on September 26, 2021 by guest. Protected copyright. 41 226 Five different visual field assessments were compared to self reported difficulty with mobility, 42 43 227 in order to determine which visual field paradigms are suitable for use as a functional visual 44 45 228 field assessment. There are four principal findings of this study. Firstly, the paradigm used to 46 47 229 assess the visual field (threshold or suprathreshold static, or kinetic) makes little difference 48 49 230 to the relationship with function: so long as the test is performed binocularly and includes 50 51 231 assessment of eccentricities to 60 deg, the visual field outcome measure reflects self 52 53 232 reported mobility function well, in terms of both correlation between visual field score and 54 55 233 mobility function, and in terms of ability of visual field scores to discriminate between 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 19 of 39 BMJ Open
1 2 3 234 individuals who do and do not have difficulty with specific mobility tasks. Secondly, tests that 4 5 235 are quicker to perform (binocular suprathreshold, binocular kinetic, Esterman; Figure 2) 6 7 236 relate just as well to self reported mobility function as tests that take longer (binocular 8 9 237 threshold; Table 2). Functional information is not lost by using kinetic or suprathreshold 10 11 238 techniques when compared to the diagnostic gold standard of measuring static thresholds. 12 13 239 Thirdly, quicker tests are also more acceptable to patients (Figure 3), especially the 14 15 240 binocular suprathresholdFor peer and kinetic review paradigms. Finally, althoughonly all participants had 16 17 241 experience of visual field assessments, over a quarter did not recall being shown their visual 18 19 242 field results before. All participants in this study found value in seeing their visual field 20 21 243 results, and clinicians are encouraged to explain visual field results with respect to likely 22 23 244 functional difficulties where possible. 24 25 26 A strength of this study is that although some previous studies44,66 68 have compared 27 245 28 different methods of quantifying the visual field and their relationship with functional vision, 29 246 30 31 247 we are not aware of any previous studies that have compared different paradigms of visual
32 http://bmjopen.bmj.com/ 33 248 field assessment with functional vision, nor that have taken the patient’s perspective into 34 35 249 consideration. Rigorous and novel visual field testing on a relative large cohort of people with 36 37 250 visual field loss has been used, along with Rasch analysed data from a validated 38 39 251 questionnaire to reflect a broad range of mobility function.
40 on September 26, 2021 by guest. Protected copyright. 41 42 252 One possible limitation of the study is that we rely on self report of mobility difficulties rather 43 44 253 than directly measuring visual performance on, for example, a mobility circuit12,30. Another 45 46 254 potential limitation is that single responses have been used in the kinetic paradigm. 47 48 255 Responses to kinetic stimuli can be variable69, and repeated presentations might be 49 50 256 necessary to confirm responses, and reduce the impact of outlying responses70. 51 52 53 257 The findings are consistent with previous studies, in that we find that the IVF relates to 54 55 258 mobility function57,71. Contrary to other studies however37,41, the IVF appears to be less 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 20 of 39
1 2 3 259 effective at relating to self related mobility function when compared to the Esterman 4 5 260 assessment. This may be due to the difference in the degree of visual field loss between the 6 7 261 sample groups. The average Esterman scores in the current study (56.4%) are less than the 8 9 262 average scores (87.4%36 and 86.7%40) reported in other studies, and would suggest that the 10 11 263 present sample has a greater degree of visual field loss. Contrary to some previous studies, 12 13 264 no significant correlations were found here between measures of clinical function and fall 14 15 265 history. VisualFor field loss peerhas been shown review to increase the risk ofonly falling in some studies6,11,13 16 17 266 18, but not in others72,73. Variation in findings reflects the multifactorial nature of falls, and also 18 19 267 suggests limitations in retrospective reporting of falls74. 20 21 22 268 In conclusion, a binocular visual field test that does not ignore the peripheral 30 60 deg of 23 24 269 the field is effective for reflecting functional difficulty, regardless of the whether a threshold, 25 26 suprathreshold or kinetic assessment paradigm is employed. A shorter duration visual field 27 270 28 test such as the binocular kinetic assessment used in this study is favoured by patients, and 29 271 30 31 272 is similarly effective to more time consuming tests in predicting perceived disability in
32 http://bmjopen.bmj.com/ 33 273 patients with peripheral field loss. To further explore the potential of a binocular kinetic 34 35 274 assessment as a functional field test, future work needs to explore the repeatability of 36 37 275 responses and the necessity for repeated presentations. 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 21 of 39 BMJ Open
1 2 3 References 4 5 6 1. Bibby SA, Maslin ER, McIlraith R, Soong GP. Vision and self reported mobility 7 8 performance in patients with low vision. Clin Exp Optom. 2007;90(2):115 23. 9 10 2. Nelson P, Aspinall P, Papasouliotis O, Worton BO, Brien C. Quality of life in 11 12 glaucoma and its relationship with visual function. J Glaucoma. 2003;12(2):139 150. 13 14 3. Tabrett DR, Latham K. Factors influencing self reported vision related activity 15 For peer review only 16 limitation in the visually impaired. Invest Ophthal Vis Sci. 2011;52:5293–5302. 17 18 4. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 19 20 field loss in the elderly and associations with impairment in daily functioning: the 21 22 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 23 24 5. Sherwood M, Garcia Siekavizza A, Meltzer M, et al. Glaucoma’s impact on quality of 25 26 life and its relation to clinical indicators. Ophthalmology. 1998;105:561–566. 27 28 6. Ivers RQ, Cumming RG, Mitchell P, et al. Visual impairment and falls in older adults: 29 30 31 the Blue Mountains Eye Study. J Am Geriatr Soc. 1998;46:58–64.
32 http://bmjopen.bmj.com/ 33 7. Latham K, Baranian M, Timmis MA, Pardhan S. Difficulties with Goals of the Dutch 34 35 ICF Activity Inventory: Perceptions of Those with Retinitis Pigmentosa and of Those 36 37 Who Support Them. Inves Ophthal Vis Sci. 2015;56(4):2381 2391. 38 39 8. Lowe J, Drasdo N. Patients' responses to retinitis pigmentosa. Optom Vis Sci.
40 on September 26, 2021 by guest. Protected copyright. 41 1992;69 (3):182 185. 42 43 9. Szlyk JP, Fishman GA, Grover S, Revelins BI, Derlacki DJ. Difficulty in performing 44 45 everyday activities in patients with juvenile macular dystrophies: comparison with 46 47 patients with. Br J Ophthalmol. 1998;82(12):1372 1376. 48 49 10. Noe G, Ferraro J, Lamoureux E, Rait J, Keeffe JE. Associations between 50 51 glaucomatous visual field loss and participation in activities of daily living. Clin Exp 52 53 Optom. 2003;31(6):482 6. 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 22 of 39
1 2 3 11. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 4 5 field loss in the elderly and associations with impairment in daily functioning: the 6 7 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 8 9 12. Lovie Kitchin JE, Soong GP, Hassan SE, Woods RL. Visual field size criteria for 10 11 mobility rehabilitation referral. Optom Vis Sci. 2010;87(12):948 57. 12 13 13. Patino C M, McKean Cowdin R, Azen S P, Allison JC, Choudhury F, Varma R. 14 15 CentralFor and Peripheral peer Visual Impairment review and the Risk only of Falls and Falls with Injury. 16 17 Ophthalmology. 2010;117(2):199. 18 19 14. Klein BEK, Klein R, Lee KE, Cruickshanks KJ. Performance based and self 20 21 assessed measures of visual function as related to history of falls, hip fractures, and 22 23 measured gait time: The Beaver Dam Study. Ophthalmology. 1998;105(1):160 164. 24 25 15. Jack CIA, Smith T, Neoh C, Lye M, Mcgalliard JN. Prevalence of low vision in elderly 26 27 patients admitted to an acute geriatric unit in Liverpool: Elderly people who fall are 28 29 more likely to have low vision. Gerontology. 1995;41(5):280 285. 30 31 16. Freeman EE, Munoz B, Rubin GS, West SK. Visual field loss increases the risk of 32 http://bmjopen.bmj.com/ 33 falls in older adults: the Salisbury eye evaluation. Inves Ophthal Vis Sci. 34 35 36 2007;48(10):4445 4450. 37 38 17. Klein BEK, Moss SE, Klein R, Lee KE, Cruickshanks KJ. Associations of visual 39 function with physical outcomes and limitations 5 years later in an older population:
40 on September 26, 2021 by guest. Protected copyright. 41 42 The Beaver Dam eye study. Ophthalmology. 2003;110(4):644 650. 43 44 18. Haymes S. A., LeBlanc R. P., Nicolela M. T., Chiasson L. A., Chauhan B. C. Risk of 45 46 falls and motor vehicle collisions in glaucoma. Invest. Ophthalmol. Vis. Sci. 47 48 2007;48:1149–1155 49 50 19. Subhi H, Latham K, Myint J, Crossland M. Functional visual fields: relationship of 51 52 visual field areas to self reported function. Ophthalmic Physiol Opt. 2017;37:399– 53 54 408. 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 23 of 39 BMJ Open
1 2 3 20. Nelson Quigg JM, Cello K, Johnson CA. Predicting binocular visual field sensitivity 4 5 from monocular visual field results. Inves Ophthal Vis Sci. 2000;41:2212–2221. 6 7 21. Schneck ME, Haegerstom Portnoy G, Lott L, Brabyn JA. Monocular vs. Binocular 8 9 Measurement of Spatial Vision in Elders. Optom Vis Sci. 2010;87(8):526 531. 10 11 22. Asaoka R, Crabb D, Yamashita T, Russell R, Wany YX, Garway Heath DF. Patients 12 13 Have Two Eyes!: Binocular versus Better Eye Visual Field Indices. Inves Ophthal Vis 14 15 Sci. 2011;52(9):7007 7011.For peer review only 16 17 23. Crabb DP, Smith ND, Glen FC, Burton R, Garway Heath DF. How does glaucoma 18 19 look? Patient perception of visual field loss. Ophthalmology. 2013;120(6):1120 6. 20 21 24. Gardiner SK, Demirel S. Assessment of patient opinions of different clinical tests 22 23 used in the management of glaucoma. Ophthalmology. 2008;115:2127–31. 24 25 25. Glen FC, Baker H, Crabb DP A qualitative investigation into patients’ views on visual 26 27 field testing for glaucoma monitoring BMJ Open 2014;4(1). 28 29 26. Hudson C, Wild JM, O’Neill EC. Fatigue effects during a single session of automated 30 31 static threshold perimetry. Invest Ophthalmol Vis Sci. 1994;35:268–80. 32 http://bmjopen.bmj.com/ 33 27. Artes, P, Monhart M, Marco M. Are big stimuli “easier” to see? An analysis of 34 35 36 response times from frequency of seeing data. In 22nd International Visual Field 37 38 Imaging Symposium. Udine, Italy, September 27th 30th 2016. 39 28. Turano KA, Rubin GS, Quigley HA Mobility performance in glaucoma. Invest
40 on September 26, 2021 by guest. Protected copyright. 41 42 Ophthalmol Vis Sci. 1999;40:2803 2309. 43 44 29. Lovie Kitchin JE, Mainstone J, Robinson J, Brown B. What areas of the visual field 45 46 are important for mobility in low vision patients? Clin Vis Sci. 1990;5:249 263. 47 48 30. Haymes SA, Johnston AW, Heyes AD. Relationship between vision impairment and 49 50 ability to perform activities of daily living. Ophthalmic Physiol Opt. 2002;22(2):79 91. 51 52 31. Haymes SA, Leblanc RP, Nicolela MT et al. Risk of falls and motor vehicle collisions 53 54 in glaucoma. Invest Ophthalmol Vis Sci. 2007;48:1149–55. 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 24 of 39
1 2 3 32. Hassan SE, Hicks JC, Lei H, Turano KA. What is the minimum field of view required 4 5 for efficient navigation? Vision Res. 2007;47(16):2115 2123. 6 7 33. Lee J, Cho H, Kee C. Assessment of the Vision Specific Quality of Life Using 8 9 Binocular Esterman Visual Field in Glaucoma Patients. J Korean Ophthalmol Soc, 10 11 2013;54(10):1567. 12 13 34. Mills RP, Drance SM. Esterman disability rating in severe glaucoma. Ophthamology. 14 15 1986;93:371 378.For peer review only 16 17 35. Fujita K, Yasuda N, Nakamoto K, Fukuda T. The relationship between difficulty in 18 19 daily living and binocular visual field in patients with glaucoma. Nihon Ganka Gakkai 20 21 Zasshi. 2008;112(5):447 50. 22 23 36. Jampel HD. Glaucoma patients’ assessment of their visual function and quality of life. 24 25 Tr Am Ophth Soc. 2001;99:301 317. 26 27 37. Jampel HD, Friedman DS, Quigley H, et al. Correlation of the binocular visual field 28 29 with patients’ assessment of vision. Inves Ophthal Vis Sci. 2002;43:1059 67. 30 31 38. Tabrett DR, Latham K. Important areas of the central binocular visual field for daily 32 http://bmjopen.bmj.com/ 33 functioning in the visually impaired. Ophthalmic Physiol Opt. 2012;32:156 163. 34 35 36 39. Black A, Lovie Kitchin J, Woods R, Arnold N, Byrnes J, Murrish J. Mobility 37 38 performance with retinitis pigmentosa. Clin Exp Optom. 1996;80(1):1 12. 39 40. Esterman B. Functional scoring of the binocular field. Ophthalmology. 1982;89:1226–
40 on September 26, 2021 by guest. Protected copyright. 41 42 34. 43 44 41. Crabb DP, Viswanathan AC. Integrated visual fields: a new approach to measuring 45 46 the binocular field of view and visual disability. Graefes Arch Clin Exp Ophthalmol. 47 48 2004;243(3):210 6. 49 50 42. Driver and Vehicle Licensing Agency. Visual disorders: assessing fitness to drive. 51 52 2016 Available: https://www.gov.uk/guidance/visual disorders assessing fitness to 53 54 drive. Last accessed February 2017. 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 25 of 39 BMJ Open
1 2 3 43. Parrish RK II, Gedde SJ, Scott IU, et al. Visual function and quality of life among 4 5 patients with glaucoma. Arch Ophthalmol. 1997;115:1447 1455. 6 7 44. Choy ES, Mills RP, Drance SM. Automated Esterman testing of disability in 8 9 glaucoma. In:Greve E, Heiji A, eds. Seventh International Visual Field Symposium. 10 11 Amsterdam, the Netherlands: Junk Publishers. 1986:527 535. 12 13 45. Mills RP. Correlation of quality of life with clinical symptoms and signs at the time of 14 15 glaucomaFor diagnosis. peer Trans Am Ophthalmolreview Soc. 1998;96:753 812. only 16 17 46. Harris ML, Jacobs, NA. Is the Esterman binocular field sensitive enough? In: Mills 18 19 RP, Wall M. Perimetry update 1994/1995. Amsterdam: Kugler, 1995:13–24. 20 21 47. Crabb DP, Viswanathan AC, McNaught AI, et al. Simulating binocular visual field 22 23 status in glaucoma. Br J Ophthalmol. 1998;82:1236–41. 24 25 48. Turano KA, Geruschat DR, Baker FH. Fixation behavior while walking: persons with 26 27 central visual field loss. Vision Res. 2002;42:2635–2644. 28 29 49. Skelton DA, Becker C, Lamb SE, Close JCT, Zijlstra W, Yardley L, Todd CJ. 30 31 Prevention of Falls Network Europe: a Thematic Network aimed at introducing good 32 http://bmjopen.bmj.com/ 33 practice in effective falls prevention across Europe. Eur J Aging. 2004;1:89 94. 34 35 36 50. Arditi A, Cagenello B. On the statistical reliability of letter chart visual acuity 37 38 measurement. Invest Ophthalmol Vis Sci. 1993;34:120–129. 39 51. Ferris FL, Kassoff A, Bresnick GH, Bailey IL. New visual acuity charts for clinical
40 on September 26, 2021 by guest. Protected copyright. 41 42 research. Am J of Ophthalmol. 1982;94:91–96. 43 44 52. Myint J, Latham K, Mann D, Gomersall P, Wilkins A, Allen P. The relationship 45 46 between visual function and performance in rifle shooting for athletes with vision 47 48 impairment. BMJ Open Sport Exerc Med. 2016;2(1). 49 50 53. Pelli DG, Robson JG, Wilkins AJ. The design of a new letter chart for measuring 51 52 contrast sensitivity. Clin Vis Sci. 1988;2:187 199. 53 54 54. Elliott DB, Bullimore MA, Bailey IL. Improving the reliability of the Pelli Robson 55 56 contrast sensitivity test. Clin Vis Sci. 1991;6(6):471 475. 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 26 of 39
1 2 3 55. Octopus, Haag Streit International, AG, Bern, Switzerland 4 5 56. HFA, Carl Zeiss Meditec, Inc., Dublin, CA. 6 7 57. Aspinall PA, Johnson ZK, Azuara Blanco A, Montarzino A, Brice R, Vickers A. 8 9 Evaluation of quality of life and priorities of patients with glaucoma. Invest 10 11 Ophthalmol Vis Sci. 2008;49(5):1907 1915. 12 13 58. Saunders LJ, Russell RA, Crabb DP. Practical landmarks for visual field disability in 14 15 glaucoma.For Br J Ophthalmol.peer 2012;96:1185–1189. review only 16 17 59. Chisholm CM, Rauscher FG, Crabb DC, et al. Assessing Visual Fields for Driving in 18 19 Patients With Paracentral Scotomata. Br J Ophthalmol. 2008;92(2):225 230. 20 21 60. Rasch G. Probabalistic models for some intelligence and attainment tests. Chicago: 22 23 Mesa Press. 1993. 24 25 61. Field AP. Discovering statistics using SPSS. 2nd ed. London: SAGE publications Ltd. 26 27 2005. 28 29 62. DeLong ER, DeLong DM, Clarke Pearson DL. Comparing the areas under two or 30 31 more correlated receiver operating characteristic curves: a nonparametric approach. 32 http://bmjopen.bmj.com/ 33 Biometrics. 1988;44(3):837 45. 34 35 36 63. Linacre JM. Dimensionality investigation an example. Retrieved 12/11/2014, from 37 38 http://www.winsteps.com/winman/multidimensionality.htm. 39 64. Fenwick EK, O'Hare F, Deverell L, Ayton LN, Luu CD, McSweeney S, Bentley SA,
40 on September 26, 2021 by guest. Protected copyright. 41 42 Guymer RH, Finger RP. Rasch Analysis of the Independent Mobility Questionnaire. 43 44 Optom Vis Sci. 2016;93(2):181 187. 45 46 65. van Nispen RMA, Hoeijmakers JGJ, de Boer MR, et al. Agreement Between Self 47 48 Reported Co morbidity of Visually Impaired Older Patients and Reports from their 49 50 General Practitioners. Visual Impair Res. 2008;10(2 3):49 56 51 52 66. Haymes S, Guest D, Heyes A, Johnston A. Mobility of people with retinitis 53 54 pigmentosa as a function of vision and psychological variables. Optom Vis Sci. 55 56 1996;73(10):621 637. 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 27 of 39 BMJ Open
1 2 3 67. Safran AB, Glaser JS. Statokinetic Dissociation in Lesions of the Anterior Visual 4 5 Pathways: A Reappraisal of the Riddoch Phenomenon. Arch Ophthalmol. 6 7 1980;98:291 295. 8 9 68. Yanagisawa M, Kato, S, Kunimatsu S, Tamura M, Ochiai M. Relationship between 10 11 vision related quality of life in Japanese patients and methods for evaluating visual 12 13 field. Jpn J Ophthalmol, 2011;55(2):132 137. 14 15 69. LynnFor J, Swanson peer W, Fellman R.review Evaluation of automated only kinetic perimetry (AKP) 16 17 with the Humphrey Field Analyzer. Perimetry Update. 1990;1991:433 452. 18 19 70. Moenter VM, Crabb DP, Artes PH. Reclaiming the Periphery: Automated Kinetic 20 21 Perimetry for Measuring Peripheral Visual Fields in Patients with Glaucoma. Invest 22 23 Ophthal Vis Sci. 2017;58(2):868 875. 24 25 71. Black AA, Wood JM, Lovie Kitchin JE. Inferior field loss increases rate of falls in older 26 27 adults with glaucoma. Optom Vis Sci. 2011;88(11):1275 82 28 29 72. Friedman SM, Munoz B, West SK, et al. Falls and fear of falling: which comes first? A 30 31 longitudinal prediction model suggests strategies for primary and secondary 32 http://bmjopen.bmj.com/ 33 prevention. J Am Geriatr Soc. 2002;50:1329 35. 34 35 36 73. Glynn RJ, Seddon JM, Krug JH, Jr, et al. Falls in elderly patients with glaucoma. Arch 37 38 Ophthalmol. 1991;109:205–10. 39 74. Cummings SR, Nevitt MC, Kidd S. Forgetting falls. The limited accuracy of recall of
40 on September 26, 2021 by guest. Protected copyright. 41 42 falls in the elderly. J Am Geriatr Soc. 1988;36:613–16. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open Page 28 of 39
Ophthalmic & Physiological Optics ISSN 0275-5408 1 2 3 4 5 Functional visual fields: relationship of visual field areas to 6 self-reported function 7 8 Hikmat Subhi1, Keziah Latham1, Joy Myint2 and Michael D. Crossland3 9 1 2 3 10 Anglia Ruskin University, Cambridge, University of Hertfordshire, Hatfield, and Moorfields Eye Hospital NHS Foundation Trust, London, UK 11 12 13 Citation information: Subhi H, Latham K, Myint J & Crossland MD. Functional visual fields: relationship of visual field areas to self-reported 14 function. Ophthalmic Physiol Opt 2017. doi: 10.1111/opo.12362 15 For peer review only 16 Keywords: mobility function, self-reported Abstract 17 function, visual fields 18 Purpose: The aim of this study is to relate areas of the visual field to functional 19 Correspondence: Hikmat Subhi difficulties to inform the development of a binocular visual field assessment that 20 E-mail address: [email protected] can reflect the functional consequences of visual field loss. 21 Methods: Fifty-two participants with peripheral visual field loss undertook binoc- 22 Received: 13 October 2016; Accepted: 24 ular assessment of visual fields using the 30-2 and 60-4 SITA Fast programs on January 2017 23 the Humphrey Field Analyser, and mean thresholds were derived. Binocular 24 visual acuity, contrast sensitivity and near reading performance were also deter- 25 mined. Self-reported overall and mobility function were assessed using the Dutch 26 ICF Activity Inventory. 27 Results: Greater visual field loss (0–60°) was associated with worse self-reported 28 function both overall (R2 = 0.50; p < 0.0001), and for mobility (R2 = 0.64; 29 p < 0.0001). Central (0–30°) and peripheral (30–60°) visual field areas were simi- 30 larly related to mobility function (R2 = 0.61, p < 0.0001 and R2 = 0.63,
31 p < 0.0001 respectively), although the peripheral (30–60°) visual field was the best http://bmjopen.bmj.com/ 32 predictor of mobility self-reported function in multiple regression analyses. Supe- 33 rior and inferior visual field areas related similarly to mobility function (R2 = 0.56, 34 p < 0.0001 and R2 = 0.67, p < 0.0001 respectively). The inferior field was found to 35 be the best predictor of mobility function in multiple regression analysis. 36 Conclusion: Mean threshold of the binocular visual field to 60° eccentricity is a 37 good predictor of self-reported function overall, and particularly of mobility 38 – ° – ° 39 function. Both the central (0 30 ) and peripheral (30 60 ) mean threshold are – ° on September 26, 2021 by guest. Protected copyright. 40 good predictors of self-reported function, but the peripheral (30 0 ) field is a 41 slightly better predictor of mobility function, and should not be ignored when 42 considering functional consequences of field loss. The inferior visual field is a 43 slightly stronger predictor of perceived overall and mobility function than the 44 superior field. 45 46 functional fields. Currently available conventional visual Introduction 47 fields tests are designed to detect and monitor the progres- 48 The functional consequences of visual field loss are known sion of disease, and no visual field test currently available is – 49 to include diminished mobility,1 3 diminished ability to optimised for reflecting the functional consequences of 50 complete activities such as reading or watching television,4 visual field loss. Although there are numerous studies that – – – 51 and an increased risk of falling.4 6 Binocular visual field relate self-reported function11 27 or performance28 34 to 52 assessment is thought to represent functional abilities better visual field parameters, few do so with the intention of 53 than monocular assessment, especially in individuals with developing a clinically applicable method of functional field 54 visual impairment, as most activities of daily living are usu- assessment. – 55 ally performed with both eyes open.7 10 However, there is Development of a binocular visual field test that can 56 no standard reference method for assessing binocular reflect functional difficulty would be a valuable tool in low 57 58 59 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
Page 29 of 39 BMJ Open
Functional visual fields H Subhi et al. 1 2 3 vision assessment. Quantification of visual field loss with under 18 years old and those unable to perform verbal 4 an understanding of how scores relate to functional diffi- evaluations in English. Ethical approval was granted by 5 culty would be helpful not only in assessing and managing Anglia Ruskin University Faculty of Science and Technol- 6 the low vision patient, but also in determining robust crite- ogy Research Ethics committee. The tenets of the Declara- 7 ria for visual impairment registration as compared to visual tion of Helsinki were upheld. All participants gave 8 field criteria currently in place in the UK that are open to informed consent after the nature of the study was 9 significant subjective interpretation.35 explained. All data collection was conducted by one experi- 10 Previous studies have related visual field loss to func- enced optometrist. – 11 tion,11 33 but many of these have used conventional A face-to-face interview elicited demographic characteris- 12 monocular visual fields tests that do not reflect the binocu- tics (Table 1). The Dutch ICF Activity Inventory (D-AI)40,41 – 13 lar field,11,12,15 20,28 or have assessed the visual field using was used to determine self-reported function in Activities of 14 monocular thresholdFor tests to constructpeer a binocular review field Daily Living, sinceonly it covers a wide range of goals relevant to 15 plot.14,21 Of the studies that have assessed the visual field people with peripheral visual field loss.43 The original ques- 16 binocularly, the majority have assessed the visual field out tionnaire assesses the difficulty of 47 rehabilitation goals, 17 to 30°.13,30 Those studies that have assessed the binocular nested within nine domains of the World Health Organisa- 18 visual field past 30° have used kinetic27,29,31,32,34 or supra- tion International Classification of Functioning frame- 19 42 threshold test strategies such as the Esterman visual field work. In the current study, participants were asked to 20 – test.22 26 Threshold sensitivities of the peripheral visual grade their perceived level of difficulty for 43 goals. These 21 field have not previously been determined. represented the 47 goals proposed by Bruijning et al.40 but 22 The current study builds on previous work that used a excluded the two goals underpinning the 10th emotional 23 threshold paradigm to assess the binocular visual field out health domain, and a further two relating to driving and 24 to 30° and found that mean thresholds can predict self- riding a bicycle that have been shown not to fit the unidi- 25 13 26 reported overall and mobility function. However, it is not mensional construct of the questionnaire in people with ° 43 27 known if testing the visual field past 30 is of further bene- peripheral vision loss. Respondents could indicate that a 28 fit. Furthermore, numerous studies have suggested the par- goal was not important or not applicable to them, which ticular significance of the inferior visual field for mobility was scored as missing data. If the goal was relevant, diffi- 29 28,29,36–39 30 function, and inferior visual field loss has been culty was rated on a five-point Likert scale (none, slight, shown to impair mobility performance28,29,39 and perceived moderate, very difficult, impossible). Responses to the four
31 http://bmjopen.bmj.com/ 36 37 32 mobility function, increase postural sway, and increase goals of the mobility domain (mobility at home, mobility 37,38 33 the risk of falling in individuals with no visual impair- indoors, mobility outdoors, and using public transport) 28,38 36,37 34 ment, with early visual field loss, or with simulated were used to determine self-reported mobility function. 39 35 field loss. It is not known whether the inferior field High contrast distance visual acuity was assessed binocu- 36 remains more significant to mobility function in individu- larly and scored by-letter with participants’ habitual dis- 37 als with a greater degree of visual field loss. The importance tance spectacle correction using a 3 m internally 38 of the inferior visual field to overall self-reported function illuminated ETDRS logMAR chart.44 If the largest letters 39 is also unclear. could not be read at 3 m, the chart was moved 50% closer on September 26, 2021 by guest. Protected copyright. 40 The current study uses a binocular threshold test to to the participant to 1.5 and 0.75 m. MNRead charts were 41 extend the findings of Tabrett and Latham13 and assesses used at 40 cm to determine binocular clinical reading per- 42 the visual field out to 60° to determine whether the func- formance with habitual near spectacle correction or a dis- 43 tional visual field would benefit from being assessed beyond tance correction with +2.50 reading addition where 44 30° from fixation, and to test the significance of the inferior appropriate.45,46 Participants with acuity that was not mea- 45 field to mobility function in a sample of individuals with sureable but with perception of light were assigned distance 46 moderate to severe visual field loss. Findings will be used to and near reading acuities of 3.00logMAR.47 Contrast sensi- 47 aid the design of binocular functional field test that assesses tivity was measured binocularly with participants’ habitual 48 important areas of the visual field, as a potential clinical distance spectacle correction using a Pelli-Robson Chart48 49 tool for the low vision assessment. at 1 m scored on a letter by letter basis.49 Participants with 50 no measurable CS function were assigned a score of 51 0.00logCS.47 Methods 52 Binocular visual field assessments were performed using 53 Participants with self-reported peripheral visual field loss the Humphrey Field Analyser50 utilising the monocular test 54 were recruited. Individuals with conditions not primarily strategy for the right eye with a standard size III Goldmann 55 affecting peripheral visual function, such as macular degen- white target. The chin rest was positioned as far right as 56 eration, were excluded from the study, along with those possible and the left hand side of the chin rest was used. 57 58 59 2 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open Page 30 of 39
H Subhi et al. Functional visual fields 1 2 3 Table 1. Descriptive statistics of the variables assessed (n = 52). The Table 1 (continued) 4 mean Æ S.D., and the median (interquartile range) are given for the Demographic variables 5 clinical visual function variables n 6 Demographic variables Have you fallen in the past 12 months?, (%) 7 Yes 23 (44) 8 Gender, n (%) No 29 (56) Male 31 (60) 9 Mean Median Female 21 (40) 10 (ÆS.D.) (25% IQ–75% IQ) Range Age (years) 11 – Median 61 (49 68) Clinical function variables 12 – (25% IQ 75% IQ) Binocular distance +0.34 (Æ0.09) +0.07 (À0.07–0.46) À0.22–3.00 13 – – Min max 31 96 visual acuity 14 Ocular diagnosis, n (%) For peer review(logMAR) only 15 RP 21 (40) Binocular contrast +1.44 (Æ0.08) +1.63 (1.20–1.95) 0.00–1.95 16 Glaucoma 22 (42) sensitivity 17 Vascular 2 (4) (logCS units) 18 occlusion Binocular reading +0.50 (Æ0.12) +0.18 (0.02–0.41) À0.13–3.00 19 Retinal 2 (4) acuity (logMAR) detachments/tears 20 Binocular visual field variables Other 5 (10) Æ – – 21 Overall visual field 11.7 ( 1.4) 11.8 (2.1 20.7) 0.0 27.1 Duration of visual impairment (years) (0–60°; mean 22 – Median 15 (6 26) threshold, dB) – 23 (25% IQ 75% IQ) Central visual field 14.1 (Æ1.6) 13.0 (3.1–24.1) 0.0–31.8 – – 24 Min max 1 63 (0–30°; mean n 25 Registration status, (%) threshold, dB) 26 Registered severely 22 (42) Peripheral visual field 8.7 (Æ1.1) 7.5 (0.0–15.5) 0.0–23.0 27 sight impaired (30–60°; mean Registered sight 6 (12) 28 threshold, dB) impaired 29 Superior visual field 11.1 (Æ1.3) 10.2 (2.0–19.2) 0.0–27.6 Not registered 24 (46) (0–60°; mean 30 n Living arrangements, (%) threshold, dB)
31 Alone 14 (27) Inferior visual field 12.2 (Æ1.4) 8.9 (1.9–22.8) 0.0–28.3 http://bmjopen.bmj.com/ 32 With partner 33 (63) (0–60°; mean 33 With other 4 (8) threshold, dB) 34 Warden assisted 1 (2) Dutch Activity Inventory Person measures n 35 Current employment status, (%) Overall self-reported 2.09 (Æ0.26) 1.48 (0.72–3.40) À0.52–6.00 36 Working full time 16 (31) function (logits) Working part time 5 (10) Self-reported 2.81 (Æ0.46) 1.66 (0.21–5.79) À4.19–7.27 37 Student 3 (6) 38 mobility Unemployed 1 (2) function (logits) 39 Retired 26 (50) on September 26, 2021 by guest. Protected copyright. 40 Number of prescribed medications (n) *Number of comorbid conditions from a list of 12 common medical 41 Median 2 (0.5–2.5) conditions representing general health status.56 42 (25% IQ–75% IQ) 43 Min-max 0–11 n Since implementing monocular tests binocularly using the 44 Number of co-morbidities* ( ) Median 2(1–3) HFA invalidates conventional methods of fixation monitor- 45 (25% IQ–75% IQ) ing51 participant’s fixation was monitored manually.13,30,52 46 Min–max 0–5 Other reliability indices provided by the HFA, including 47 n Use of mobility aids, (%) false positives and false negatives, were also reviewed. The 48 White cane or 20 (38) test was stopped, the participant reinstructed, and a new 49 guide dog test started if during the first attempt false negative or false 50 No mobility 32 (62) 51 aids used positive responses exceeded 50%, or if poor fixation was 52 Use of low vision aids, n (%) 32 (62) observed by the practitioner. The subsequent test attempt 53 Yes 23 (44) was not interrupted if poor reliability indices or poor No 29 (56) 54 fixation was observed. All cases were used in subsequent 55 (continued) analyses. 56 57 58 59 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 3 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
Page 31 of 39 BMJ Open
Functional visual fields H Subhi et al. 1 2 3 The SITA Fast 30-2 threshold test was used to evaluate Analyses 4 binocular central visual field to 30° eccentricity with 76 5 points spaced six degrees apart (Figure 1). The SITA Fast The absolute threshold values from the SITA Fast 30-2 and 6 60-4 threshold test was used to assess the binocular periph- 60-4 tests were used to calculate mean threshold of the cen- 7 eral field from 30 to 60° eccentricity with 60 points spaced tral (0–30°) and the peripheral (30–60°), and superior and 8 12 degrees apart (Figure 1). These grid patterns were cho- inferior visual field. 9 sen as they are familiar to clinicians and arguably a gold Responses to the D-AI were converted to person 10 standard pattern for visual field assessment used for diag- measures using Rasch analysis (see Supporting Information 11 nostic purposes. For the 30-2 assessment, near correction for details), where higher person measures indicate greater 12 adapted from the habitual distance correction was provided perceived ability. To determine overall self-reported func- 13 in full aperture trial lenses used in adult half-eye trial tion, responses to all 43 goals were assessed. Responses to 14 frames with lens centrationFor corrected peer for near. Thereview 60-4 the four goals only of the mobility domain (mobility at home, 15 test was performed uncorrected to minimise the possibility mobility indoors, mobility outdoors, and using public 16 of lens and frame artefacts. The purpose of the study was to transport) were used to determine self-reported mobility 17 assess functional visual fields in the most habitual form that function and were Rasch analysed in isolation. 18 was appropriate. Whilst it might be considered that the use To investigate the relationship between the predictor 19 of the refractive correction normally used for mobility (in- variables and self-reported function Mann–Whitney U tests 20 cluding multifocals) would be most habitual, the perimeter were conducted for the dichotomous predictors to establish 21 used had a working distance of 33 cm and therefore assess- whether the means of the independent samples significantly 22 ment with correction incorporating a distance element differed, Kruskal–Wallis tests were performed on the nomi- 23 would have underestimated sensitivity. Instead, the estab- nal/categorical data as a non-parametric determination of 24 lished HFA protocols of using near correction for the cen- differences between the independent groups, and non-para- 25 ° – ° 26 tral 30 and no correction in the 30 60 field were used. It metric two-tailed Spearman’s rho bivariate correlations 27 should therefore be noted that the results will not reflect were conducted to investigate the relationship between the 28 variations in visual field sensitivity resulting from refractive continuous predictor variables and self-reported visual 29 correction, such as rim artefacts or near blur from progres- function. A Bonferroni corrected significance level of = = 30 sive addition lenses. p 0.0025 ( 0.05/20) was used.
31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 on September 26, 2021 by guest. Protected copyright. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
55 Figure 1. Threshold values representing the median threshold of the sample (n = 52) at each test location for (a) central 30-2, and (b) peripheral 56 60-4 tests. 57 58 59 4 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open Page 32 of 39
H Subhi et al. Functional visual fields 1 2 3 Table 2. Relationship between the variables assessed, and self- Collinearity statistics were assessed to determine whether 4 reported overall and mobility function scores for different visual field areas were independent. 5 Overall Mobility These measures included the tolerance and variance infla- 6 Demographic variables D-AI score function tion factor statistics. Variance inflation factors (VIF) 7 greater than 10,53 and a tolerance statistic below 0.154,55 8 Dichotomous variables (U) U = U = would indicate a multi-collinearity bias. 9 Gender 286.00, 275.00, p = p = 10 0.52 0.39 Use of mobility aids *U = 112.50, *U = 101.50, Results 11 p < 0.001 p < 0.001 12 Use of low vision aids *U = 107.50, U = 166.50, Fifty-two participants took part, and Table 1 illustrates the 13 p < 0.001 p = 0.020 descriptive statistics for the parameters assessed. All partici- 14 Have you fallen in the U = 208.50, U = 225.00, pants were able to complete both visual field tests binocu- Forp = peerp = review only 15 past 12 months? 0.020 0.046 larly. Twelve percent of participants had difficulty either Nominal variables (v²) 16 seeing the standard fixation target or maintaining single 17 Ocular diagnosis v²=13.57, v²=15.35, p = p = vision during the assessment. For these participants, a cus- 18 0.009 0.004 Living arrangements v²=0.98, v²=1.79, tom fixation target consisting of a black 2 mm high con- 19 p = 0.81 p = 0.62 trast pericentral ring around the fixation spot was used to 20 Current v²=2.71, v²=2.00, aid fixation. The SITA Fast 30-2 assessment took on aver- 21 employment status p = 0.61 p = 0.74 age 5 min 14 s (Æ11 s), and the 60-4 took on average 22 v²= v²= Sight loss registration * 26.92, * 23.66, 4 min 59 s (Æ7 s). The median threshold values of the 23 p < 0.001 p < 0.001 2 sample for each location within the central 30-2 and 24 Continuous variables (R ) 2 2 peripheral 60-4 test programmes are demonstrated in Fig- 25 Age R = 0.03, R = 0.05, 26 p = 0.23 p = 0.10 ure 1. R2 = R2 = 27 Duration of 0.16, * 0.23, Table 2 shows the relationships between the parameters visual impairment p = 0.004 p < 0.001 assessed and the outcome measures of overall self-reported 28 2 2 Number of medications R = 0.00, R = 0.01, function and self-reported mobility function. 29 p = p = 0.50 0.52 Self-reported overall and mobility related function were 30 Number of R2 = 0.12, R2 = 0.07, p = p = significantly worse for individuals who reported using 31 comorbidities 0.015 0.060 http://bmjopen.bmj.com/ = 32 Clinical function variables (R2) mobility aids than for those who did not (U 112.50, 2 2 < = < 33 Distance visual acuity *R = 0.52, *R = 0.40, p 0.0001 and U 101.50, p 0.0001 respectively), and 34 (logMAR) p < 0.001 p < 0.001 for those with more severe visual impairment as indicated R2 = R2 = 2 35 Contrast sensitivity * 0.52, * 0.38, by their visual registration status (R = 0.50 p < 0.0001; (logCS units) p < 0.001 p < 0.001 R2 = 0.45 p < 0.0001 respectively). Overall self-reported 36 R2 = R2 = Binocular reading * 0.54, * 0.42, function was more difficult for participants who reported 37 acuity (logMAR) p < 0.001 p < 0.001 2 using low vision aids than those who did not (U = 107.50, 38 Binocular visual field variables (R ) < 39 Overall field (0–60°)*R2 = 0.50, *R2 = 0.64, p 0.0001). Longer duration of visual impairment was a on September 26, 2021 by guest. Protected copyright. 40 p < 0.001 p < 0.001 significant predictor of worse self-reported mobility func- 41 Central field (0–30°)*R = 0.49, *R2 = 0.61, tion (R2 = 0.23, p < 0.0001). Acuity and contrast sensitiv- 42 p < 0.001 p < 0.001 ity variables all correlated significantly (p ≤ 0.0025) with 2 2 43 Peripheral *R = 0.48, *R = 0.63, overall and mobility related self-reported visual function, – ° p < p < 44 field (30 60 ) 0.001 0.001 where worse visual function measures were associated with Superior visual *R2 = 0.41, *R2 = 0.56, 45 worse perceived function. field (0–60°) p < 0.001 p < 0.001 – ° 46 Inferior visual *R2 = 0.55, *R2 = 0.67, Greater visual field loss (0 60 ) was associated with 2 47 field (0–60°) p < 0.001 p < 0.001 worse self-reported overall function (R = 0.50; 48 p < 0.0001). Worse overall binocular visual field was in p ≤ p > 49 * 0.0025, for all others 0.0025. particular a good predictor of greater self-reported diffi- 50 culty in mobility related activities (R2 = 0.64, p < 0.0001). 51 To allow for the prediction of self-reported visual func- Splitting the visual field into central (0–30°) and periph- 52 tion by a linear combination of two or more predictor vari- eral (30–60°) areas, the relationship between the visual field 53 ables, and to explore the unique variance explained by each and self-reported function did not appear to be greatly 54 predictor variable, clinical function variables were entered dependent on eccentricity. The peripheral (30–60°) and 55 into the regression model in a forward stepwise manner central (0–30°) visual field were similarly correlated with 56 using an alpha of 0.05. self-reported function for both overall function (R2 = 0.49, 57 58 59 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 5 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
Page 33 of 39 BMJ Open
Functional visual fields H Subhi et al. 1 2 3 p < 0.0001 central and R2 = 0.48, p < 0.0001 peripheral) Table 3. Results of stepwise regression analyses to determine which 4 and mobility related function (R2 = 0.61, p < 0.0001 cen- clinical function variables best represent overall self-reported function, n = – ° 5 tral and R2 = 0.63, p < 0.0001 peripheral), where greater and mobility function ( 52). Variables included: (a) central (0 30 ) and peripheral (30–60°) visual field, contrast sensitivity, distance visual 6 visual field loss is associated with worse perceived function. acuity; (b) superior and inferior visual field, contrast sensitivity, distance 7 The superior and inferior visual field areas were also simi- visual acuity 8 larly related to overall (R2 = 0.41, p < 0.0001 superior and 2 9 R2 = 0.56, p < 0.0001 inferior) and mobility related func- B SE B b R change p 10 2 = < 2 = tion R 0.55, p 0.0001 superior and R 0.67, (a) 11 p < 0.0001 inferior), but with a tendency for the inferior Overall D-AI score 12 fields to be better correlated to perceived function. Peripheral (30–60°) field 0.12 0.03 0.51 0.50 <0.001 13 Although there was a strong correlation between the cen- Binocular CS 1.11 0.33 0.36 0.09 0.002 2 14 tral (0–30°) and theFor peripheral (30peer–60°) visual field review scores R only0.59 15 (R2 = 0.85, p < 0.0001), and between the superior and Mobility function Peripheral (30–60°) field 0.24 0.04 0.58 0.59 <0.001 16 inferior visual field scores (R2 = 0.85, p < 0.0001), further 17 Binocular CS 1.83 0.53 0.34 0.08 0.001 investigation suggested these relationships were unlikely to R2 18 0.67 adversely affect the results obtained by including both vari- (b) 19 ables in subsequent multiple regression analyses due to Overall D-AI function 20 multi-collinearity. Variance inflation factors were lower Inferior visual field 0.10 0.02 0.56 0.54 <0.001 21 than required at 5.85 and 4.2953 and the tolerance statistics Binocular CS 0.99 0.33 0.32 0.07 0.004 22 R2 of 0.17 and 0.23 were greater than required.54,55 Therefore, 0.61 23 Mobility function although thresholds are related to one another, the relative 24 Inferior visual field 0.19 0.03 0.61 0.61 <0.001 roles of the central (0–30°) and peripheral (30–60°) field, 25 Binocular CS 1.68 0.53 0.31 0.07 0.002 2 26 and of the superior and inferior field in reflecting func- R 0.68 tional difficulty can be appropriately investigated with 27 b regression analyses. B, unstandardised regression coefficients; SE B, standard errors; , stan- 28 dardised regression coefficients; R2 change, amount of additional vari- The variables of central (0–30°) and peripheral (30–60) 29 ance by including predictors from sample. 30 visual field mean thresholds, binocular visual acuity, and binocular contrast sensitivity were entered into stepwise order to best reflect self-reported function. Both central
31 http://bmjopen.bmj.com/ – ° – ° 32 multiple regressions to determine which independently (0 30 ) and peripheral (30 60 ) mean thresholds related 33 explained significant amounts of variance in overall and well to self-reported function (Table 2), but it was the 34 mobility self-reported function (Table 3a). The peripheral peripheral (30–60°) field that was the best predictor of both 35 (30–60°) visual field was found to account for most vari- overall and mobility related self-reported function in this 36 ance (50%) in overall self-reported function. Also signifi- sample (Table 3a). Therefore, in order to accurately deter- 37 cant in this model was binocular contrast sensitivity, which mine the functional consequences of visual field loss, the 38 explained a further 9% of variance. For self-reported peripheral visual field should also be considered (Table 4). 39 mobility function, the peripheral (30–60°) visual field That the peripheral visual field is important for mobility on September 26, 2021 by guest. Protected copyright. 40 explained 59% of the variance in self-reported mobility function is supported by previous findings using several 41 function. When combined with binocular contrast sensitiv- different outcome measures. In a sample of older adults, 42 ity this increased to 67%. the status of peripheral visual field (20–60°) was signifi- 43 Further stepwise regression analyses were conducted cantly correlated with the risk of falling, whereas the central 44 including the superior and inferior visual field mean visual field (0–20°) was not.57 Also, correcting for central 45 thresholds, binocular visual acuity, and binocular contrast visual impairment alone may be insufficient to effectively 46 sensitivity (Table 3b). The inferior visual field explained decrease rates of falls relating to visual impairment.58 47 54% of overall self-reported function, and 61% when com- Monocular kinetic visual field extent in people with Retini- 48 bined with binocular contrast sensitivity. The inferior visual tis Pigmentosa has also been found to be significantly asso- 49 field was also found to account for most variance (61%) in ciated with mobility function as assessed on a mobility 50 self-reported mobility function. Binocular contrast sensitiv- course, with worse function in subjects with fields con- 51 ity explained a further 7% of variance. tained within the central 20°.59 Orientation and mobility 52 performance has been shown to be worse when the visual 53 field loss is peripheral than when the loss is central,60 with Discussion 54 an increased risk of tripping over obstacles also associated 55 The aim of this study was to investigate which areas of the with peripheral visual impairment.28 The relationship 56 binocular threshold visual field should be measured in between visual field loss and function is also likely to be 57 58 59 6 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open Page 34 of 39
H Subhi et al. Functional visual fields 1 2 3 influenced by compensatory scanning behaviour used by results in reduced step length when walking across 4 individuals with visual field loss which may improve task uneven terrain.39 – 5 performance.61 64 It should be noted that previous studies demonstrating 6 Other studies have indicated that the central visual field the importance of the inferior visual field for mobility func- 7 is more strongly related to mobility function, however. The tion have assessed subjects whose degree of visual field loss 8 field of view required for navigation in subjects with artifi- was likely less than that of many of the participants in our 9 cially restricted fields is between 10.9 and 32.1° depending sample. Over half of the current sample were registered 10 on contrast conditions.34 It has been proposed that the cen- sight impaired (Table 1), and although this may relate to 11 tral 37° is most important for mobility function in individ- visual acuity rather than fields, individuals with visual field 12 uals with low vision,29 and Tabrett and Latham13 found the loss were recruited, and Table 1 shows a relatively good 13 central 10–30° best predicted visual related activity limita- median acuity, suggesting visual field loss was the primary 14 tion in mobility tasksFor (although fields peer were not assessed reviewreason for registration. only Previous studies have evaluated 15 beyond 30°). In subjects with glaucoma, perceived mobility older adults with no visual impairment,28,29 normally 16 function was best explained by the function of the inferior sighted subjects with simulated field loss,39 participants 17 5° from fixation.65 with glaucoma,36,37 and a mixed sample with half the par- 18 The variance in findings of these studies is likely attribu- ticipants having visual impairment.29 In the present study 19 ted to differences in the methods of assessing the visual we have demonstrated that loss in the inferior visual field 20 field, and how mobility is assessed. Several previous studies remains a better indicator of perceived mobility function 21 have assessed the visual field binocularly using kinetic para- than the superior field in individuals with a greater degree 22 digms that include the peripheral visual field,28,29,31,32 while of established visual impairment. 23 others have used threshold techniques restricted to the cen- While the association between the inferior visual field 24 tral 30° visual field.13,30,52 Only the present study has con- and mobility is well documented, the significance of the 25 sidered threshold fields beyond 30°. In terms of mobility, inferior visual field to overall function has not previously 26 57,58 27 some studies have used falls as an outcome measure, been investigated. The inferior visual field was also selected others have used objective performance on mobility as the primary predictor of overall perceived function in 28 – courses,28 34 while others including the present study have the present study, indicating the significance of the inferior 29 1,13,21,22,24 30 used questionnaires to assess perceived function. visual field for general function. This could be explained by Since it has been suggested that the different areas of the the presence of more ecologically relevant information in
31 http://bmjopen.bmj.com/ 68 32 visual field are used for different purposes in mobility, with this region of space. 33 the central field used to guide walking and the peripheral A further predictor of overall and mobility self-reported 34 field used to establish and update an accurate representa- function was binocular contrast sensitivity. This supports 66 35 tion of spatial structure for navigation, relationships previous research that has shown that while visual acuity, 36 between field and mobility may depend on the specific visual field, and contrast sensitivity correlate significantly 37 mobility issues that the outcome measure used taps into. with mobility performance, the visual field and contrast – 38 The results of this study also support the significance sensitivity are stronger predictors than visual acuity.60,69 71 39 of the inferior visual field for mobility function that has The combined effect of visual field and contrast sensitivity – on September 26, 2021 by guest. Protected copyright. 40 been previously demonstrated.28,29,36 39 Visual field loss in other studies of low vision groups has been shown to 41 in the inferior mid-periphery (20–40°) has been shown account for between 39% and 64% of the variance in mea- 42 to adversely affect mobility more than loss of the visual sured mobility performance,32,59,72 similar to the 59–67% 43 field in other areas.29 Similarly, visual field loss in the found here. 44 lower peripheral region has shown comparable decre- Whilst this study assesses binocular static thresholds to 45 ments in walking speed on a mobility course.28,39 The 60° eccentricity, previous studies have assessed the binocu- 46 inferior visual field has also been shown to relate signifi- lar visual field past 30° with supra-threshold or kinetic 47 cantly to perceived mobility function.36 One potential paradigms. In people with glaucoma, the Esterman supra- 48 reason for the increased importance of inferior field in threshold field test has been shown to relate well to self- 49 mobility is that the inferior field may provide a stronger reported function in some studies23,24 but not in 50 contribution to postural stability than the superior visual others.22,25,26 Other studies have found an association 51 field.37 It has also been suggested that the inferior visual between the visual field assessed kinetically and self- – 52 field contributes a greater proportion of the visual infor- reported function,1,31 and mobility performance.29,31 33 53 mation used in determining lower limb movements, foot Whether a binocular threshold visual field is a more appro- 54 placement, and obstacle detection.39 Individuals tend to priate way to assess functional field loss is not clear as few 55 fixate approximately two steps ahead when walking,67 studies to date have compared different visual field strate- 56 but loss of information from the lower visual field gies and their ability to reflect functional loss. One study 57 58 59 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 7 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
Page 35 of 39 BMJ Open
Functional visual fields H Subhi et al. 1 2 3 did not find significant differences in how well different References 4 visual field protocols related to self-reported function,73 5 whereas another found that of several visual field protocols 1. Bibby SA, Maslin ER, McIlraith R & Soong GP. Vision and 6 only the Esterman field score correlated with self-reported self-reported mobility performance in patients with low 7 function.74 It is yet to be determined whether a threshold vision. Clin Exp Optom 2007; 90: 115–123. 8 binocular visual field test assessing both central and periph- 2. Nelson P, Aspinall P, Papasouliotis O, Worton BO & Brien 9 eral field could yield stronger correlation with functional C. Quality of life in glaucoma and its relationship with – 10 ability than other methods of visual field assessment, visual function. J Glaucoma 2003; 12: 139 150. 11 although the extent of variance explained in the present 3. Tabrett DR & Latham K. Factors influencing self-reported 12 study is promising. vision-related activity limitation in the visually impaired. – 13 In the present study, a cohort with a wide range of visual Invest Ophthalmol Vis Sci 2011; 52: 5293 5302. 14 field sensitivities from mild to profound loss were assessed 4. Ramrattan RS, Wolfs RC, Panda-Jonas S et al. Prevalence For peer reviewand caused only of visual field loss in the elderly and associations 15 in order to examine the relationship between field loss and with impairment in daily functioning: the Rotterdam Study. 16 functional ability. In future work, a larger sample and Arch Ophthalmol 2001; 119: 1788–1794. 17 inclusion of a normally sighted cohort may be needed to 5. Sherwood M, Garcia-Siekavizza A, Meltzer M et al. Glau- 18 provide robust regressions and to examine the ability of 19 coma’s impact on quality of life and its relation to clinical any proposed functional field test to discriminate between 20 indicators. Ophthalmology 1998; 105: 561–566. people with and without field loss. Alternative considera- 21 6. Ivers RQ, Cumming RG, Mitchell P et al. Visual impair- tions of multi-colinearity in the data might also need to be 22 ment and falls in older adults: the Blue Mountains Eye considered. The use of a single self-reported outcome pro- – 23 Study. J Am Geriatr Soc 1998; 46: 58 64. vides a broad indication of general mobility function, but 24 7. Nelson-Quigg JM, Cello K & Johnson CA. Predicting binoc- does not explore specific mobility tasks. These broad 25 ular visual field sensitivity from monocular visual field – 26 mobility goals are however less likely to be biased by field results. Invest Ophthalmol Vis Sci 2000; 41: 2212 2221. 27 area-specific tasks such as tripping over obstacles or walk- 8. Schneck ME, Haegerstom-Portnoy G, Lott L & Brabyn JA. 28 ing into overhanging objects. Monocular vs. binocular measurement of spatial vision in – 29 In conclusion, peripheral and central fields both have a elders. Optom Vis Sci 2010; 87: 526 531. 30 role in reflecting the functional difficulties of people with 9. Asaoka R, Crabb D, Yamashita T, Russell R, Wany YX & field loss and should be considered in a functional visual Garway-Heath DF. Patients have two eyes!: binocular versus
31 http://bmjopen.bmj.com/ 32 field assessment. The significance of the inferior field to better eye visual field indices. Invest Ophthalmol Vis Sci 2011; 52: 7007–7011. 33 both mobility function and overall function has also been 10. Crabb DP, Smith ND, Glen FC, Burton R & Garway-Heath 34 demonstrated in individuals with moderate to severe visual DF. How does glaucoma look? Patient perception of visual 35 field loss. A binocular threshold test that assesses to 60° can field loss. Ophthalmology 2013; 120: 1120–1126. 36 represent the functional abilities of individuals with periph- 11. Gutierrez P, Wilson MR, Johnson C et al. Influence of glau- 37 eral visual impairment. However, what remains to be deter- comatous visual field loss on health-related quality of life. 38 mined is whether a threshold method is preferable to Arch Ophthalmol 1997; 115: 777–784. 39 alternative paradigms such as supra-threshold or kinetic 12. El-Gasim M, Munoz B, West S & Scott A. Associations on September 26, 2021 by guest. Protected copyright. fields in producing an outcome that can be used clinically 40 between self-rated vision score, vision tests and self-reported and best describes functional difficulty. Further research 41 visual function in the Salisbury Eye Evaluation Study. Invest will determine the most appropriate method of assessing 42 Ophthalmol Vis Sci 2013; 54: 6439–6445. 43 visual fields in the low vision assessment by comparing 13. Tabrett DR & Latham K. Important areas of the central 44 different visual field protocols. binocular visual field for daily functioning in the visually 45 impaired. Ophthalmic Physiol Opt 2012; 32: 156–163. 46 Disclosure 14. Aspinall PA, Johnson ZK, Azuara-Blanco A, Montarzino A, 47 Brice R & Vickers A. Evaluation of quality of life and priori- 48 The authors report no conflicts of interest and have no pro- ties of patients with glaucoma. Invest Ophthalmol Vis Sci 49 prietary interest in any of the materials mentioned in this 2008; 49:1907–1915. 50 article. 15. Seo JH, Yu HG & Lee BJ. Assessment of functional vision 51 score and vision-specific quality of life in individuals with 52 retinitis pigmentosa. Korean J Ophthalmol 2009; 23: 164– Acknowledgement 53 168. 54 This research is funded by a College of Optometrists Post- 16. Varma R, Wu J, Chong K, Azen SP, Hays RD; Los Angeles 55 graduate Research Scholarship. Latino Eye Study Group. Impact of severity and bilaterality 56 57 58 59 8 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
BMJ Open Page 36 of 39
H Subhi et al. Functional visual fields 1 2 3 of visual impairment on health-related quality of life. Oph- 33. Lovie-Kitchin JE, Soong GP, Hassan SE & Woods RL. Visual 4 thalmology 2006; 113: 1846–1853. field size criteria for mobility rehabilitation referral. Optom 5 17. Nelson P, Aspinall P & O’Brien C. Patient’s perception of Vis Sci 2010; 87: 948–957. 6 visual impairment in glaucoma: a pilot study. Br J Ophthal- 34. Hassan SE, Hicks JC, Lei H & Turano KA. What is the mini- 7 mol 1999; 83: 546–552. mum field of view required for efficient navigation? Vision 8 18. Szlyk JP, Fishman GA, Alexander KR et al. Relationship Res 2007; 47: 2115–2123. 9 between difficulty in performing daily activities and clinical 35. Guerin E, Bouliotis G & King A. Visual impairment registra- 10 measures of visual function in patients with retinitis pig- tion: evaluation of agreement among ophthalmologists. Eye 11 mentosa. Arch Ophthalmol 1997; 115: 53–59. 2014; 28: 808–813. 12 19. Szlyk JP, Fishman GA, Grover S, Revelins BI & Derlacki DJ. 36. Black AA, Wood JM & Lovie-Kitchin JE. Inferior field loss 13 Difficulty in performing everyday activities in patients with increases rate of falls in older adults with glaucoma. Optom 14 juvenile macular dystrophies: comparison with patients with Vis Sci 2011; 88: 1275–1282. 15 retinitis pigmentosa. BrFor J Ophthalmol peer1998; 82: 1372–1376. review37. Black AA, Wood only JM, Lovie-Kitchin JE & Newman BM. 16 20. Parrish RK II, Gedde SJ, Scott IU et al. Visual function and Visual impairment and postural sway among older adults 17 quality of life among patients with glaucoma. Arch Ophthal- with glaucoma. Optom Vis Sci 2008; 85: 489–497. – 18 mol 1997; 115: 1447 1455. 38. Coleman AL, Cummings SR, Yu F et al. Binocular visual- 19 21. Crabb DP & Viswanathan AC. Integrated visual fields: a new field loss increases the risk of future falls in older white – 20 approach to measuring the binocular field of view and visual women. J Am Geriatr Soc 2007; 55: 357 364. – 21 disability. Graefes Arch Clin Exp Ophthalmol 2004; 243: 210 39. Marigold DS & Patla AE. Visual information from the lower 22 216. visual field is important for walking across multi-surface ter- – 23 22. Noe G, Ferraro J, Lamoureux E, Rait J & Keeffe JE. Associa- rain. Exp Brain Res 2008; 188: 23 31. 24 tions between glaucomatous visual field loss and participa- 40. Bruijning JE, van Nispen RMA & van Rens GHMB. A Dutch tion in activities of daily living. Clin Exp Optom 2003; 31: ICF version of the Activity Inventory: results from a pilot 25 – 26 482 486. study among visually impaired persons. BMC Health Serv 27 23. Mills RP & Drance SM. Esterman disability rating in severe Res 2010; 10: 318. glaucoma. Ophthamology 1986; 93: 371–378. 41. Bruijning JE, van Rens GHMB, Knol DL & van Nispen 28 24. Fujita K, Yasuda N, Nakamoto K & Fukuda T. The relation- RMA. Psychometric analyses to improve the Dutch ICF 29 ship between difficulty in daily living and binocular visual Activity Inventory. Optom Vis Sci 2013; 90: 806–819. 30 field in patients with glaucoma. Nihon Ganka Gakkai Zasshi 42. World Health Organisation. International Classification of
31 http://bmjopen.bmj.com/ 2008; 112: 447–450. Functioning, Disability and Health, World Health Organisa- 32 25. Jampel HD. Glaucoma patients’ assessment of their visual tion: ICF, Geneva, Switzerland, 2001. 33 function and quality of life. Tr Am Ophth Soc 2001; 99: 301– 43. Latham K, Baranian M, Timmis MA & Pardhan S. Difficul- 34 317. ties with goals of the Dutch ICF Activity Inventory: percep- 35 26. Jampel HD, Friedman DS, Quigley H et al. Correlation of tions of those with retinitis pigmentosa and of those who 36 the binocular visual field with patients’ assessment of vision. support them. Invest Ophthalmol Vis Sci 2015; 56: 2381– 37 Invest Ophthalmol Vis Sci 2002; 43: 1059–1067. 2391. 38 27. Turano KA, Rubin GS & Quigley HA. Mobility perfor- 44. Ferris FL, Kassoff A, Bresnick GH & Bailey IL. New visual 39 mance in glaucoma. Invest Ophthalmol Vis Sci 1999; 40: acuity charts for clinical research. Am J Ophthalmol 1982; on September 26, 2021 by guest. Protected copyright. 40 2803–2309. 94: 91–96. 41 28. Turano KA, Broman AT, Bandeen-Roche K, Munoz B, 45. Mansfield J, Ahn SJ, Legge GE & Luebker A. A new reading- 42 Rubin GS & West S. Association of visual field loss and acuity chart for normal and low vision. Ophthalmic & visual 43 mobility performance in older adults: Salisbury Eye Evalua- optics/noninvasive assessment of the visual system. OSA 44 tion Study. Optom Vis Sci 2004; 81: 298–307. Tech Digest 1993; 3: 232–235. 45 29. Lovie-Kitchin JE, Mainstone J, Robinson J & Brown B. What 46. Ahn SJ, Legge GE & Luebker A. Printed cards for measuring 46 areas of the visual field are important for mobility in low low-vision reading speed. Vision Res 1995; 35: 1939–1944. 47 vision patients? Clin Vis Sci 1990; 5: 249–263. 47. Myint J, Latham K, Mann D, Gomersall P, Wilkins A & 48 30. Black A, Lovie-Kitchin J, Woods R, Arnold N, Byrnes J & Allen P. The relationship between visual function and 49 Murrish J. Mobility performance with retinitis pigmentosa. performance in rifle shooting for athletes with vision impair- 50 Clin Exp Optom 1996; 80: 1–12. ment. BMJ Open Sport Exerc Med 2016; 2(1): e000080. 51 31. Haymes SA, Johnston AW & Heyes AD. Relationship 48. Pelli DG, Robson JG & Wilkins AJ. The design of a new let- 52 between vision impairment and ability to perform activities ter chart for measuring contrast sensitivity. Clin Vis Sci 53 of daily living. Ophthalmic Physiol Opt 2002; 22: 79–91. 1988; 2: 187–199. 54 32. Haymes S, Guest D, Heyes A & Johnston A. Mobility of peo- 49. Elliott DB, Bullimore MA & Bailey IL. Improving the relia- 55 ple with retinitis pigmentosa as a function of vision and psy- bility of the Pelli-Robson contrast sensitivity test. Clin Vis 56 chological variables. Optom Vis Sci 1996; 73: 621–637. Sci 1991; 6: 471–475. 57 58 59 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 9 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from
Page 37 of 39 BMJ Open
Functional visual fields H Subhi et al. 1 2 3 50. HFA. Carl Zeiss Meditec, Inc., Dublin, CA. 65. Sumi I, Shirato S, Matsumoto S & Araie M. The relation- 4 51. Heijl A & Krakau C. An automatic perimeter for glaucoma ships between visual disability in patients with glaucoma. 5 visual field screening and control. Graefes Arch Clin Exp Ophthalmology 2003; 110: 332–339. 6 Ophthalmol 1975; 197: 13–23. 66. Turano KA, Yu D, Hao L & Hicks JC. Optic-flow and ego- 7 52. Leat SJ & Lovie-Kitchin J. Visual impairment and the useful centric-direction strategies in walking: central vs peripheral 8 field of vision. Ophthalmic Physiol Opt 2006; 26: 392–403. visual field. Vision Res 2005; 45: 3117–3132. 9 53. Field AP. Discovering Statistics Using SPSS, 2nd edn. SAGE 67. Land MF. Eye movements and the control of actions in 10 publications Ltd.: London, 2005. everyday life. Prog Retin Eye Res 2006; 25: 296–324. 11 54. Menard S. Applied Logistic Regression Analysis: Sage Univer- 68. Rezec AA & Dobkins KR. Attentional weighting: a possible 12 sity Series on Quantitative Applications in the Social Sciences. account of visual field asymmetries in visual search? Spat Vis 13 Sage: Thousand Oaks, CA, 1995. 2004; 17: 269–293. 14 55. O’Brien RM. A caution regarding rules of thumb for vari- 69. Bailey IL. New procedures for detecting early vision losses in 15 ance inflation factors.ForQual Quant peer2007; 41: 673–690. reviewthe elderly. Optomonly Vis Sci 1993; 70: 299–305. 16 56. van Nispen RMA, Hoeijmakers JGJ, de Boer MR et al. 70. Kuyk T, Elliott JL & Fuhr PS. Visual correlates of mobility in 17 Agreement between self-reported co-morbidity of visually real world settings in older adults with low vision. Optom – 18 impaired older patients and reports from their general prac- Vis Sci 1998; 75: 538 547. – 19 titioners. Visual Impair Res 2008; 10: 49 56. 71. Hassan SE, Lovie-Kitchin JE & Woods RL. Vision and 20 57. Freeman EE, Munoz B, Rubin GS & West SK. Visual field mobility performance of subjects with age-related macular – 21 loss increases the risk of falls in older adults: the Salisbury eye degeneration. Optom Vis Sci 2002; 79: 697 707. – 22 evaluation. Invest Ophthalmol Vis Sci 2007; 48: 4445 4450. 72. Long RG, Reiser JJ & Hill EW. Mobility in individuals with 23 58. Patino CM, McKean-Cowdin R, Azen SP, Allison JC, moderate visual impairments. J Vis Impair Blind 1990; 84: – 24 Choudhury F & Varma R. Central and peripheral visual 111 118. 25 impairment and the risk of falls and falls with injury. Oph- 73. Choy ES, Mills RP & Drance SM. Automated Esterman test- 26 thalmology 2010; 117: 199. ing of disability in glaucoma. In: Seventh International Visual 27 59. Geruschat DR, Turano KA & Stahl JW. Traditional measures Field Symposium (Greve E & Heiji A, editors), Junk Publish- of mobility performance and retinitis pigmentosa. Optom ers: Amsterdam, the Netherlands, 1986; pp. 527–535. 28 Vis Sci 1998; 75: 525–537. 74. Yanagisawa M, Kato S, Kunimatsu S & Tamura M. Relation- 29 60. Marron JA & Bailey IL. Visual factors and orientation-mobi- ship between vision-related quality of life in Japanese 30 lity performance. Am J Optom Physiol Opt 1982; 59: 413–426. patients and methods for evaluating visual field. Jpn J Oph-
31 http://bmjopen.bmj.com/ 61. Pambakian ALM, Wooding DS, Patel N, Morland AB, Ken- thalmol 2011; 55: 132–137. 32 nard C & Mannan SK. Scanning the visual world: a study of 33 patients with homonymous hemianopia. J Neurol Neurosurg 34 Supporting Information Psychiatry 2000; 69: 751–759. 35 62. Zihl J. Visual scanning behavior in patients with homony- Additional Supporting Information may be found in the 36 mous hemianopia. Neuropsychologia 1995; 33: 287–303. online version of this article: 37 63. Chen CS, Lee AW, Clarke G et al. Vision-related quality of Appendix S1. 38 Calculation of Person Measures with life in patients with complete homonymous hemianopia Rasch analysis. 39 – post stroke. Top Stroke Rehabil 2009; 16: 445 453. Table S1. on September 26, 2021 by guest. Protected copyright. 40 Item parameters of the 43 goals of the Dutch 64. Loetscher T, Chen C, Wignall S et al. A study on the natural ICF Activity Inventory as determined by Rasch analysis. 41 history of scanning behaviour in patients with visual field 42 defects after stroke. BMC Neurol 2015; 15: 321–325. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 10 © 2017 The Authors Ophthalmic & Physiological Optics © 2017 The College of Optometrists 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 38 of 39 7-10 7-10 NA NA 1 1
11 11 10-11 10-11 11 11 3 3 6 6 6 6
6 6 5 5
10-11 10-11 7-10 7-10 10-11 10-11 Reported on page # Reported on page 2 2 6 6
cross-sectional studies cross-sectional
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
) Indicate ) study’sthe design commonlywith useda in term ortitlethe abstract the ) ) theGive eligibility criteria, and sourcesthe methodsand of selection of participants ) ) statisticalDescribe all methods, thoseincluding confoundingusedforcontrol to ) ) Describe any methods used subgroups examine to interactionsand ) If ) applicable, describe methodsanalytical taking account of sampling strategy ) ) in theProvide abstract informativean and balanced summary of what was donewhat wasand found ) Describe) sensitivity any analyses
For peer review only ) howExplain missing data were addressed a a a b e d c b For eachof interest,For variable sourcesdatagive of and details of methods of assessment (measurement). Describe comparability of of comparability assessment methods isthanthere more if one group ( ( why why
applicable collection collection ( ( ( ( Recommendation STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in be included should that items of Statement—Checklist (v4) 2007 STROBE
6 1 ( 3 specific objectives, State including prespecified any hypotheses 5 the Describe setting, locations, relevantand dates, including periods exposure,of recruitment, follow-up, data and 9 Describe any efforts to address sourcespotential of bias 7 define Clearly outcomes,all exposures,predictors, potential confounders,effect and modifiers. diagnostic Give criteria, if 4 key Present of elements study inthedesign early paper 8* 10 10 how Explain studythe size was arrived at # Item Item Results
Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 thebackground Explain scientific for rationale investigation the and being reported Introduction Title andTitle abstract Section/Topic
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from 17-19 17-19 12 12 NA
17-19 17-19 3 3 17-19 17-19
12 12 NA NA 11 11 NA NA 4 4 14 14 NA
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report) boundaries category when continuous variables were categorized ) ) unadjustedGive estimates if applicable, and, confounder-adjusted estimates their and precision95% (eg, confidence
For peer )relevant, If consider translating estimatesrelativerisk of into absolute for risk meaningfulperioda time review only c b a similar studies,similar and other relevant evidence confounders magnitude of magnitude potential any bias confirmed eligible,confirmed included in study, the completing follow-up, analysed and interval). Make interval). clear confounderswhich were adjusted they why and for were included which the presentthe which is based article
( ( number Indicate (b) of participants missingwith eachfor data of interest variable Consider (c) diagram use flow of a reasons Give (b) non-participationfor each at stage 18 18 Summarise results key to reference withstudy objectives 16 16 ( 22 22 sourcethe Give of funding and role the of fundersthe thefor studypresent if and, applicable, originalthe for study on 19 19 Discuss of study, the limitations taking into sourcesaccount biasof potential imprecision.or Discussboth direction and 13* 13* Report (a) numbers of individuals each at stage of study—eg numbers eligible, potentially examinedfor eligibility, An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/,at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (externalgeneralisabilitythe of validity) studythe results Interpretation 20 cautious Give interpretationa of overall resultsconsidering limitations, objectives, multiplicity of resultsanalyses, from Limitations Key Key results Discussion Other analyses 17 other Report analyses done—eg analyses of subgroups sensitivity interactions, and and analyses
Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures
Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and
Participants Page 39 of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open
Functional visual fields: A cross-sectional study to determine which visual field paradigms best reflect difficulty with mobility function
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2017-018831.R1
Article Type: Research
Date Submitted by the Author: 15-Sep-2017
Complete List of Authors: Subhi, Hikmat; Anglia Ruskin University, Vision and Hearing Sciences Latham, Keziah; Anglia Ruskin University, Vision and Hearing Sciences Myint, Joy; University of Hertfordshire School of Life and Medical Sciences Crossland, Michael; Moorfields Eye Hospital NHS Foundation Trust, Optometry
Primary Subject Ophthalmology Heading:
Secondary Subject Heading: Ophthalmology
Low vision, Visual fields, Perimetry, Mobility function, Visual impairment, http://bmjopen.bmj.com/ Keywords: Self-reported function
on September 26, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 1 of 34 BMJ Open
1 2 3 Title Functional visual fields: A cross sectional study to determine which 4 visual field paradigms best reflect difficulty with mobility function 5 Running head Functional visual fields 6 Authors *Subhi, Hikmat; Anglia Ruskin University, Department of Vision and 7 Hearing Sciences and Vision and Eye Research Unit 8 9 Latham, Keziah; Anglia Ruskin University, Department of Vision and 10 Hearing Sciences and Vision and Eye Research Unit 11 12 Myint, Joy; University of Hertfordshire, Life and Medical Sciences, 13 Postgraduate Medicine 14 15 For peer review only 16 Crossland, Michael D.; Moorfields Eye Hospital, London 17 18 *Corresponding author: [email protected] 19 Word count 4814 20 Key words Visual fields, self reported function, mobility function, functional vision, 21 visual impairment 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 2 of 34
1 2 3 Abstract 4 5 6 Objectives 7 8 9 To develop an appropriate method of assessing visual field (VF) loss which reflects its 10 11 functional consequences, this study aims to determine which method(s) of assessing visual 12 13 fields best reflect mobility difficulty. 14 15 For peer review only 16 Setting 17 18 19 20 This cross sectional observational study took place within a single primary care setting. 21 22 Participants attended a single session at a University Eye Clinic, Cambridge, UK, with data 23 24 collected by a single researcher (HS), a qualified optometrist. 25 26 27 Participants 28 29 30 50 adult participants with peripheral field impairment were recruited for this study. Individuals 31
32 with conditions not primarily affecting peripheral visual function, such as macular http://bmjopen.bmj.com/ 33 34 degeneration, were excluded from the study. 35 36 37 Primary and secondary outcome measures 38 39
40 on September 26, 2021 by guest. Protected copyright. Participants undertook three custom and one standard binocular VF tests assessing visual 41 42 43 field to 60 degrees, and also integrated monocular threshold 24 2 visual fields (IVF). Primary 44 45 VF outcomes were average mean threshold, percentage of stimuli seen, and VF area. VF 46 47 outcomes were compared to self reported mobility function assessed with the Independent 48 49 Mobility Questionnaire, and time taken and patient acceptability were also considered. 50 51 Receiver Operating Characteristic (ROC) curves determined which tests best predicted 52 53 difficulty with mobility tasks. 54 55 56 Results 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 3 of 34 BMJ Open
1 2 3 Greater VF loss was associated with greater self reported mobility difficulty with all field 4 5 paradigms (R2 0.38 0.48, all p<0.001). All four binocular tests were better than the IVF at 6 7 predicting difficulty with at least three mobility tasks in ROC analysis. Mean duration of the 8 9 tests ranged from 1min 26sec (±9sec) for kinetic assessment to 9min 23sec (±24 sec) for 10 11 IVF. 12 13 14 Conclusions 15 For peer review only 16 17 The binocular VF tests extending to 60 deg eccentricity all relate similarly to self reported 18 19 mobility function, and slightly better than integrated monocular VFs. A kinetic assessment of 20 21 VF area is quicker than and as effective at predicting mobility function as static threshold 22 23 24 assessment. 25 26 27 Strengths and limitations of this study 28 29 30 Strengths of this study include: 31
32 http://bmjopen.bmj.com/ 33 o Rigorous and novel visual field testing on a cohort of people with visual field 34 35 loss 36 37 o Consideration of the patient’s perspective of visual field assessment 38 39 o Use of Rasch analysed data from a validated questionnaire to reflect a broad
40 on September 26, 2021 by guest. Protected copyright. 41 range of mobility function 42 43 o Different visual field paradigms compared to functional vision, which has not 44 45 previously been reported 46 47 48 Potential limitations include: 49 50 51 o 52 Single, rather than repeated, measures of visual field function are used 53 54 55 Copyright 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 4 of 34
1 2 3 The Corresponding Author has the right to grant on behalf of all authors and does grant on 4 5 behalf of all authors, an exclusive licence (or non exclusive for government employees) on a 6 7 worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be 8 9 published in BMJ editions and any other BMJPGL products and sublicences such use and 10 11 exploit all subsidiary rights, as set out in our licence. 12 13 14 Funding statement 15 For peer review only 16 17 This work was supported by a College of Optometrists’ Postgraduate Research Scholarship. 18 19 20 21 Competing interests 22 23 24 The authors report no conflicts of interest and have no proprietary interest in any of the 25 26 materials mentioned in this article. 27 28 29 Results included in this manuscript were presented in part at Vision 2017, The Hague, The 30 31 Netherlands (June 2017).
32 http://bmjopen.bmj.com/ 33 34 Author contribution 35 36 37 All authors (HS, KL, JM and MC) made substantial contributions to the design of the work. 38 39 HS acquired the data for the study. All authors contributed to the analysis and interpretation
40 on September 26, 2021 by guest. Protected copyright. 41 of the data, drafting the manuscript, and critical revision of the article. All authors give final 42 43 approval for the manuscript to be published and agree to be accountable for all aspects of 44 45 the work. 46 47 48 49 Transparency declaration 50 51 52 I affirm that the manuscript is an honest, accurate, and transparent account of the study 53 54 being reported; that no important aspects of the study have been omitted; and that any 55 56 discrepancies from the study as planned (and, if relevant, registered) have been explained. 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 5 of 34 BMJ Open
1 2 3 Data sharing statement 4 5 6 Data are available from FigShare, doi to be confirmed. 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 6 of 34
1 2 3 Introduction 4 5 6 1 Visual field loss is an extremely important dimension of visual impairment. There are 7 8 2 significant consequences of visual field loss for functional ability1 6, including increased 9 10 3 difficulty with mobility7 12, and potentially also increased risk of falling11,13 18 and of car 11 12 4 accidents19. However, currently available conventional visual fields tests are designed to 13 14 5 detect and monitor the progression of disease, and are not optimised for reflecting the 15 For peer review only 16 6 functional consequences of visual field loss. Although binocular visual fields are known to be 17 18 7 important in individuals with visual impairment, particularly visual fields beyond 30 degrees20, 19 20 8 conventional visual field tests only measure monocular visual fields of the central 30 degrees 21 22 21 24 9 or less . 23 24
25 25 26 10 Visual field assessments are demanding procedures and many people dislike performing
27 25,26 28 11 visual field tests . An ideal functional field test would take into account the acceptability of
29 27 30 12 tests to a patient, on the basis that an acceptable test and duration will provide better data . 31
32 13 Patients’ opinion of vision testing is largely unreported, although it has been suggested that http://bmjopen.bmj.com/ 33 34 14 this is due to difficulty objectively quantifying subjective, or “human factors”, of field 35 25,28 36 15 assessment . 37 38 39 16 One aspect of an ideal binocular functional visual field test that is not clear is what test
40 on September 26, 2021 by guest. Protected copyright. 41 17 strategies are most appropriate to use. Of the studies that have previously assessed the 42 43 18 visual field binocularly, strategies including kinetic29 33, suprathreshold strategies including 44 45 19 the Esterman visual field test10,33 38, and threshold tests24,39,40, have been used. 46 47 48 20 The closest current visual field tests to a ‘gold standard’ for assessing functional loss are the 49 50 21 binocular Esterman test41, and the integrated visual field (IVF)42 although this is a monocular 51 52 22 assessment used to produce binocular results. The Esterman test, the only standard 53 54 23 binocular field test available, is used to determine the extent of visual fields in UK drivers43, 55 56 24 although it was not originally designed for this purpose40. Although some studies suggest the 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 7 of 34 BMJ Open
1 2 3 25 Esterman test is a good predictor of visual function36,44,45, its limitations are well documented 4 5 26 and include a substantial ceiling effect which limits the test’s applicability for people with low 6 7 27 to moderate levels of field loss2,10,29,37,38,44,46,47, and an inability to differentiate effectively 8 9 28 between people with different levels of field loss38. As an alternative, Crabb and 10 11 29 Viswanathan42 constructed integrated visual fields from monocular field plots using software 12 13 30 that estimates sensitivities at points of overlapping locations using a “best location” 14 15 31 algorithm21,38For. While integrated peer visual fields review provide a rapid estimate only of a patient’s binocular 16 17 32 field without extra perimetric examination21,23,48, they do assume that previous visual field 18 19 33 results are available which is not always the case in low vision clinics. Binocular function is 20 21 34 derived rather than measured, and the plots used usually only assess the central 30 deg of 22 23 35 visual field. 24 25 26 The purpose of this study is to develop an optimal binocular visual field test to reflect the 27 36 28 functional consequences of visual field loss by assessing threshold, suprathreshold and 29 37 30 31 38 kinetic binocular visual field paradigms examining the visual field to 60 deg eccentricity, and
32 http://bmjopen.bmj.com/ 33 39 existing standard Esterman and IVF fields. Results are compared to self reported difficulty 34 35 40 with mobility to determine which visual field assessment paradigms best reflect functional 36 37 41 difficulty. The visual field testing experience of individuals with visual impairment is also 38 39 42 considered to help devise optimal strategies for functional field assessment.
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 Methods 43 44 45 44 Participants with self reported peripheral visual field loss were recruited for this cross 46 47 45 sectional observational study through advertisement of the study with local voluntary agency 48 49 46 Cam Sight, and the charities Retinitis Pigmentosa Fighting Blindness and the International 50 51 47 Glaucoma Association. Recruiting participants with a range of ocular conditions allowed the 52 53 48 assessment of a range of visual field loss, and reflects the need of a functional field 54 55 49 assessment to be applicable to patients with a range of visual disorders. Individuals with 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 8 of 34
1 2 3 50 conditions not primarily affecting peripheral visual function, such as macular degeneration, 4 5 51 were excluded from the study, along with those under 18 years old and those unable to 6 7 52 perform verbal evaluations in English. Ethical approval was granted by Anglia Ruskin 8 9 53 University Faculty of Science and Technology Research Ethics committee. All relevant 10 11 54 tenets of the Declaration of Helsinki were upheld. All participants gave informed consent 12 13 55 after the nature of the study was explained. Participants attended a single session at the 14 15 56 University EyeFor Clinic, Angliapeer Ruskin University, review with data collected only between 3/7/2015 and 16 17 57 4/3/2016 by a single researcher (HS), a qualified optometrist. 18 19 20 58 Part 1 of the Independent Mobility Questionnaire (IMQ)29,49 was used to assess self 21 22 59 perceived ability in mobility. Participants were asked to report if they encountered difficulty in 23 24 60 each of 35 mobility situations (Table 1) to obtain a binary response (Yes/No), and then to 25 26 rate the level the difficulty they experienced on a 5 point Likert scale. Respondents could 27 61 28 indicate that a mobility task was not applicable to them, which was scored as missing data. 29 62 30 31
32 How difficult is it for you to undertake the following tasks independently/without http://bmjopen.bmj.com/ 33 assistance (but with the aid of any mobility aids as required): 19. Adjusting to lighting changed during the 34 1. Walking in familiar areas 35 day: indoor to outdoor 20. Adjusting to lighting changed during the 36 2. Walking in unfamiliar areas 37 day: outdoor to indoor 38 21. Adjusting to lighting changed at night: 3. Moving about: at home 39 indoor to streetlights
40 22. Adjusting to lighting changed at night: on September 26, 2021 by guest. Protected copyright. 4. Moving about: at work 41 streetlights to indoor 42 43 5. Moving about: in the classroom 23. Walking in dimly lit indoor areas 44 24. Being aware of another person’s 6. Moving about: in stores 45 presence 46 47 7. Moving about: outdoors 25. Avoiding bumping into: people 48 49 8. Moving about: in crowded situations 26. Avoiding bumping into: walls 50 27. Avoiding bumping into: head height 51 9. Walking at night objects 52 28. Avoiding bumping into: shoulder height 53 10. Using public transportation 54 objects 29. Avoiding bumping into: waist height 55 11. Detecting ascending stairwells 56 objects 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 9 of 34 BMJ Open
1 2 3 30. Avoiding bumping into: knee height 12. Detecting descending stairwells 4 objects 5 6 13. Walking up steps 31. Avoiding bumping into: low lying objects 7 32. Avoiding tripping over uneven travel 14. Walking down steps 8 surfaces 9 10 15. Stepping onto curbs 33. Moving around in social gatherings 11 12 16. Stepping off curbs 34. Finding restrooms in public places 13 14 17. Walking through doorways 35. Seeing cars at intersections 15 For peer review only 16 18. Walking in high glare areas 17 18 63 19 20 21 64 Table 1. 35 mobility situations from Part 1 of the Independent Mobility Questionnaire 29,49 22 65 (IMQ) . 23 24 66 Participants were also asked to report whether they had fallen in the last 12 months, defined 25 26 67 as an event which results in a person coming to rest inadvertently on the ground, floor or 27 28 68 other lower level50. 29 30 31 69 High contrast distance visual acuity with habitual distance spectacle correction was
32 http://bmjopen.bmj.com/ 33 70 assessed binocularly and scored by letter51 using a 3m internally illuminated EDTRS chart52. 34 35 71 If the largest letters could not be read at 3m, the chart was moved 50% closer to the 36 37 72 participant to 1.5m and 0.75m. Participants with acuity that was not measureable were 38 39 73 assigned an acuity of 3.00logMAR53. Contrast sensitivity was measured binocularly with
40 on September 26, 2021 by guest. Protected copyright. 41 74 habitual distance spectacle correction using a Pelli Robson Chart54 at 1m scored on a by 42 43 75 letter basis55. Participants with no measurable CS function were assigned a score of 44 45 53 76 0.00logCS . 46 47 48 56 49 77 Five visual field assessments were performed using the Octopus 900 Perimeter and the
50 57 51 78 Humphrey Field Analyser . Visual field test order was randomised, and regular breaks were 52 53 79 provided. In all assessments, participants responded to seeing a visual field stimulus by 54 55 80 pressing a response button. The duration of all assessments was noted. The standard size 56 57 81 III Goldmann white stimulus was used throughout. Participants fixated on the standard 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 10 of 34
1 2 3 82 orange central point target (HFA) or green cross target (Octopus 900). When participants 4 5 83 had difficulty seeing the target or maintaining binocular fixation, a 6 deg ring target was 6 7 84 utilised on the Octopus, and an adaption to the fixation target on the HFA that slotted into the 8 9 85 fixation target hole to provide a black 2mm high contrast pericentral ring around the fixation 10 11 86 spot were used. 12 13 14 87 The three tests conducted on the Octopus 900 Perimeter were as follows, using the 15 For peer review only 16 88 ‘binocular’ setting in all cases: 17 18 19 89 1. Binocular threshold 20 21 A custom test point pattern was used, assessing the binocular field to 60 deg from 22 90 23 24 91 fixation with 52 points spaced every 7.5 deg in the central 30 deg, and 36 points 25 26 92 spaced every 15 deg in the peripheral 30 60 deg. Full aperture trial lenses were used 27 28 93 in adult half eye trial frames with lens centration distances corrected for near for 29 30 94 assessment of the central 30 deg. The peripheral 30 60 deg field was assessed 31
32 95 without correction to minimise the possibility of lens and frame artefacts. The ‘low http://bmjopen.bmj.com/ 33 34 96 vision’ test strategy was utilised: stimuli are presented using a 4 2 1 dB bracketing 35 36 97 test method starting at 0dB (4000asb) in order to arrive quickly at the expected 37 38 98 threshold level in subjects with impaired visual fields. The absolute thresholds 39 99 achieved at each test location were used to calculate the mean threshold39,42. 40 on September 26, 2021 by guest. Protected copyright. 41 42 100 43 44 101 2. Binocular suprathreshold 45 46 102 The same custom test pattern as for binocular threshold assessment was utilized. 47 48 103 Each point was assessed with a stimulus of 10dB intensity. The number of points 49 50 104 seen from the total of 88 was used to calculate a percentage score. 51 52 105 53 54 106 3. Binocular kinetic 55 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 11 of 34 BMJ Open
1 2 3 107 A white III 4e (10dB) target was presented kinetically at an angular velocity of 5 4 5 108 deg/sec. Vectors were presented from 60 deg eccentricity and moved centrally in 12 6 7 109 meridians spaced every 30 degrees. This assessment was completed without 8 9 110 refractive correction. The solid angle (deg2) subtended by the isopter was determined 10 11 111 automatically using the Eye Suite software55. The perimeter corrected results for 12 13 112 reaction time bias. 14 15 For peer review only 16 113 The two tests conducted on the HFA were as follows: 17 18 19 114 4. Integrated monocular threshold fields (IVF) 20 21 The central 24 2 threshold test with SITA Fast strategy was used to assess 22 115 23 24 116 monocular visual fields. Full aperture trial lenses provided refractive correction. 25 26 117 Integrated visual field scores were calculated using the best location
27 21,23,24,37,38,42,47,58 60 28 118 algorithm . 29 30 119 31
32 120 5. Esterman http://bmjopen.bmj.com/ 33 34 121 The binocular Esterman visual field test examines 120 test points out to 80 deg 35 36 122 eccentricity using a suprathreshold paradigm with a stimulus intensity of 10dB. This 37 38 123 assessment was completed without refractive correction. The number of points seen 39 124 was used to calculate the percentage Esterman Efficiency score. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 125 Participants were asked to rank the field tests in order from 1 to 5 based on how acceptable 44 45 126 they felt the tests were, and also to rank the perceived usefulness of results presented. 46 47 127 Participants were also invited to make further comments relating to acceptability and output 48 49 128 of the tests, and qualitative themes were evaluated from this data. 50 51 52 129 Analysis 53 54 55 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 12 of 34
1 2 3 130 Interval data was derived from the ordinal IMQ responses by the use of Rasch analysis61. 4 5 131 The Rasch model, a probabilistic logistic model, is a paradigm for the analysis and scoring of 6 7 132 questionnaires that can be used to produce interval level data from ordinal responses, and in 8 9 133 turn improves sensitivity to change in function62 and correlations with other variables62,63, 10 11 134 allows the use of robust parametric statistics on the data63, and provides more accurate 12 13 135 measurements of perceived function64. In the Rasch model, items and respondents are 14 15 136 scaled accordingFor to responses peer to a groupreview of items61,65. The only underlying construct being 16 17 137 assessed is used to define the relative difficulty of each item. On the same linear scale of the 18 19 138 construct, respondents are ordered from least to most ability, and items are ordered from 20 21 139 most to least difficult. 22 23 24 140 Rasch analysis derives person and item measures in logits from raw ordinal data. Person 25 26 measures are an estimate of a person’s underlying ability based on their performance on a 27 141 28 set of items that measure a single trait. The item measure is the Rasch estimate of item 29 142 30 31 143 difficulty.
32 http://bmjopen.bmj.com/ 33 34 144 Person measures were derived using all 35 items, which have been found to constitute a 35 29 49 36 145 unidimensional scale in people with peripheral field loss due to RP and glaucoma . Higher 37 38 146 person measures indicate greater perceived ability. 39
40 on September 26, 2021 by guest. Protected copyright. 41 147 The reliability indices were assessed in terms of person separation statistics, which provide 42 43 148 an indication of the instrument’s ability to discriminate between respondents: person 44 45 149 separation and person reliability should be greater than the suggested minima of 2.0 and 46 47 150 0.80 respectively66. Further, item separation statistics provide an indication of how reliably 48 49 151 ordered the items are in terms of difficulty: item separation and item reliability should be in 50 51 152 excess of suggested minima of 3.0 and 0.9066. Targeting, or the difference between mean 52 53 153 item and person measures, should ideally be less than 1.0 logit67,68. 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 13 of 34 BMJ Open
1 2 3 154 Additionally, the fit of individual items to a unidimensional construct is assessed. It is 4 5 155 considered that items with infit and outfit meansquare (mnsq) values within a range of 0.5 to 6 7 156 1.5 contribute usefully to a scale8. Items with mnsq values greater than two have the 8 9 157 potential to damage the integrity of the scale69. 10 11 12 158 Mann Whitney U tests compared visual function in participants who had reported a fall in the 13 14 159 previous 12 months and those who had not. Continuous clinical function variables were 15 For peer review only 16 160 compared to self reported mobility function in non parametric 2 tailed Spearman’s rho 17 18 161 bivariate correlations. 19 20 21 For the binary responses to the 35 mobility tasks of the IMQ, difficulty was compared to 22 162 23 24 163 different visual field test scores using receiver operating characteristic (ROC) analysis to 25 26 164 evaluate how effective the tests were at selecting participants with perceived mobility 27 28 165 difficulty (sensitivity), and without perceived mobility difficulty (specificity). Sensitivity and 29 30 166 specificity were determined for all possible cut off values for the visual field scores and 31
32 167 plotted as ROC curves. An area under the ROC curve of 1 indicates a perfect diagnostic http://bmjopen.bmj.com/ 33 34 168 procedure, whereas 0.5 indicates a poor procedure. A statistical technique appropriate 35 70 36 169 where two measures are applied to the same set of participants was used to compare 37 38 170 areas under the ROC curves and establish if any visual field test was statistically significantly 39 171 better at predicting perceived difficulty. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 172 Results 44 45 46 173 Table 2 shows the characteristics of the 50 participants who were recruited to the study: all 47 48 174 completed the study. Sample size is consistent with previous similar studies 1,29,34,42. Person 49 50 175 measures for the IMQ are shown. Other parameters derived from the Rasch analysis include 51 52 176 person separation of 3.43 (reliability 0.92), indicating that individuals can be reliably ordered 53 54 177 by the instrument in terms of their level of perceived ability. Item separation is 2.95 (reliability 55 56 178 0.90), slightly less than the minimum ideal value of 3, and indicating the instrument might not 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 14 of 34
1 2 3 179 be able to reliably order items in terms of their difficulty. Targeting ( 1.23±1.64 logits) is close 4 5 180 to the ideal of within ±1 logits of the mean item difficulty7. In terms of item fit, as a 6 7 181 representation of how well the questions fitted a unidimensional construct, there were three 8 9 182 mis fitting items with fits in the range 1.5 to 2.0 meansquare. These fits do not diminish the 10 11 183 validity of the measures and can be considered acceptable71, and so all items are 12 13 184 considered in the analysis. Item fits are also comparable to other Rasch analyses of this 14 15 185 instrument, For with three itemspeer with slight review mis fit also found foronly samples with RP29,72 and 16 17 186 glaucoma49. Reliability statistics indicate adequate visual field reliability for the majority of the 18 19 187 sample. Fixation losses for the monocular threshold tests suggest reliable results, with only 20 21 188 18% of participants losing fixation more than 20% during both assessments. 90% of false 22 23 189 positive and false negative statistics from binocular threshold data were better than 20%. 24 25 190 This figure is similar to statistics obtained from Esterman (86%) and binocular 26 27 191 suprathreshold results (78%) results. 28 29 30 31 192
32 http://bmjopen.bmj.com/ 33 34 Demographic variables 35 Gender (n) 29 Males, 21 Females 36 Age (years) 37 Median (25% IQ 75% IQ) 64(55 71) 38 Min max 24 84 39 Ocular diagnosis (n)
40 RP 14 on September 26, 2021 by guest. Protected copyright. 41 Glaucoma 23 42 Retinal detachments/tears 4 43 Other 9 44 Registration status (n) 45 Registered severely sight 18 46 impaired 47 Registered sight impaired 8 48 49 Not registered 24 50 Use of mobility aids (n) 51 White cane or guide dog 23 52 No mobility aids used 27 53 Have you been shown your visual field 54 results before? (n) 55 Yes 36 56 No 14 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 15 of 34 BMJ Open
1 2 3 Have you fallen in the previous 12mos? 4 Yes 28 5 No 22 6 Clinical function variables Mean (±std) Median (25% Range 7 IQ-75% IQ)) 8 Binocular distance visual acuity 0.28(±0.08) 0.09( 0.06 0.50) 0.28 3.00 9 (logMAR) 10 Binocular contrast sensitivity 1.51(±0.07) 1.65(1.30 1.95) 0.00 1.95 11 (logCS units) 12 Binocular visual field variables 13 Binocular threshold (dB) 10.87(±1.19) 10.14(2.13 0.27 25.60 14 19.40) 15 For peer review only 16 Binocular suprathreshold (%) 54.48(±5.09) 58.53(18.8 2.27 98.86 17 93.18) Binocular kinetic solid angle 5966.77(±541.19) 7355.7(1783.80 64.20 18 2 19 (deg ) 9566.70) 10320.50 20 Esterman (%) 59.43(±4.81) 67.08(33.33 0.00 21 90.83) 100.00 22 Integrated monocular threshold 15.69(±1.52) 15.17(4.88 0.90 31.96 23 (dB) 26.48) 24 Independent Mobility Questionnaire 1.23(±0.23) 1.26( 2.29 ( 5.92 1.84 25 Person Measures 0.09)) 26 27 193 Table 2. Descriptive statistics of the variables assessed (n=50). The mean ± standard 28 194 deviation, and the median (interquartile range) are given for the clinical visual function 29 195 variables. *Number of comorbid conditions from a list of 12 common medical conditions 30 196 representing general health status73. 31
32 197 http://bmjopen.bmj.com/ 33 34 35 Clinical function variables IMQ score (R2) Have you fallen in the 36 previous 12mos? (U) 37 38 Distance visual acuity (logMAR) 0.31, p≤0.001 288.0, p=0.701 39 Contrast sensitivity (logCS units) 0.33, p≤0.001 302.5, p=0.913 Binocular threshold (dB) 0.47, p≤0.001 236.0, p=0.157
40 on September 26, 2021 by guest. Protected copyright. 41 Binocular suprathreshold (%) 0.47, p≤0.001 235.0, p=0.161 2 42 Binocular kinetic solid angle (deg ) 0.48, p≤0.001 236.0, p=0.159 43 Esterman (%) 0.46, p≤0.001 209.0, p=0.053 44 IVF (dB) 0.38, p≤0.001 235.0, p=0.149 45 46 198 Table 3. Relationship between the variables assessed, and self reported mobility function 47 199 and falls history. Mann Whitney U values are provided for falls data and Spearman’s rho 48 200 bivariate correlations are provided for the IMQ score. 49 50 201 51 52 202 Table 3 shows the relationships between the parenfameters assessed and the outcome 53 54 203 measures of self reported mobility function, and fall history. Greater visual field loss is 55 56 significantly associated with greater self reported difficulty regardless of the method of field 57 204 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 16 of 34
1 2 3 205 assessment, although the relationship is a little weaker for the IVF score. 56% of the sample 4 5 206 reported falling at least once in the previous 12 months, but none of the clinical function 6 7 207 variables significantly associate with falls history (Table 3). 8 9 10 208 Figure 1. 11 12 13 209 The ability of the VF tests to discriminate between people with and without difficulty with 14 15 210 mobility tasksFor was reasonable peer (AUC>0.7) review for at least one VF onlytest for all except 2 of the 35 16 17 211 questions (Figure 1). Statistically significant differences in the areas under the ROC curves 18 19 212 were seen (indicated by * in Figure 1), in that binocular threshold and supra threshold 20 21 assessments were better than the IVF at predicting difficulty walking in familiar areas, 22 213 23 24 214 walking in unfamiliar areas, walking at home, walking in crowded areas, avoiding bumping 25 26 215 into knee height objects, and finding public toilets. The binocular threshold assessment was 27 28 216 also better than the IVF at predicting difficulty avoiding bumping into people, while the 29 30 217 Esterman was found to better predict difficulty walking in high glare when compared with the 31
32 218 IVF. All three custom tests and the Esterman assessment were better than the IVF at http://bmjopen.bmj.com/ 33 34 219 predicting difficulty avoiding bumping into waist height objects, and at predicting difficulty 35 36 220 avoiding bumping into low lying objects. 37 38 39 221 Figure 2.
40 on September 26, 2021 by guest. Protected copyright. 41 42 222 Figure 3. 43 44 45 223 The length of time taken to undertake each visual fields assessment is shown in Figure 2. 46 47 224 The quickest test was the kinetic which took a mean of 1min 26sec (±9sec), while the mean 48 49 225 duration of the longest test, the IVF, was 9min 23sec (±24 sec). 50 51 52 In terms of the participants’ view of how they found the tests to undertake, the most favoured 53 226 54 55 227 assessment was the kinetic (Figure 3), while the IVF was ranked the least favourite test by 56 57 228 over 60% of participants. In addition to ranking the tests, participants also made comments 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 17 of 34 BMJ Open
1 2 3 229 on the different test strategies, and common themes are considered here. Three participants 4 5 230 (6%) commented that shorter tests were preferred, the kinetic assessment in particular. 6 7 231 Participants found the kinetic assessment pleasant, less stressful, and encouraging, with 8 9 232 three participants commenting on being reassured by knowing that a light would be seen 10 11 233 eventually. Seven participants (14%) suggested that the kinetic assessment was more fun 12 13 234 and engaging than the static tests, and one remarked on the assessment’s novelty value. 14 15 235 However, sixFor participants peer (12%) expressed review concern that the testonly was too basic or too short, 16 17 236 and that the accuracy of results would be compromised by the test’s rapidity. Participants 18 19 237 were happy to conduct a longer test if they knew that results would be more beneficial to the 20 21 238 practitioner. Eleven participants (22%) indicated preference for binocular tests, which were 22 23 239 reported as more comfortable and less tiring than monocular assessments. Five participants 24 25 240 (10%) preferred the uniformly bright lights on the binocular suprathreshold assessment; 26 27 241 however, four (8%) found the bright lights at the start of the binocular threshold assessment 28 29 242 encouraging. 30 31
32 http://bmjopen.bmj.com/ 33 243 All participants in the study reported previous experience of visual field testing, although 34 35 244 28% had not seen their visual field results before (Table 2). All perceived value in being 36 37 245 shown their visual fields. In terms of visual field outputs, 48% of participants preferred the 38 39 246 grey scale plots from the binocular threshold analysis, and individuals remarked on the
40 on September 26, 2021 by guest. Protected copyright. 41 247 greater level of detail provided on a grey scale plot compared to others. The kinetic plot was 42 43 248 most favoured by 37% of participants, with preference expressed for plots that indicated 44 45 249 greater levels of residual field. Four participants commented on the usefulness of having 46 47 250 their peripheral field represented, with one participant suggesting that the monocular 48 49 251 threshold plots were an inaccurate depiction that overestimated the extent of their visual 50 51 252 field. The Esterman output was the least favoured by 38% of participants. Three participants 52 53 253 commented on the size of points on suprathreshold outputs, and expressed difficulty viewing 54 55 254 the results. 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 18 of 34
1 2 3 255 Discussion 4 5 6 256 Five different visual field assessments were compared to self reported difficulty with mobility, 7 8 257 in order to determine which visual field paradigms are suitable for use as a functional visual 9 10 258 field assessment. There are four principal findings of this study. Firstly, the paradigm used to 11 12 259 assess the visual field (threshold or suprathreshold static, or kinetic) makes little difference 13 14 260 to the relationship with function: so long as the test is performed binocularly and includes 15 For peer review only 16 261 assessment of eccentricities to 60 deg, the visual field outcome measure reflects self 17 18 262 reported mobility function well, in terms of both correlation between visual field score and 19 20 263 mobility function, and in terms of ability of visual field scores to discriminate between 21 22 264 individuals who do and do not have difficulty with specific mobility tasks. Secondly, tests that 23 24 265 are quicker to perform (binocular suprathreshold, binocular kinetic, Esterman; Figure 2) 25 26 relate just as well to self reported mobility function as tests that take longer (binocular 27 266 28 threshold; Table 2). Functional information is not lost by using kinetic or suprathreshold 29 267 30 31 268 techniques when compared to the diagnostic gold standard of measuring static thresholds.
32 http://bmjopen.bmj.com/ 33 269 Thirdly, quicker tests are also more acceptable to patients (Figure 3), especially the 34 35 270 binocular suprathreshold and kinetic paradigms. Finally, although all participants had 36 37 271 experience of visual field assessments, over a quarter did not recall being shown their visual 38 39 272 field results before. All participants in this study found value in seeing their visual field
40 on September 26, 2021 by guest. Protected copyright. 41 273 results, and clinicians are encouraged to explain visual field results with respect to likely 42 43 274 functional difficulties where possible. 44 45 46 275 A strength of this study is that although some previous studies45,74 76 have compared 47 48 276 different methods of quantifying the visual field and their relationship with functional vision, 49 50 277 we are not aware of any previous studies that have compared different paradigms of visual 51 52 278 field assessment with functional vision, nor that have taken the patient’s perspective into 53 54 279 consideration. Rigorous and novel visual field testing on a cohort of people with visual field 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 19 of 34 BMJ Open
1 2 3 280 loss has been used, along with Rasch analysed data from a validated questionnaire to 4 5 281 reflect a broad range of mobility function. 6 7 8 282 One possible limitation of the study is that we rely on self report of mobility difficulties rather 9 10 283 than directly measuring visual performance on, for example, a mobility circuit12,31. Another 11 12 284 potential limitation is that single responses have been used in the kinetic paradigm. 13 14 285 Responses to kinetic stimuli can be variable77, and repeated presentations might be 15 For peer review only 16 286 necessary to confirm responses, and reduce the impact of outlying responses78. 17 18 19 287 The findings are consistent with previous studies, in that we find that the IVF relates to 20 21 mobility function58,79. Contrary to other studies however38,42, the IVF appears to be less 22 288 23 24 289 effective at relating to self related mobility function when compared to the Esterman 25 26 290 assessment. This may be due to the difference in the degree of visual field loss between the 27 28 291 sample groups. The average Esterman scores in the current study (56.4%) are less than the
29 37 41 30 292 average scores (87.4% and 86.7% ) reported in other studies, and would suggest that the 31
32 293 present sample has a greater degree of visual field loss. Contrary to some previous studies, http://bmjopen.bmj.com/ 33 34 294 no significant correlations were found here between measures of clinical function and fall 35 6,11,13 36 295 history. Visual field loss has been shown to increase the risk of falling in some studies 37 18 80,81 38 296 , but not in others . Variation in findings reflects the multifactorial nature of falls, and also 39 297 suggests limitations in retrospective reporting of falls82. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 298 In conclusion, a binocular visual field test that does not ignore the peripheral 30 60 deg of 44 45 299 the field is effective for reflecting functional difficulty, regardless of the whether a threshold, 46 47 300 suprathreshold or kinetic assessment paradigm is employed. A shorter duration visual field 48 49 301 test such as the binocular kinetic assessment used in this study is favoured by patients, and 50 51 302 is similarly effective to more time consuming tests in predicting perceived disability in 52 53 303 patients with peripheral field loss. To further explore the potential of a binocular kinetic 54 55 304 assessment as a functional field test, future work needs to explore the repeatability of 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 20 of 34
1 2 3 305 responses, the necessity for repeated presentations, and the assessment’s ability to 4 5 306 discriminate between individuals with normal and abnormal fields using control subjects. 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 21 of 34 BMJ Open
1 2 3 References 4 5 6 307 1. Bibby SA, Maslin ER, McIlraith R, Soong GP. Vision and self reported mobility 7 8 308 performance in patients with low vision. Clin Exp Optom. 2007;90(2):115 23. 9 10 309 2. Nelson P, Aspinall P, Papasouliotis O, Worton BO, Brien C. Quality of life in 11 12 310 glaucoma and its relationship with visual function. J Glaucoma. 2003;12(2):139 150. 13 14 311 3. Tabrett DR, Latham K. Factors influencing self reported vision related activity 15 For peer review only 16 312 limitation in the visually impaired. Invest Ophthal Vis Sci. 2011;52:5293–5302. 17 18 313 4. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 19 20 314 field loss in the elderly and associations with impairment in daily functioning: the 21 22 315 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 23 24 316 5. Sherwood M, Garcia Siekavizza A, Meltzer M, et al. Glaucoma’s impact on quality of 25 26 life and its relation to clinical indicators. Ophthalmology. 1998;105:561–566. 27 317 28 6. Ivers RQ, Cumming RG, Mitchell P, et al. Visual impairment and falls in older adults: 29 318 30 31 319 the Blue Mountains Eye Study. J Am Geriatr Soc. 1998;46:58–64.
32 http://bmjopen.bmj.com/ 33 320 7. Latham K, Baranian M, Timmis MA, Pardhan S. Difficulties with Goals of the Dutch 34 35 321 ICF Activity Inventory: Perceptions of Those with Retinitis Pigmentosa and of Those 36 37 322 Who Support Them. Inves Ophthal Vis Sci. 2015;56(4):2381 2391. 38 39 323 8. Lowe J, Drasdo N. Patients' responses to retinitis pigmentosa. Optom Vis Sci.
40 on September 26, 2021 by guest. Protected copyright. 41 324 1992;69 (3):182 185. 42 43 325 9. Szlyk JP, Fishman GA, Grover S, Revelins BI, Derlacki DJ. Difficulty in performing 44 45 326 everyday activities in patients with juvenile macular dystrophies: comparison with 46 47 327 patients with. Br J Ophthalmol. 1998;82(12):1372 1376. 48 49 328 10. Noe G, Ferraro J, Lamoureux E, Rait J, Keeffe JE. Associations between 50 51 329 glaucomatous visual field loss and participation in activities of daily living. Clin Exp 52 53 330 Optom. 2003;31(6):482 6. 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 22 of 34
1 2 3 331 11. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 4 5 332 field loss in the elderly and associations with impairment in daily functioning: the 6 7 333 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 8 9 334 12. Lovie Kitchin JE, Soong GP, Hassan SE, Woods RL. Visual field size criteria for 10 11 335 mobility rehabilitation referral. Optom Vis Sci. 2010;87(12):948 57. 12 13 336 13. Patino CM, McKean Cowdin R, Azen SP, Allison JC, Choudhury F, Varma R. Central 14 15 337 and PeripheralFor Visualpeer Impairment review and the Risk of Falls onlyand Falls with Injury. 16 17 338 Ophthalmology. 2010;117(2):199. 18 19 339 14. Klein BEK, Klein R, Lee KE, Cruickshanks KJ. Performance based and self 20 21 340 assessed measures of visual function as related to history of falls, hip fractures, and 22 23 341 measured gait time: The Beaver Dam Study. Ophthalmology. 1998;105(1):160 164. 24 25 342 15. Jack CIA, Smith T, Neoh C, Lye M, Mcgalliard JN. Prevalence of low vision in elderly 26 27 343 patients admitted to an acute geriatric unit in Liverpool: Elderly people who fall are 28 29 344 more likely to have low vision. Gerontology. 1995;41(5):280 285. 30 31 345 16. Freeman EE, Munoz B, Rubin GS, West SK. Visual field loss increases the risk of 32 http://bmjopen.bmj.com/ 33 falls in older adults: the Salisbury eye evaluation. Inves Ophthal Vis Sci. 34 346 35 36 347 2007;48(10):4445 4450. 37 38 348 17. Klein BEK, Moss SE, Klein R, Lee KE, Cruickshanks KJ. Associations of visual 39 349 function with physical outcomes and limitations 5 years later in an older population:
40 on September 26, 2021 by guest. Protected copyright. 41 42 350 The Beaver Dam eye study. Ophthalmology. 2003;110(4):644 650. 43 44 351 18. Haymes SA, LeBlanc RP, Nicolela MT, Chiasson LA, Chauhan B.C. Risk of falls and 45 46 352 motor vehicle collisions in glaucoma. Invest. Ophthalmol. Vis. Sci. 2007;48:1149 47 48 353 1155. 49 50 354 19. McGwin G Jr, Xie A, Mays A, Joiner W, DeCarlo DK, Hall TA, Owsley C. Visual field 51 52 355 defects and the risk of motor vehicle collisions among patients with glaucoma. Invest 53 54 356 Ophthalmol Vis Sci. 2005;46(12):4437 41. 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 23 of 34 BMJ Open
1 2 3 357 20. Subhi H, Latham K, Myint J, Crossland M. Functional visual fields: relationship of 4 5 358 visual field areas to self reported function. Ophthalmic Physiol Opt. 2017;37:399– 6 7 359 408. 8 9 360 21. Nelson Quigg JM, Cello K, Johnson CA. Predicting binocular visual field sensitivity 10 11 361 from monocular visual field results. Inves Ophthal Vis Sci. 2000;41:2212–2221. 12 13 362 22. Schneck ME, Haegerstom Portnoy G, Lott L, Brabyn JA. Monocular vs. Binocular 14 15 363 MeasurementFor of Spatialpeer Vision inreview Elders. Optom Vis Sci. only 2010;87(8):526 531. 16 17 364 23. Asaoka R, Crabb D, Yamashita T, Russell R, Wany YX, Garway Heath DF. Patients 18 19 365 Have Two Eyes!: Binocular versus Better Eye Visual Field Indices. Inves Ophthal Vis 20 21 366 Sci. 2011;52(9):7007 7011. 22 23 367 24. Crabb DP, Smith ND, Glen FC, Burton R, Garway Heath DF. How does glaucoma 24 25 368 look? Patient perception of visual field loss. Ophthalmology. 2013;120(6):1120 6. 26 27 369 25. Gardiner SK, Demirel S. Assessment of patient opinions of different clinical tests 28 29 370 used in the management of glaucoma. Ophthalmology. 2008;115:2127–31. 30 31 371 26. Glen FC, Baker H, Crabb DP A qualitative investigation into patients’ views on visual 32 http://bmjopen.bmj.com/ 33 field testing for glaucoma monitoring BMJ Open 2014;4(1). 34 372 35 36 373 27. Hudson C, Wild JM, O’Neill EC. Fatigue effects during a single session of automated 37 38 374 static threshold perimetry. Invest Ophthalmol Vis Sci. 1994;35:268–80. 39 375 28. Artes, P, Monhart M, Marco M. Are big stimuli “easier” to see? An analysis of
40 on September 26, 2021 by guest. Protected copyright. 41 42 376 response times from frequency of seeing data. In 22nd International Visual Field 43 44 377 Imaging Symposium. Udine, Italy, September 27th 30th 2016. 45 46 378 29. Turano KA, Rubin GS, Quigley HA Mobility performance in glaucoma. Invest 47 48 379 Ophthalmol Vis Sci. 1999;40:2803 2309. 49 50 380 30. Lovie Kitchin JE, Mainstone J, Robinson J, Brown B. What areas of the visual field 51 52 381 are important for mobility in low vision patients? Clin Vis Sci. 1990;5:249 263. 53 54 382 31. Haymes SA, Johnston AW, Heyes AD. Relationship between vision impairment and 55 56 383 ability to perform activities of daily living. Ophthalmic Physiol Opt. 2002;22(2):79 91. 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 24 of 34
1 2 3 384 32. Haymes SA, Leblanc RP, Nicolela MT et al. Risk of falls and motor vehicle collisions 4 5 385 in glaucoma. Invest Ophthalmol Vis Sci. 2007;48:1149–55. 6 7 386 33. Hassan SE, Hicks JC, Lei H, Turano KA. What is the minimum field of view required 8 9 387 for efficient navigation? Vision Res. 2007;47(16):2115 2123. 10 11 388 34. Lee J, Cho H, Kee C. Assessment of the Vision Specific Quality of Life Using 12 13 389 Binocular Esterman Visual Field in Glaucoma Patients. J Korean Ophthalmol Soc, 14 15 390 2013;54(10):1567.For peer review only 16 17 391 35. Mills RP, Drance SM. Esterman disability rating in severe glaucoma. Ophthamology. 18 19 392 1986;93:371 378. 20 21 393 36. Fujita K, Yasuda N, Nakamoto K, Fukuda T. The relationship between difficulty in 22 23 394 daily living and binocular visual field in patients with glaucoma. Nihon Ganka Gakkai 24 25 395 Zasshi. 2008;112(5):447 50. 26 27 396 37. Jampel HD. Glaucoma patients’ assessment of their visual function and quality of life. 28 29 397 Tr Am Ophth Soc. 2001;99:301 317. 30 31 398 38. Jampel HD, Friedman DS, Quigley H, et al. Correlation of the binocular visual field 32 http://bmjopen.bmj.com/ 33 with patients’ assessment of vision. Inves Ophthal Vis Sci. 2002;43:1059 67. 34 399 35 36 400 39. Tabrett DR, Latham K. Important areas of the central binocular visual field for daily 37 38 401 functioning in the visually impaired. Ophthalmic Physiol Opt. 2012;32:156 163. 39 402 40. Black A, Lovie Kitchin J, Woods R, Arnold N, Byrnes J, Murrish J. Mobility
40 on September 26, 2021 by guest. Protected copyright. 41 42 403 performance with retinitis pigmentosa. Clin Exp Optom. 1996;80(1):1 12. 43 44 404 41. Esterman B. Functional scoring of the binocular field. Ophthalmology. 1982;89:1226– 45 46 405 34. 47 48 406 42. Crabb DP, Viswanathan AC. Integrated visual fields: a new approach to measuring 49 50 407 the binocular field of view and visual disability. Graefes Arch Clin Exp Ophthalmol. 51 52 408 2004;243(3):210 6. 53 54 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 25 of 34 BMJ Open
1 2 3 409 43. Driver and Vehicle Licensing Agency. Visual disorders: assessing fitness to drive. 4 5 410 2016 Available: https://www.gov.uk/guidance/visual disorders assessing fitness to 6 7 411 drive. Last accessed February 2017. 8 9 412 44. Parrish RK II, Gedde SJ, Scott IU, et al. Visual function and quality of life among 10 11 413 patients with glaucoma. Arch Ophthalmol. 1997;115:1447 1455. 12 13 414 45. Choy ES, Mills RP, Drance SM. Automated Esterman testing of disability in 14 15 415 glaucoma.For In:Greve peer E, Heiji A, eds. review Seventh International only Visual Field Symposium. 16 17 416 Amsterdam, the Netherlands: Junk Publishers. 1986:527 535. 18 19 417 46. Mills RP. Correlation of quality of life with clinical symptoms and signs at the time of 20 21 418 glaucoma diagnosis. Trans Am Ophthalmol Soc. 1998;96:753 812. 22 23 419 47. Harris ML, Jacobs, NA. Is the Esterman binocular field sensitive enough? In: Mills 24 25 420 RP, Wall M. Perimetry update 1994/1995. Amsterdam: Kugler, 1995:13–24. 26 27 421 48. Crabb DP, Viswanathan AC, McNaught AI, et al. Simulating binocular visual field 28 29 422 status in glaucoma. Br J Ophthalmol. 1998;82:1236–41. 30 31 423 49. Turano KA, Geruschat DR, Baker FH. Fixation behavior while walking: persons with 32 http://bmjopen.bmj.com/ 33 central visual field loss. Vision Res. 2002;42:2635–2644. 34 424 35 36 425 50. Skelton DA, Becker C, Lamb SE, Close JCT, Zijlstra W, Yardley L, Todd CJ. 37 38 426 Prevention of Falls Network Europe: a Thematic Network aimed at introducing good 39 427 practice in effective falls prevention across Europe. Eur J Aging. 2004;1:89 94.
40 on September 26, 2021 by guest. Protected copyright. 41 42 428 51. Arditi A, Cagenello B. On the statistical reliability of letter chart visual acuity 43 44 429 measurement. Invest Ophthalmol Vis Sci. 1993;34:120–129. 45 46 430 52. Ferris FL, Kassoff A, Bresnick GH, Bailey IL. New visual acuity charts for clinical 47 48 431 research. Am J of Ophthalmol. 1982;94:91–96. 49 50 432 53. Myint J, Latham K, Mann D, Gomersall P, Wilkins A, Allen P. The relationship 51 52 433 between visual function and performance in rifle shooting for athletes with vision 53 54 434 impairment. BMJ Open Sport Exerc Med. 2016;2(1). 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 26 of 34
1 2 3 435 54. Pelli DG, Robson JG, Wilkins AJ. The design of a new letter chart for measuring 4 5 436 contrast sensitivity. Clin Vis Sci. 1988;2:187 199. 6 7 437 55. Elliott DB, Bullimore MA, Bailey IL. Improving the reliability of the Pelli Robson 8 9 438 contrast sensitivity test. Clin Vis Sci. 1991;6(6):471 475. 10 11 439 56. Octopus, Haag Streit International, AG, Bern, Switzerland 12 13 440 57. HFA, Carl Zeiss Meditec, Inc., Dublin, CA. 14 15 441 58. AspinallFor PA, Johnson peer ZK, Azuara Blanco review A, Montarzino only A, Brice R, Vickers A. 16 17 442 Evaluation of quality of life and priorities of patients with glaucoma. Invest 18 19 443 Ophthalmol Vis Sci. 2008;49(5):1907 1915. 20 21 444 59. Saunders LJ, Russell RA, Crabb DP. Practical landmarks for visual field disability in 22 23 445 glaucoma. Br J Ophthalmol. 2012;96:1185–1189. 24 25 446 60. Chisholm CM, Rauscher FG, Crabb DC, et al. Assessing Visual Fields for Driving in 26 27 447 Patients With Paracentral Scotomata. Br J Ophthalmol. 2008;92(2):225 230. 28 29 448 61. Rasch G. Probabalistic models for some intelligence and attainment tests. Chicago: 30 31 449 Mesa Press. 1993. 32 http://bmjopen.bmj.com/ 33 62. Norquist, J., Fitzpatrick, R., Dawson, J. and Jenkinson, C. (2004). Comparing 34 450 35 36 451 Alternative Rasch Based Methods vs Raw Scores in Measuring Change in Health. 37 38 452 Medical Care, 42, p.25. 39 453 63. Khadka J, McAlinden C, Pesudovs K. Quality Assessment of Ophthalmic
40 on September 26, 2021 by guest. Protected copyright. 41 42 454 Questionnaires: Review and Recommendations. Optom Vis Sci. 2013;8:720–744. 43 44 455 64. Stelmack J, Massof R. (2007). Using the VA LV VFQ 48 and LV VFQ 20 in Low 45 46 456 Vision Rehabilitation. Optom Vis Sci, 84(8), pp.705 709. 47 48 457 65. Reise S, Henson J. (2003). A Discussion of Modern Versus Traditional 49 50 458 Psychometrics As Applied to Personality Assessment Scales. J Pers Assess, 81(2), 51 52 459 pp.93 103. 53 54 460 66. Linacre JM. Reliability and separation of measures. Retrieved 30/06/2015, from 55 56 461 http://www.winsteps.com/winman/reliability.htm. 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 27 of 34 BMJ Open
1 2 3 462 67. Gothwal VK, Wright TA, Lamoureux EL, and Pesudovs K. Rasch Analysis of Visual 4 5 463 Function and Quality of Life Questionnaires. Optom Vis Sci. 2009;86(10):1160 1168. 6 7 464 68. Marella M, Gothwal VK, Pesudovs K, Lamoureux EL. Validation of the Visual 8 9 465 Disability Questionnaire (VDQ) in India. Optom Vis Sci. 2009;86:826 835. 10 11 466 69. Wright B, Linacre JM. Reasonable mean square fit values. Retrieved 29/02/2016, 12 13 467 from www.rasch.org/rmt/rmt83b.htm. 14 15 468 70. DeLongFor ER, DeLong peer DM, Clarke Pearson review DL. Comparing only the areas under two or 16 17 469 more correlated receiver operating characteristic curves: a nonparametric approach. 18 19 470 Biometrics. 1988;44(3):837 45. 20 21 471 71. Linacre JM. Dimensionality investigation an example. Retrieved 12/11/2014, from 22 23 472 http://www.winsteps.com/winman/multidimensionality.htm. 24 25 473 72. Fenwick EK, O'Hare F, Deverell L, Ayton LN, Luu CD, McSweeney S, Bentley SA, 26 27 474 Guymer RH, Finger RP. Rasch Analysis of the Independent Mobility Questionnaire. 28 29 475 Optom Vis Sci. 2016;93(2):181 187. 30 31 476 73. van Nispen RMA, Hoeijmakers JGJ, de Boer MR, et al. Agreement Between Self 32 http://bmjopen.bmj.com/ 33 Reported Co morbidity of Visually Impaired Older Patients and Reports from their 34 477 35 36 478 General Practitioners. Visual Impair Res. 2008;10(2 3):49 56. 37 38 479 74. Haymes S, Guest D, Heyes A, Johnston A. Mobility of people with retinitis 39 480 pigmentosa as a function of vision and psychological variables. Optom Vis Sci.
40 on September 26, 2021 by guest. Protected copyright. 41 42 481 1996;73(10):621 637. 43 44 482 75. Safran AB, Glaser JS. Statokinetic Dissociation in Lesions of the Anterior Visual 45 46 483 Pathways: A Reappraisal of the Riddoch Phenomenon. Arch Ophthalmol. 47 48 484 1980;98:291 295. 49 50 485 76. Yanagisawa M, Kato, S, Kunimatsu S, Tamura M, Ochiai M. Relationship between 51 52 486 vision related quality of life in Japanese patients and methods for evaluating visual 53 54 487 field. Jpn J Ophthalmol, 2011;55(2):132 137. 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 28 of 34
1 2 3 488 77. Lynn J, Swanson W, Fellman R. Evaluation of automated kinetic perimetry (AKP) 4 5 489 with the Humphrey Field Analyzer. Perimetry Update. 1990;1991:433 452. 6 7 490 78. Moenter VM, Crabb DP, Artes PH. Reclaiming the Periphery: Automated Kinetic 8 9 491 Perimetry for Measuring Peripheral Visual Fields in Patients with Glaucoma. Invest 10 11 492 Ophthal Vis Sci. 2017;58(2):868 875. 12 13 493 79. Black AA, Wood JM, Lovie Kitchin JE. Inferior field loss increases rate of falls in older 14 15 494 adultsFor with glaucoma. peer Optom Vis review Sci. 2011;88(11):1275 82. only 16 17 495 80. Friedman SM, Munoz B, West SK, et al. Falls and fear of falling: which comes first? A 18 19 496 longitudinal prediction model suggests strategies for primary and secondary 20 21 497 prevention. J Am Geriatr Soc. 2002;50:1329 35. 22 23 498 81. Glynn RJ, Seddon JM, Krug JH, Jr, et al. Falls in elderly patients with glaucoma. Arch 24 25 499 Ophthalmol. 1991;109:205–10. 26 27 500 82. Cummings SR, Nevitt MC, Kidd S. Forgetting falls. The limited accuracy of recall of 28 29 501 falls in the elderly. J Am Geriatr Soc. 1988;36:613–16. 30 31
32 http://bmjopen.bmj.com/ 502 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 28 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 29 of 34 BMJ Open
1 2 3 503 4 5 504 Figure 1. Graphical representation of areas under ROC curves for the different visual field 6 505 tests. The questions to which the numbers on the x axis refer are given in Table 1. * 7 506 indicates a visual field assessment with a statistically significant greater area under the curve 8 507 (AUC) than an alternative assessment (p≤0.05). All other comparisons were non significant. 9 10 508 Figure 2. Test durations of each of the five visual field assessments. 11 12 13 14 509 Figure 3. Participants’ ranking of visual field tests acceptability. 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 29 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 30 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 Graphical representation of areas under ROC curves for the different visual field tests. The questions to 27 which the numbers on the x axis refer are given in Table 1. * indicates a visual field assessment with a 28 statistically significant greater area under the curve (AUC) than an alternative assessment (p≤0.05). All 29 other comparisons were non significant. 30 31 542x305mm (300 x 300 DPI)
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 31 of 34 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 Test durations of each of the five visual field assessments. 36 37 165x137mm (300 x 300 DPI) 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 32 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 Participants’ ranking of visual field tests acceptability. 26 27 579x304mm (300 x 300 DPI) 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from 7-10 7-10 NA NA 1 1
11 11 10-11 10-11 11 11 3 3 6 6 6 6
6 6 5 5
10-11 10-11 7-10 7-10 10-11 10-11 Reported on page # Reported on page 2 2 6 6
cross-sectional studies cross-sectional
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
) Indicate ) study’sthe design commonlywith useda in term ortitlethe abstract the ) ) theGive eligibility criteria, and sourcesthe methodsand of selection of participants ) ) statisticalDescribe all methods, thoseincluding confoundingusedforcontrol to ) ) Describe any methods used subgroups examine to interactionsand ) If ) applicable, describe methodsanalytical taking account of sampling strategy ) ) in theProvide abstract informativean and balanced summary of what was donewhat wasand found ) Describe) sensitivity any analyses
For peer review only ) howExplain missing data were addressed a a a b e d c b For eachof interest,For variable sourcesdatagive of and details of methods of assessment (measurement). Describe comparability of of comparability assessment methods isthanthere more if one group ( ( why why
applicable collection collection ( ( ( ( Recommendation STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in be included should that items of Statement—Checklist (v4) 2007 STROBE
6 1 ( 3 specific objectives, State including prespecified any hypotheses 5 the Describe setting, locations, relevantand dates, including periods exposure,of recruitment, follow-up, data and 9 Describe any efforts to address sourcespotential of bias 7 define Clearly outcomes,all exposures,predictors, potential confounders,effect and modifiers. diagnostic Give criteria, if 4 key Present of elements study inthedesign early paper 8* 10 10 how Explain studythe size was arrived at # Item Item Results
Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 thebackground Explain scientific for rationale investigation the and being reported Introduction Title andTitle abstract Section/Topic
Page 33 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 34 of 17-19 17-19 12 12 NA
17-19 17-19 3 3 17-19 17-19
12 12 NA NA 11 11 NA NA 4 4 14 14 NA
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report) boundaries category when continuous variables were categorized ) ) unadjustedGive estimates if applicable, and, confounder-adjusted estimates their and precision95% (eg, confidence
For peer )relevant, If consider translating estimatesrelativerisk of into absolute for risk meaningfulperioda time review only c b a similar studies,similar and other relevant evidence confounders magnitude of magnitude potential any bias confirmed eligible,confirmed included in study, the completing follow-up, analysed and interval). Make interval). clear confounderswhich were adjusted they why and for were included which the presentthe which is based article
( ( number Indicate (b) of participants missingwith eachfor data of interest variable Consider (c) diagram use flow of a reasons Give (b) non-participationfor each at stage 18 18 Summarise results key to reference withstudy objectives 16 16 ( 22 22 sourcethe Give of funding and role the of fundersthe thefor studypresent if and, applicable, originalthe for study on 19 19 Discuss of study, the limitations taking into sourcesaccount biasof potential imprecision.or Discussboth direction and 13* 13* Report (a) numbers of individuals each at stage of study—eg numbers eligible, potentially examinedfor eligibility, An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/,at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (externalgeneralisabilitythe of validity) studythe results Interpretation 20 cautious Give interpretationa of overall resultsconsidering limitations, objectives, multiplicity of resultsanalyses, from Limitations Key Key results Discussion Other analyses 17 other Report analyses done—eg analyses of subgroups sensitivity interactions, and and analyses
Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures
Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and
Participants 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open
Functional visual fields: A cross-sectional UK study to determine which visual field paradigms best reflect difficulty with mobility function.
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2017-018831.R2
Article Type: Research
Date Submitted by the Author: 10-Oct-2017
Complete List of Authors: Subhi, Hikmat; Anglia Ruskin University, Vision and Hearing Sciences Latham, Keziah; Anglia Ruskin University, Vision and Hearing Sciences Myint, Joy; University of Hertfordshire School of Life and Medical Sciences Crossland, Michael; Moorfields Eye Hospital NHS Foundation Trust, Optometry
Primary Subject Ophthalmology Heading:
Secondary Subject Heading: Ophthalmology
Low vision, Visual fields, Perimetry, Mobility function, Visual impairment, http://bmjopen.bmj.com/ Keywords: Self-reported function
on September 26, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 1 of 34 BMJ Open
1 2 3 Title Functional visual fields: A cross sectional UK study to determine which 4 visual field paradigms best reflect difficulty with mobility function. 5 Running head Functional visual fields 6 Authors *Subhi, Hikmat; Anglia Ruskin University, Department of Vision and 7 Hearing Sciences and Vision and Eye Research Unit 8 9 Latham, Keziah; Anglia Ruskin University, Department of Vision and 10 Hearing Sciences and Vision and Eye Research Unit 11 12 Myint, Joy; University of Hertfordshire, Life and Medical Sciences, 13 Postgraduate Medicine 14 15 For peer review only 16 Crossland, Michael D.; Moorfields Eye Hospital, London 17 18 *Corresponding author: [email protected] 19 Word count 4813 20 Key words Visual fields, self reported function, mobility function, functional vision, 21 visual impairment 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 2 of 34
1 2 3 Abstract 4 5 6 Objectives 7 8 9 To develop an appropriate method of assessing visual field (VF) loss which reflects its 10 11 functional consequences, this study aims to determine which method(s) of assessing visual 12 13 fields best reflect mobility difficulty. 14 15 For peer review only 16 Setting 17 18 19 20 This cross sectional observational study took place within a single primary care setting. 21 22 Participants attended a single session at a University Eye Clinic, Cambridge, UK, with data 23 24 collected by a single researcher (HS), a qualified optometrist. 25 26 27 Participants 28 29 30 50 adult participants with peripheral field impairment were recruited for this study. Individuals 31
32 with conditions not primarily affecting peripheral visual function, such as macular http://bmjopen.bmj.com/ 33 34 degeneration, were excluded from the study. 35 36 37 Primary and secondary outcome measures 38 39
40 on September 26, 2021 by guest. Protected copyright. Participants undertook three custom and one standard binocular VF tests assessing visual 41 42 43 field to 60 degrees, and also integrated monocular threshold 24 2 visual fields (IVF). Primary 44 45 VF outcomes were average mean threshold, percentage of stimuli seen, and VF area. VF 46 47 outcomes were compared to self reported mobility function assessed with the Independent 48 49 Mobility Questionnaire, and time taken and patient acceptability were also considered. 50 51 Receiver Operating Characteristic (ROC) curves determined which tests best predicted 52 53 difficulty with mobility tasks. 54 55 56 Results 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 3 of 34 BMJ Open
1 2 3 Greater VF loss was associated with greater self reported mobility difficulty with all field 4 5 paradigms (R2 0.38 0.48, all p<0.001). All four binocular tests were better than the IVF at 6 7 predicting difficulty with at least three mobility tasks in ROC analysis. Mean duration of the 8 9 tests ranged from 1min 26sec (±9sec) for kinetic assessment to 9min 23sec (±24 sec) for 10 11 IVF. 12 13 14 Conclusions 15 For peer review only 16 17 The binocular VF tests extending to 60 deg eccentricity all relate similarly to self reported 18 19 mobility function, and slightly better than integrated monocular VFs. A kinetic assessment of 20 21 VF area is quicker than and as effective at predicting mobility function as static threshold 22 23 24 assessment. 25 26 27 Strengths and limitations of this study 28 29 30 Strengths of this study include: 31
32 http://bmjopen.bmj.com/ 33 o Rigorous and novel visual field testing on a cohort of people with visual field 34 35 loss 36 37 o Consideration of the patient’s perspective of visual field assessment 38 39 o Use of Rasch analysed data from a validated questionnaire to reflect a broad
40 on September 26, 2021 by guest. Protected copyright. 41 range of mobility function 42 43 o Different visual field paradigms compared to functional vision, which has not 44 45 previously been reported 46 47 48 Potential limitations include: 49 50 51 o 52 Single, rather than repeated, measures of visual field function are used 53 54 55 Copyright 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 4 of 34
1 2 3 The Corresponding Author has the right to grant on behalf of all authors and does grant on 4 5 behalf of all authors, an exclusive licence (or non exclusive for government employees) on a 6 7 worldwide basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be 8 9 published in BMJ editions and any other BMJPGL products and sublicences such use and 10 11 exploit all subsidiary rights, as set out in our licence. 12 13 14 Funding statement 15 For peer review only 16 17 This work was supported by a College of Optometrists’ Postgraduate Research Scholarship. 18 19 20 21 Competing interests 22 23 24 The authors report no conflicts of interest and have no proprietary interest in any of the 25 26 materials mentioned in this article. 27 28 29 Results included in this manuscript were presented in part at Vision 2017, The Hague, The 30 31 Netherlands (June 2017).
32 http://bmjopen.bmj.com/ 33 34 Author contribution 35 36 37 All authors (HS, KL, JM and MC) made substantial contributions to the design of the work. 38 39 HS acquired the data for the study. All authors contributed to the analysis and interpretation
40 on September 26, 2021 by guest. Protected copyright. 41 of the data, drafting the manuscript, and critical revision of the article. All authors give final 42 43 approval for the manuscript to be published and agree to be accountable for all aspects of 44 45 the work. 46 47 48 49 Transparency declaration 50 51 52 I affirm that the manuscript is an honest, accurate, and transparent account of the study 53 54 being reported; that no important aspects of the study have been omitted; and that any 55 56 discrepancies from the study as planned (and, if relevant, registered) have been explained. 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 5 of 34 BMJ Open
1 2 3 Data sharing statement 4 5 6 Data are available from FigShare, doi: 10.6084/m9.figshare.5484109.v1 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 6 of 34
1 2 3 Introduction 4 5 6 1 Visual field loss is an extremely important dimension of visual impairment. There are 7 8 2 significant consequences of visual field loss for functional ability1 6, including increased 9 10 3 difficulty with mobility7 12, and potentially also increased risk of falling11,13 18 and of car 11 12 4 accidents19. However, currently available conventional visual fields tests are designed to 13 14 5 detect and monitor the progression of disease, and are not optimised for reflecting the 15 For peer review only 16 6 functional consequences of visual field loss. Although binocular visual fields are known to be 17 18 7 important in individuals with visual impairment, particularly visual fields beyond 30 degrees20, 19 20 8 conventional visual field tests only measure monocular visual fields of the central 30 degrees 21 22 21 24 9 or less . 23 24
25 25 26 10 Visual field assessments are demanding procedures and many people dislike performing
27 25,26 28 11 visual field tests . An ideal functional field test would take into account the acceptability of
29 27 30 12 tests to a patient, on the basis that an acceptable test and duration will provide better data . 31
32 13 Patients’ opinion of vision testing is largely unreported, although it has been suggested that http://bmjopen.bmj.com/ 33 34 14 this is due to difficulty objectively quantifying subjective, or “human factors”, of field 35 25,28 36 15 assessment . 37 38 39 16 One aspect of an ideal binocular functional visual field test that is not clear is what test
40 on September 26, 2021 by guest. Protected copyright. 41 17 strategies are most appropriate to use. Of the studies that have previously assessed the 42 43 18 visual field binocularly, strategies including kinetic29 33, suprathreshold strategies including 44 45 19 the Esterman visual field test10,33 38, and threshold tests24,39,40, have been used. 46 47 48 20 The closest current visual field tests to a ‘gold standard’ for assessing functional loss are the 49 50 21 binocular Esterman test41, and the integrated visual field (IVF)42 although this is a monocular 51 52 22 assessment used to produce binocular results. The Esterman test, the only standard 53 54 23 binocular field test available, is used to determine the extent of visual fields in UK drivers43, 55 56 24 although it was not originally designed for this purpose40. Although some studies suggest the 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 7 of 34 BMJ Open
1 2 3 25 Esterman test is a good predictor of visual function36,44,45, its limitations are well documented 4 5 26 and include a substantial ceiling effect which limits the test’s applicability for people with low 6 7 27 to moderate levels of field loss2,10,29,37,38,44,46,47, and an inability to differentiate effectively 8 9 28 between people with different levels of field loss38. As an alternative, Crabb and 10 11 29 Viswanathan42 constructed integrated visual fields from monocular field plots using software 12 13 30 that estimates sensitivities at points of overlapping locations using a “best location” 14 15 31 algorithm21,38For. While integrated peer visual fields review provide a rapid estimate only of a patient’s binocular 16 17 32 field without extra perimetric examination21,23,48, they do assume that previous visual field 18 19 33 results are available which is not always the case in low vision clinics. Binocular function is 20 21 34 derived rather than measured, and the plots used usually only assess the central 30 deg of 22 23 35 visual field. 24 25 26 The purpose of this study is to develop an optimal binocular visual field test to reflect the 27 36 28 functional consequences of visual field loss by assessing threshold, suprathreshold and 29 37 30 31 38 kinetic binocular visual field paradigms examining the visual field to 60 deg eccentricity, and
32 http://bmjopen.bmj.com/ 33 39 existing standard Esterman and IVF fields. Results are compared to self reported difficulty 34 35 40 with mobility to determine which visual field assessment paradigms best reflect functional 36 37 41 difficulty. The visual field testing experience of individuals with visual impairment is also 38 39 42 considered to help devise optimal strategies for functional field assessment.
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 Methods 43 44 45 44 Participants with self reported peripheral visual field loss were recruited for this cross 46 47 45 sectional observational study through advertisement of the study with local voluntary agency 48 49 46 Cam Sight, and the charities Retinitis Pigmentosa Fighting Blindness and the International 50 51 47 Glaucoma Association. Recruiting participants with a range of ocular conditions allowed the 52 53 48 assessment of a range of visual field loss, and reflects the need of a functional field 54 55 49 assessment to be applicable to patients with a range of visual disorders. Individuals with 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 8 of 34
1 2 3 50 conditions not primarily affecting peripheral visual function, such as macular degeneration, 4 5 51 were excluded from the study, along with those under 18 years old and those unable to 6 7 52 perform verbal evaluations in English. Ethical approval was granted by Anglia Ruskin 8 9 53 University Faculty of Science and Technology Research Ethics committee. All relevant 10 11 54 tenets of the Declaration of Helsinki were upheld. All participants gave informed consent 12 13 55 after the nature of the study was explained. Participants attended a single session at the 14 15 56 University EyeFor Clinic, Angliapeer Ruskin University, review with data collected only between 3/7/2015 and 16 17 57 4/3/2016 by a single researcher (HS), a qualified optometrist. 18 19 20 58 Part 1 of the Independent Mobility Questionnaire (IMQ)29,49 was used to assess self 21 22 59 perceived ability in mobility. Participants were asked to report if they encountered difficulty in 23 24 60 each of 35 mobility situations (Table 1) to obtain a binary response (Yes/No), and then to 25 26 rate the level the difficulty they experienced on a 5 point Likert scale. Respondents could 27 61 28 indicate that a mobility task was not applicable to them, which was scored as missing data. 29 62 30 31
32 How difficult is it for you to undertake the following tasks independently/without http://bmjopen.bmj.com/ 33 assistance (but with the aid of any mobility aids as required): 19. Adjusting to lighting changed during the 34 1. Walking in familiar areas 35 day: indoor to outdoor 20. Adjusting to lighting changed during the 36 2. Walking in unfamiliar areas 37 day: outdoor to indoor 38 21. Adjusting to lighting changed at night: 3. Moving about: at home 39 indoor to streetlights
40 22. Adjusting to lighting changed at night: on September 26, 2021 by guest. Protected copyright. 4. Moving about: at work 41 streetlights to indoor 42 43 5. Moving about: in the classroom 23. Walking in dimly lit indoor areas 44 24. Being aware of another person’s 6. Moving about: in stores 45 presence 46 47 7. Moving about: outdoors 25. Avoiding bumping into: people 48 49 8. Moving about: in crowded situations 26. Avoiding bumping into: walls 50 27. Avoiding bumping into: head height 51 9. Walking at night objects 52 28. Avoiding bumping into: shoulder height 53 10. Using public transportation 54 objects 29. Avoiding bumping into: waist height 55 11. Detecting ascending stairwells 56 objects 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 9 of 34 BMJ Open
1 2 3 30. Avoiding bumping into: knee height 12. Detecting descending stairwells 4 objects 5 6 13. Walking up steps 31. Avoiding bumping into: low lying objects 7 32. Avoiding tripping over uneven travel 14. Walking down steps 8 surfaces 9 10 15. Stepping onto curbs 33. Moving around in social gatherings 11 12 16. Stepping off curbs 34. Finding restrooms in public places 13 14 17. Walking through doorways 35. Seeing cars at intersections 15 For peer review only 16 18. Walking in high glare areas 17 18 63 19 20 21 64 Table 1. 35 mobility situations from Part 1 of the Independent Mobility Questionnaire 29,49 22 65 (IMQ) . 23 24 66 Participants were also asked to report whether they had fallen in the last 12 months, defined 25 26 67 as an event which results in a person coming to rest inadvertently on the ground, floor or 27 28 68 other lower level50. 29 30 31 69 High contrast distance visual acuity with habitual distance spectacle correction was
32 http://bmjopen.bmj.com/ 33 70 assessed binocularly and scored by letter51 using a 3m internally illuminated EDTRS chart52. 34 35 71 If the largest letters could not be read at 3m, the chart was moved 50% closer to the 36 37 72 participant to 1.5m and 0.75m. Participants with acuity that was not measureable were 38 39 73 assigned an acuity of 3.00logMAR53. Contrast sensitivity was measured binocularly with
40 on September 26, 2021 by guest. Protected copyright. 41 74 habitual distance spectacle correction using a Pelli Robson Chart54 at 1m scored on a by 42 43 75 letter basis55. Participants with no measurable CS function were assigned a score of 44 45 53 76 0.00logCS . 46 47 48 56 49 77 Five visual field assessments were performed using the Octopus 900 Perimeter and the
50 57 51 78 Humphrey Field Analyser . Visual field test order was randomised, and regular breaks were 52 53 79 provided. In all assessments, participants responded to seeing a visual field stimulus by 54 55 80 pressing a response button. The duration of all assessments was noted. The standard size 56 57 81 III Goldmann white stimulus was used throughout. Participants fixated on the standard 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 10 of 34
1 2 3 82 orange central point target (HFA) or green cross target (Octopus 900). When participants 4 5 83 had difficulty seeing the target or maintaining binocular fixation, a 6 deg ring target was 6 7 84 utilised on the Octopus, and an adaption to the fixation target on the HFA that slotted into the 8 9 85 fixation target hole to provide a black 2mm high contrast pericentral ring around the fixation 10 11 86 spot were used. 12 13 14 87 The three tests conducted on the Octopus 900 Perimeter were as follows, using the 15 For peer review only 16 88 ‘binocular’ setting in all cases: 17 18 19 89 1. Binocular threshold 20 21 A custom test point pattern was used, assessing the binocular field to 60 deg from 22 90 23 24 91 fixation with 52 points spaced every 7.5 deg in the central 30 deg, and 36 points 25 26 92 spaced every 15 deg in the peripheral 30 60 deg. Full aperture trial lenses were used 27 28 93 in adult half eye trial frames with lens centration distances corrected for near for 29 30 94 assessment of the central 30 deg. The peripheral 30 60 deg field was assessed 31
32 95 without correction to minimise the possibility of lens and frame artefacts. The ‘low http://bmjopen.bmj.com/ 33 34 96 vision’ test strategy was utilised: stimuli are presented using a 4 2 1 dB bracketing 35 36 97 test method starting at 0dB (4000asb) in order to arrive quickly at the expected 37 38 98 threshold level in subjects with impaired visual fields. The absolute thresholds 39 99 achieved at each test location were used to calculate the mean threshold39,42. 40 on September 26, 2021 by guest. Protected copyright. 41 42 100 43 44 101 2. Binocular suprathreshold 45 46 102 The same custom test pattern as for binocular threshold assessment was utilized. 47 48 103 Each point was assessed with a stimulus of 10dB intensity. The number of points 49 50 104 seen from the total of 88 was used to calculate a percentage score. 51 52 105 53 54 106 3. Binocular kinetic 55 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 11 of 34 BMJ Open
1 2 3 107 A white III 4e (10dB) target was presented kinetically at an angular velocity of 5 4 5 108 deg/sec. Vectors were presented from 60 deg eccentricity and moved centrally in 12 6 7 109 meridians spaced every 30 degrees. This assessment was completed without 8 9 110 refractive correction. The solid angle (deg2) subtended by the isopter was determined 10 11 111 automatically using the Eye Suite software55. The perimeter corrected results for 12 13 112 reaction time bias. 14 15 For peer review only 16 113 The two tests conducted on the HFA were as follows: 17 18 19 114 4. Integrated monocular threshold fields (IVF) 20 21 The central 24 2 threshold test with SITA Fast strategy was used to assess 22 115 23 24 116 monocular visual fields. Full aperture trial lenses provided refractive correction. 25 26 117 Integrated visual field scores were calculated using the best location
27 21,23,24,37,38,42,47,58 60 28 118 algorithm . 29 30 119 31
32 120 5. Esterman http://bmjopen.bmj.com/ 33 34 121 The binocular Esterman visual field test examines 120 test points out to 80 deg 35 36 122 eccentricity using a suprathreshold paradigm with a stimulus intensity of 10dB. This 37 38 123 assessment was completed without refractive correction. The number of points seen 39 124 was used to calculate the percentage Esterman Efficiency score. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 125 Participants were asked to rank the field tests in order from 1 to 5 based on how acceptable 44 45 126 they felt the tests were, and also to rank the perceived usefulness of results presented. 46 47 127 Participants were also invited to make further comments relating to acceptability and output 48 49 128 of the tests, and qualitative themes were evaluated from this data. 50 51 52 129 Analysis 53 54 55 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 12 of 34
1 2 3 130 Interval data was derived from the ordinal IMQ responses by the use of Rasch analysis61. 4 5 131 The Rasch model, a probabilistic logistic model, is a paradigm for the analysis and scoring of 6 7 132 questionnaires that can be used to produce interval level data from ordinal responses, and in 8 9 133 turn improves sensitivity to change in function62 and correlations with other variables62,63, 10 11 134 allows the use of robust parametric statistics on the data63, and provides more accurate 12 13 135 measurements of perceived function64. In the Rasch model, items and respondents are 14 15 136 scaled accordingFor to responses peer to a groupreview of items61,65. The only underlying construct being 16 17 137 assessed is used to define the relative difficulty of each item. On the same linear scale of the 18 19 138 construct, respondents are ordered from least to most ability, and items are ordered from 20 21 139 most to least difficult. 22 23 24 140 Rasch analysis derives person and item measures in logits from raw ordinal data. Person 25 26 measures are an estimate of a person’s underlying ability based on their performance on a 27 141 28 set of items that measure a single trait. The item measure is the Rasch estimate of item 29 142 30 31 143 difficulty.
32 http://bmjopen.bmj.com/ 33 34 144 Person measures were derived using all 35 items, which have been found to constitute a 35 29 49 36 145 unidimensional scale in people with peripheral field loss due to RP and glaucoma . Higher 37 38 146 person measures indicate greater perceived ability. 39
40 on September 26, 2021 by guest. Protected copyright. 41 147 The reliability indices were assessed in terms of person separation statistics, which provide 42 43 148 an indication of the instrument’s ability to discriminate between respondents: person 44 45 149 separation and person reliability should be greater than the suggested minima of 2.0 and 46 47 150 0.80 respectively66. Further, item separation statistics provide an indication of how reliably 48 49 151 ordered the items are in terms of difficulty: item separation and item reliability should be in 50 51 152 excess of suggested minima of 3.0 and 0.9066. Targeting, or the difference between mean 52 53 153 item and person measures, should ideally be less than 1.0 logit67,68. 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 13 of 34 BMJ Open
1 2 3 154 Additionally, the fit of individual items to a unidimensional construct is assessed. It is 4 5 155 considered that items with infit and outfit meansquare (mnsq) values within a range of 0.5 to 6 7 156 1.5 contribute usefully to a scale68. Items with mnsq values greater than two have the 8 9 157 potential to damage the integrity of the scale69. 10 11 12 158 Mann Whitney U tests compared visual function in participants who had reported a fall in the 13 14 159 previous 12 months and those who had not. Continuous clinical function variables were 15 For peer review only 16 160 compared to self reported mobility function in non parametric 2 tailed Spearman’s rho 17 18 161 bivariate correlations. 19 20 21 For the binary responses to the 35 mobility tasks of the IMQ, difficulty was compared to 22 162 23 24 163 different visual field test scores using receiver operating characteristic (ROC) analysis to 25 26 164 evaluate how effective the tests were at selecting participants with perceived mobility 27 28 165 difficulty (sensitivity), and without perceived mobility difficulty (specificity). Sensitivity and 29 30 166 specificity were determined for all possible cut off values for the visual field scores and 31
32 167 plotted as ROC curves. An area under the ROC curve of 1 indicates a perfect diagnostic http://bmjopen.bmj.com/ 33 34 168 procedure, whereas 0.5 indicates a poor procedure. A statistical technique appropriate 35 70 36 169 where two measures are applied to the same set of participants was used to compare 37 38 170 areas under the ROC curves and establish if any visual field test was statistically significantly 39 171 better at predicting perceived difficulty. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 172 Results 44 45 46 173 Table 2 shows the characteristics of the 50 participants who were recruited to the study: all 47 48 174 completed the study. Sample size is consistent with previous similar studies 1,29,34,42. Person 49 50 175 measures for the IMQ are shown. Other parameters derived from the Rasch analysis include 51 52 176 person separation of 3.43 (reliability 0.92), indicating that individuals can be reliably ordered 53 54 177 by the instrument in terms of their level of perceived ability. Item separation is 2.95 (reliability 55 56 178 0.90), slightly less than the minimum ideal value of 3, and indicating the instrument might not 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 14 of 34
1 2 3 179 be able to reliably order items in terms of their difficulty. Targeting ( 1.23±1.64 logits) is close 4 5 180 to the ideal of within ±1 logits of the mean item difficulty7. In terms of item fit, as a 6 7 181 representation of how well the questions fitted a unidimensional construct, there were three 8 9 182 mis fitting items with fits in the range 1.5 to 2.0 meansquare. These fits do not diminish the 10 11 183 validity of the measures and can be considered acceptable71, and so all items are 12 13 184 considered in the analysis. Item fits are also comparable to other Rasch analyses of this 14 15 185 instrument, For with three itemspeer with slight review mis fit also found foronly samples with RP29,72 and 16 17 186 glaucoma49. Reliability statistics indicate adequate visual field reliability for the majority of the 18 19 187 sample. Fixation losses for the monocular threshold tests suggest reliable results, with only 20 21 188 18% of participants losing fixation more than 20% during both assessments. 90% of false 22 23 189 positive and false negative statistics from binocular threshold data were less than 20%. This 24 25 190 figure is similar to statistics obtained from Esterman (86%) and binocular suprathreshold 26 27 191 results (78%) results. 28 29 30 31 192
32 http://bmjopen.bmj.com/ 33 34 Demographic variables 35 Gender (n) 29 Males, 21 Females 36 Age (years) 37 Median (25% IQ 75% IQ) 64(55 71) 38 Min max 24 84 39 Ocular diagnosis (n)
40 RP 14 on September 26, 2021 by guest. Protected copyright. 41 Glaucoma 23 42 Retinal detachments/tears 4 43 Other 9 44 Registration status (n) 45 Registered severely sight 18 46 impaired 47 Registered sight impaired 8 48 49 Not registered 24 50 Use of mobility aids (n) 51 White cane or guide dog 23 52 No mobility aids used 27 53 Have you been shown your visual field 54 results before? (n) 55 Yes 36 56 No 14 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 15 of 34 BMJ Open
1 2 3 Have you fallen in the previous 12mos? 4 Yes 28 5 No 22 6 Clinical function variables Mean (±std) Median (25% Range 7 IQ-75% IQ)) 8 Binocular distance visual acuity 0.28(±0.08) 0.09( 0.06 0.50) 0.28 3.00 9 (logMAR) 10 Binocular contrast sensitivity 1.51(±0.07) 1.65(1.30 1.95) 0.00 1.95 11 (logCS units) 12 Binocular visual field variables 13 Binocular threshold (dB) 10.87(±1.19) 10.14(2.13 0.27 25.60 14 19.40) 15 For peer review only 16 Binocular suprathreshold (%) 54.48(±5.09) 58.53(18.8 2.27 98.86 17 93.18) Binocular kinetic solid angle 5966.77(±541.19) 7355.7(1783.80 64.20 18 2 19 (deg ) 9566.70) 10320.50 20 Esterman (%) 59.43(±4.81) 67.08(33.33 0.00 21 90.83) 100.00 22 Integrated monocular threshold 15.69(±1.52) 15.17(4.88 0.90 31.96 23 (dB) 26.48) 24 Independent Mobility Questionnaire 1.23(±0.23) 1.26( 2.29 ( 5.92 1.84 25 Person Measures (logits) 0.09)) 26 27 193 Table 2. Descriptive statistics of the variables assessed (n=50). The mean ± standard 28 194 deviation, and the median (interquartile range) are given for the clinical visual function 29 195 variables. *Number of comorbid conditions from a list of 12 common medical conditions 30 196 representing general health status73. 31
32 197 http://bmjopen.bmj.com/ 33 34 35 Clinical function variables IMQ score (R2) Have you fallen in the 36 previous 12mos? (U) 37 38 Distance visual acuity (logMAR) 0.31, p≤0.001 288.0, p=0.701 39 Contrast sensitivity (logCS units) 0.33, p≤0.001 302.5, p=0.913 Binocular threshold (dB) 0.47, p≤0.001 236.0, p=0.157
40 on September 26, 2021 by guest. Protected copyright. 41 Binocular suprathreshold (%) 0.47, p≤0.001 235.0, p=0.161 2 42 Binocular kinetic solid angle (deg ) 0.48, p≤0.001 236.0, p=0.159 43 Esterman (%) 0.46, p≤0.001 209.0, p=0.053 44 IVF (dB) 0.38, p≤0.001 235.0, p=0.149 45 46 198 Table 3. Relationship between the variables assessed, and self reported mobility function 47 199 and falls history. Mann Whitney U values are provided for falls data and Spearman’s rho 48 200 bivariate correlations are provided for the IMQ score. 49 50 201 51 52 202 Table 3 shows the relationships between the parameters assessed and the outcome 53 54 203 measures of self reported mobility function, and fall history. Greater visual field loss is 55 56 significantly associated with greater self reported difficulty regardless of the method of field 57 204 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 16 of 34
1 2 3 205 assessment, although the relationship is a little weaker for the IVF score. 56% of the sample 4 5 206 reported falling at least once in the previous 12 months, but none of the clinical function 6 7 207 variables significantly associate with falls history (Table 3). 8 9 10 208 Figure 1. 11 12 13 209 The ability of the VF tests to discriminate between people with and without difficulty with 14 15 210 mobility tasksFor was reasonable peer (AUC>0.7) review for at least one VF onlytest for all except 2 of the 35 16 17 211 questions (Figure 1). Statistically significant differences in the areas under the ROC curves 18 19 212 were seen (indicated by * in Figure 1), in that binocular threshold and supra threshold 20 21 assessments were better than the IVF at predicting difficulty walking in familiar areas, 22 213 23 24 214 walking in unfamiliar areas, walking at home, walking in crowded areas, avoiding bumping 25 26 215 into knee height objects, and finding public toilets. The binocular threshold assessment was 27 28 216 also better than the IVF at predicting difficulty avoiding bumping into people, while the 29 30 217 Esterman was found to better predict difficulty walking in high glare when compared with the 31
32 218 IVF. All three custom tests and the Esterman assessment were better than the IVF at http://bmjopen.bmj.com/ 33 34 219 predicting difficulty avoiding bumping into waist height objects, and at predicting difficulty 35 36 220 avoiding bumping into low lying objects. 37 38 39 221 Figure 2.
40 on September 26, 2021 by guest. Protected copyright. 41 42 222 Figure 3. 43 44 45 223 The length of time taken to undertake each visual fields assessment is shown in Figure 2. 46 47 224 The quickest test was the kinetic which took a mean of 1min 26sec (±9sec), while the mean 48 49 225 duration of the longest test, the IVF, was 9min 23sec (±24 sec). 50 51 52 In terms of the participants’ view of how they found the tests to undertake, the most favoured 53 226 54 55 227 assessment was the kinetic (Figure 3), while the IVF was ranked the least favourite test by 56 57 228 over 60% of participants. In addition to ranking the tests, participants also made comments 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 17 of 34 BMJ Open
1 2 3 229 on the different test strategies, and common themes are considered here. Three participants 4 5 230 (6%) commented that shorter tests were preferred, the kinetic assessment in particular. 6 7 231 Participants found the kinetic assessment pleasant, less stressful, and encouraging, with 8 9 232 three participants commenting on being reassured by knowing that a light would be seen 10 11 233 eventually. Seven participants (14%) suggested that the kinetic assessment was more fun 12 13 234 and engaging than the static tests, and one remarked on the assessment’s novelty value. 14 15 235 However, sixFor participants peer (12%) expressed review concern that the testonly was too basic or too short, 16 17 236 and that the accuracy of results would be compromised by the test’s rapidity. Participants 18 19 237 were happy to conduct a longer test if they knew that results would be more beneficial to the 20 21 238 practitioner. Eleven participants (22%) indicated preference for binocular tests, which were 22 23 239 reported as more comfortable and less tiring than monocular assessments. Five participants 24 25 240 (10%) preferred the uniformly bright lights on the binocular suprathreshold assessment; 26 27 241 however, four (8%) found the bright lights at the start of the binocular threshold assessment 28 29 242 encouraging. 30 31
32 http://bmjopen.bmj.com/ 33 243 All participants in the study reported previous experience of visual field testing, although 34 35 244 28% had not seen their visual field results before (Table 2). All perceived value in being 36 37 245 shown their visual fields. In terms of visual field outputs, 48% of participants preferred the 38 39 246 grey scale plots from the binocular threshold analysis, and individuals remarked on the
40 on September 26, 2021 by guest. Protected copyright. 41 247 greater level of detail provided on a grey scale plot compared to others. The kinetic plot was 42 43 248 most favoured by 37% of participants, with preference expressed for plots that indicated 44 45 249 greater levels of residual field. Four participants commented on the usefulness of having 46 47 250 their peripheral field represented, with one participant suggesting that the monocular 48 49 251 threshold plots were an inaccurate depiction that overestimated the extent of their visual 50 51 252 field. The Esterman output was the least favoured by 38% of participants. Three participants 52 53 253 commented on the size of points on suprathreshold outputs, and expressed difficulty viewing 54 55 254 the results. 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 18 of 34
1 2 3 255 Discussion 4 5 6 256 Five different visual field assessments were compared to self reported difficulty with mobility, 7 8 257 in order to determine which visual field paradigms are suitable for use as a functional visual 9 10 258 field assessment. There are four principal findings of this study. Firstly, the paradigm used to 11 12 259 assess the visual field (threshold or suprathreshold static, or kinetic) makes little difference 13 14 260 to the relationship with function: so long as the test is performed binocularly and includes 15 For peer review only 16 261 assessment of eccentricities to 60 deg, the visual field outcome measure reflects self 17 18 262 reported mobility function well, in terms of both correlation between visual field score and 19 20 263 mobility function, and in terms of ability of visual field scores to discriminate between 21 22 264 individuals who do and do not have difficulty with specific mobility tasks. Secondly, tests that 23 24 265 are quicker to perform (binocular suprathreshold, binocular kinetic, Esterman; Figure 2) 25 26 relate just as well to self reported mobility function as tests that take longer (binocular 27 266 28 threshold; Table 2). Functional information is not lost by using kinetic or suprathreshold 29 267 30 31 268 techniques when compared to the diagnostic gold standard of measuring static thresholds.
32 http://bmjopen.bmj.com/ 33 269 Thirdly, quicker tests are also more acceptable to patients (Figure 3), especially the 34 35 270 binocular suprathreshold and kinetic paradigms. Finally, although all participants had 36 37 271 experience of visual field assessments, over a quarter did not recall being shown their visual 38 39 272 field results before. All participants in this study found value in seeing their visual field
40 on September 26, 2021 by guest. Protected copyright. 41 273 results, and clinicians are encouraged to explain visual field results with respect to likely 42 43 274 functional difficulties where possible. 44 45 46 275 A strength of this study is that although some previous studies45,74 76 have compared 47 48 276 different methods of quantifying the visual field and their relationship with functional vision, 49 50 277 we are not aware of any previous studies that have compared different paradigms of visual 51 52 278 field assessment with functional vision, nor that have taken the patient’s perspective into 53 54 279 consideration. Rigorous and novel visual field testing on a cohort of people with visual field 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 19 of 34 BMJ Open
1 2 3 280 loss has been used, along with Rasch analysed data from a validated questionnaire to 4 5 281 reflect a broad range of mobility function. 6 7 8 282 One possible limitation of the study is that we rely on self report of mobility difficulties rather 9 10 283 than directly measuring visual performance on, for example, a mobility circuit12,31. Another 11 12 284 potential limitation is that single responses have been used in the kinetic paradigm. 13 14 285 Responses to kinetic stimuli can be variable77, and repeated presentations might be 15 For peer review only 16 286 necessary to confirm responses, and reduce the impact of outlying responses78. 17 18 19 287 The findings are consistent with previous studies, in that we find that the IVF relates to 20 21 mobility function58,79. Contrary to other studies however38,42, the IVF appears to be less 22 288 23 24 289 effective at relating to self related mobility function when compared to the Esterman 25 26 290 assessment. This may be due to the difference in the degree of visual field loss between the 27 28 291 sample groups. The average Esterman scores in the current study (56.4%) are less than the
29 37 41 30 292 average scores (87.4% and 86.7% ) reported in other studies, and would suggest that the 31
32 293 present sample has a greater degree of visual field loss. Contrary to some previous studies, http://bmjopen.bmj.com/ 33 34 294 no significant correlations were found here between measures of clinical function and fall 35 6,11,13 36 295 history. Visual field loss has been shown to increase the risk of falling in some studies 37 18 80,81 38 296 , but not in others . Variation in findings reflects the multifactorial nature of falls, and also 39 297 suggests limitations in retrospective reporting of falls82. 40 on September 26, 2021 by guest. Protected copyright. 41 42 43 298 Participants with a range of ocular conditions affecting peripheral vision (but predominantly 44 45 299 glaucoma and retinitis pigmentosa) and a range of severity of visual field loss were used in 46 47 300 this study. Results would be expected to be generalizable to people with mild to severe 48 49 301 peripheral field loss, but not necessarily to those with visual impairment affecting central 50 51 302 visual function, such as macular degeneration. 52 53 54 303 In conclusion, a binocular visual field test that does not ignore the peripheral 30 60 deg of 55 56 304 the field is effective for reflecting functional difficulty, regardless of the whether a threshold, 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 20 of 34
1 2 3 305 suprathreshold or kinetic assessment paradigm is employed. A shorter duration visual field 4 5 306 test such as the binocular kinetic assessment used in this study is favoured by patients, and 6 7 307 this study did not find significantly different effectiveness for this method compared to more 8 9 308 time consuming tests in predicting perceived disability in patients with peripheral field loss. 10 11 309 To further explore the potential of a binocular kinetic assessment as a functional field test, 12 13 310 future work needs to explore the repeatability of responses, the necessity for repeated 14 15 311 presentations,For and the assessment’speer ability review to discriminate between only individuals with normal 16 17 312 and abnormal fields using control subjects. 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 21 of 34 BMJ Open
1 2 3 References 4 5 6 313 1. Bibby SA, Maslin ER, McIlraith R, Soong GP. Vision and self reported mobility 7 8 314 performance in patients with low vision. Clin Exp Optom. 2007;90(2):115 23. 9 10 315 2. Nelson P, Aspinall P, Papasouliotis O, Worton BO, Brien C. Quality of life in 11 12 316 glaucoma and its relationship with visual function. J Glaucoma. 2003;12(2):139 150. 13 14 317 3. Tabrett DR, Latham K. Factors influencing self reported vision related activity 15 For peer review only 16 318 limitation in the visually impaired. Invest Ophthal Vis Sci. 2011;52:5293–5302. 17 18 319 4. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 19 20 320 field loss in the elderly and associations with impairment in daily functioning: the 21 22 321 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 23 24 322 5. Sherwood M, Garcia Siekavizza A, Meltzer M, et al. Glaucoma’s impact on quality of 25 26 life and its relation to clinical indicators. Ophthalmology. 1998;105:561–566. 27 323 28 6. Ivers RQ, Cumming RG, Mitchell P, et al. Visual impairment and falls in older adults: 29 324 30 31 325 the Blue Mountains Eye Study. J Am Geriatr Soc. 1998;46:58–64.
32 http://bmjopen.bmj.com/ 33 326 7. Latham K, Baranian M, Timmis MA, Pardhan S. Difficulties with Goals of the Dutch 34 35 327 ICF Activity Inventory: Perceptions of Those with Retinitis Pigmentosa and of Those 36 37 328 Who Support Them. Inves Ophthal Vis Sci. 2015;56(4):2381 2391. 38 39 329 8. Lowe J, Drasdo N. Patients' responses to retinitis pigmentosa. Optom Vis Sci.
40 on September 26, 2021 by guest. Protected copyright. 41 330 1992;69 (3):182 185. 42 43 331 9. Szlyk JP, Fishman GA, Grover S, Revelins BI, Derlacki DJ. Difficulty in performing 44 45 332 everyday activities in patients with juvenile macular dystrophies: comparison with 46 47 333 patients with. Br J Ophthalmol. 1998;82(12):1372 1376. 48 49 334 10. Noe G, Ferraro J, Lamoureux E, Rait J, Keeffe JE. Associations between 50 51 335 glaucomatous visual field loss and participation in activities of daily living. Clin Exp 52 53 336 Optom. 2003;31(6):482 6. 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 22 of 34
1 2 3 337 11. Ramrattan RS, Wolfs RC, Panda Jonas S, et al. Prevalence and caused of visual 4 5 338 field loss in the elderly and associations with impairment in daily functioning: the 6 7 339 Rotterdam Study. Arch Ophthalmol. 2001;119:1788 1794. 8 9 340 12. Lovie Kitchin JE, Soong GP, Hassan SE, Woods RL. Visual field size criteria for 10 11 341 mobility rehabilitation referral. Optom Vis Sci. 2010;87(12):948 57. 12 13 342 13. Patino CM, McKean Cowdin R, Azen SP, Allison JC, Choudhury F, Varma R. Central 14 15 343 and PeripheralFor Visualpeer Impairment review and the Risk of Falls onlyand Falls with Injury. 16 17 344 Ophthalmology. 2010;117(2):199. 18 19 345 14. Klein BEK, Klein R, Lee KE, Cruickshanks KJ. Performance based and self 20 21 346 assessed measures of visual function as related to history of falls, hip fractures, and 22 23 347 measured gait time: The Beaver Dam Study. Ophthalmology. 1998;105(1):160 164. 24 25 348 15. Jack CIA, Smith T, Neoh C, Lye M, Mcgalliard JN. Prevalence of low vision in elderly 26 27 349 patients admitted to an acute geriatric unit in Liverpool: Elderly people who fall are 28 29 350 more likely to have low vision. Gerontology. 1995;41(5):280 285. 30 31 351 16. Freeman EE, Munoz B, Rubin GS, West SK. Visual field loss increases the risk of 32 http://bmjopen.bmj.com/ 33 falls in older adults: the Salisbury eye evaluation. Inves Ophthal Vis Sci. 34 352 35 36 353 2007;48(10):4445 4450. 37 38 354 17. Klein BEK, Moss SE, Klein R, Lee KE, Cruickshanks KJ. Associations of visual 39 355 function with physical outcomes and limitations 5 years later in an older population:
40 on September 26, 2021 by guest. Protected copyright. 41 42 356 The Beaver Dam eye study. Ophthalmology. 2003;110(4):644 650. 43 44 357 18. Haymes SA, LeBlanc RP, Nicolela MT, Chiasson LA, Chauhan B.C. Risk of falls and 45 46 358 motor vehicle collisions in glaucoma. Invest. Ophthalmol. Vis. Sci. 2007;48:1149 47 48 359 1155. 49 50 360 19. McGwin G Jr, Xie A, Mays A, Joiner W, DeCarlo DK, Hall TA, Owsley C. Visual field 51 52 361 defects and the risk of motor vehicle collisions among patients with glaucoma. Invest 53 54 362 Ophthalmol Vis Sci. 2005;46(12):4437 41. 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 23 of 34 BMJ Open
1 2 3 363 20. Subhi H, Latham K, Myint J, Crossland M. Functional visual fields: relationship of 4 5 364 visual field areas to self reported function. Ophthalmic Physiol Opt. 2017;37:399– 6 7 365 408. 8 9 366 21. Nelson Quigg JM, Cello K, Johnson CA. Predicting binocular visual field sensitivity 10 11 367 from monocular visual field results. Inves Ophthal Vis Sci. 2000;41:2212–2221. 12 13 368 22. Schneck ME, Haegerstom Portnoy G, Lott L, Brabyn JA. Monocular vs. Binocular 14 15 369 MeasurementFor of Spatialpeer Vision inreview Elders. Optom Vis Sci. only 2010;87(8):526 531. 16 17 370 23. Asaoka R, Crabb D, Yamashita T, Russell R, Wany YX, Garway Heath DF. Patients 18 19 371 Have Two Eyes!: Binocular versus Better Eye Visual Field Indices. Inves Ophthal Vis 20 21 372 Sci. 2011;52(9):7007 7011. 22 23 373 24. Crabb DP, Smith ND, Glen FC, Burton R, Garway Heath DF. How does glaucoma 24 25 374 look? Patient perception of visual field loss. Ophthalmology. 2013;120(6):1120 6. 26 27 375 25. Gardiner SK, Demirel S. Assessment of patient opinions of different clinical tests 28 29 376 used in the management of glaucoma. Ophthalmology. 2008;115:2127–31. 30 31 377 26. Glen FC, Baker H, Crabb DP A qualitative investigation into patients’ views on visual 32 http://bmjopen.bmj.com/ 33 field testing for glaucoma monitoring BMJ Open 2014;4(1). 34 378 35 36 379 27. Hudson C, Wild JM, O’Neill EC. Fatigue effects during a single session of automated 37 38 380 static threshold perimetry. Invest Ophthalmol Vis Sci. 1994;35:268–80. 39 381 28. Artes, P, Monhart M, Marco M. Are big stimuli “easier” to see? An analysis of
40 on September 26, 2021 by guest. Protected copyright. 41 42 382 response times from frequency of seeing data. In 22nd International Visual Field 43 44 383 Imaging Symposium. Udine, Italy, September 27th 30th 2016. 45 46 384 29. Turano KA, Rubin GS, Quigley HA Mobility performance in glaucoma. Invest 47 48 385 Ophthalmol Vis Sci. 1999;40:2803 2309. 49 50 386 30. Lovie Kitchin JE, Mainstone J, Robinson J, Brown B. What areas of the visual field 51 52 387 are important for mobility in low vision patients? Clin Vis Sci. 1990;5:249 263. 53 54 388 31. Haymes SA, Johnston AW, Heyes AD. Relationship between vision impairment and 55 56 389 ability to perform activities of daily living. Ophthalmic Physiol Opt. 2002;22(2):79 91. 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 24 of 34
1 2 3 390 32. Haymes SA, Leblanc RP, Nicolela MT et al. Risk of falls and motor vehicle collisions 4 5 391 in glaucoma. Invest Ophthalmol Vis Sci. 2007;48:1149–55. 6 7 392 33. Hassan SE, Hicks JC, Lei H, Turano KA. What is the minimum field of view required 8 9 393 for efficient navigation? Vision Res. 2007;47(16):2115 2123. 10 11 394 34. Lee J, Cho H, Kee C. Assessment of the Vision Specific Quality of Life Using 12 13 395 Binocular Esterman Visual Field in Glaucoma Patients. J Korean Ophthalmol Soc, 14 15 396 2013;54(10):1567.For peer review only 16 17 397 35. Mills RP, Drance SM. Esterman disability rating in severe glaucoma. Ophthamology. 18 19 398 1986;93:371 378. 20 21 399 36. Fujita K, Yasuda N, Nakamoto K, Fukuda T. The relationship between difficulty in 22 23 400 daily living and binocular visual field in patients with glaucoma. Nihon Ganka Gakkai 24 25 401 Zasshi. 2008;112(5):447 50. 26 27 402 37. Jampel HD. Glaucoma patients’ assessment of their visual function and quality of life. 28 29 403 Tr Am Ophth Soc. 2001;99:301 317. 30 31 404 38. Jampel HD, Friedman DS, Quigley H, et al. Correlation of the binocular visual field 32 http://bmjopen.bmj.com/ 33 with patients’ assessment of vision. Inves Ophthal Vis Sci. 2002;43:1059 67. 34 405 35 36 406 39. Tabrett DR, Latham K. Important areas of the central binocular visual field for daily 37 38 407 functioning in the visually impaired. Ophthalmic Physiol Opt. 2012;32:156 163. 39 408 40. Black A, Lovie Kitchin J, Woods R, Arnold N, Byrnes J, Murrish J. Mobility
40 on September 26, 2021 by guest. Protected copyright. 41 42 409 performance with retinitis pigmentosa. Clin Exp Optom. 1996;80(1):1 12. 43 44 410 41. Esterman B. Functional scoring of the binocular field. Ophthalmology. 1982;89:1226– 45 46 411 34. 47 48 412 42. Crabb DP, Viswanathan AC. Integrated visual fields: a new approach to measuring 49 50 413 the binocular field of view and visual disability. Graefes Arch Clin Exp Ophthalmol. 51 52 414 2004;243(3):210 6. 53 54 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 25 of 34 BMJ Open
1 2 3 415 43. Driver and Vehicle Licensing Agency. Visual disorders: assessing fitness to drive. 4 5 416 2016 Available: https://www.gov.uk/guidance/visual disorders assessing fitness to 6 7 417 drive. Last accessed February 2017. 8 9 418 44. Parrish RK II, Gedde SJ, Scott IU, et al. Visual function and quality of life among 10 11 419 patients with glaucoma. Arch Ophthalmol. 1997;115:1447 1455. 12 13 420 45. Choy ES, Mills RP, Drance SM. Automated Esterman testing of disability in 14 15 421 glaucoma.For In:Greve peer E, Heiji A, eds. review Seventh International only Visual Field Symposium. 16 17 422 Amsterdam, the Netherlands: Junk Publishers. 1986:527 535. 18 19 423 46. Mills RP. Correlation of quality of life with clinical symptoms and signs at the time of 20 21 424 glaucoma diagnosis. Trans Am Ophthalmol Soc. 1998;96:753 812. 22 23 425 47. Harris ML, Jacobs, NA. Is the Esterman binocular field sensitive enough? In: Mills 24 25 426 RP, Wall M. Perimetry update 1994/1995. Amsterdam: Kugler, 1995:13–24. 26 27 427 48. Crabb DP, Viswanathan AC, McNaught AI, et al. Simulating binocular visual field 28 29 428 status in glaucoma. Br J Ophthalmol. 1998;82:1236–41. 30 31 429 49. Turano KA, Geruschat DR, Baker FH. Fixation behavior while walking: persons with 32 http://bmjopen.bmj.com/ 33 central visual field loss. Vision Res. 2002;42:2635–2644. 34 430 35 36 431 50. Skelton DA, Becker C, Lamb SE, Close JCT, Zijlstra W, Yardley L, Todd CJ. 37 38 432 Prevention of Falls Network Europe: a Thematic Network aimed at introducing good 39 433 practice in effective falls prevention across Europe. Eur J Aging. 2004;1:89 94.
40 on September 26, 2021 by guest. Protected copyright. 41 42 434 51. Arditi A, Cagenello B. On the statistical reliability of letter chart visual acuity 43 44 435 measurement. Invest Ophthalmol Vis Sci. 1993;34:120–129. 45 46 436 52. Ferris FL, Kassoff A, Bresnick GH, Bailey IL. New visual acuity charts for clinical 47 48 437 research. Am J of Ophthalmol. 1982;94:91–96. 49 50 438 53. Myint J, Latham K, Mann D, Gomersall P, Wilkins A, Allen P. The relationship 51 52 439 between visual function and performance in rifle shooting for athletes with vision 53 54 440 impairment. BMJ Open Sport Exerc Med. 2016;2(1). 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 26 of 34
1 2 3 441 54. Pelli DG, Robson JG, Wilkins AJ. The design of a new letter chart for measuring 4 5 442 contrast sensitivity. Clin Vis Sci. 1988;2:187 199. 6 7 443 55. Elliott DB, Bullimore MA, Bailey IL. Improving the reliability of the Pelli Robson 8 9 444 contrast sensitivity test. Clin Vis Sci. 1991;6(6):471 475. 10 11 445 56. Octopus, Haag Streit International, AG, Bern, Switzerland 12 13 446 57. HFA, Carl Zeiss Meditec, Inc., Dublin, CA. 14 15 447 58. AspinallFor PA, Johnson peer ZK, Azuara Blanco review A, Montarzino only A, Brice R, Vickers A. 16 17 448 Evaluation of quality of life and priorities of patients with glaucoma. Invest 18 19 449 Ophthalmol Vis Sci. 2008;49(5):1907 1915. 20 21 450 59. Saunders LJ, Russell RA, Crabb DP. Practical landmarks for visual field disability in 22 23 451 glaucoma. Br J Ophthalmol. 2012;96:1185–1189. 24 25 452 60. Chisholm CM, Rauscher FG, Crabb DC, et al. Assessing Visual Fields for Driving in 26 27 453 Patients With Paracentral Scotomata. Br J Ophthalmol. 2008;92(2):225 230. 28 29 454 61. Rasch G. Probabalistic models for some intelligence and attainment tests. Chicago: 30 31 455 Mesa Press. 1993. 32 http://bmjopen.bmj.com/ 33 62. Norquist, J., Fitzpatrick, R., Dawson, J. and Jenkinson, C. (2004). Comparing 34 456 35 36 457 Alternative Rasch Based Methods vs Raw Scores in Measuring Change in Health. 37 38 458 Medical Care, 42, p.25. 39 459 63. Khadka J, McAlinden C, Pesudovs K. Quality Assessment of Ophthalmic
40 on September 26, 2021 by guest. Protected copyright. 41 42 460 Questionnaires: Review and Recommendations. Optom Vis Sci. 2013;8:720–744. 43 44 461 64. Stelmack J, Massof R. (2007). Using the VA LV VFQ 48 and LV VFQ 20 in Low 45 46 462 Vision Rehabilitation. Optom Vis Sci, 84(8), pp.705 709. 47 48 463 65. Reise S, Henson J. (2003). A Discussion of Modern Versus Traditional 49 50 464 Psychometrics As Applied to Personality Assessment Scales. J Pers Assess, 81(2), 51 52 465 pp.93 103. 53 54 466 66. Linacre JM. Reliability and separation of measures. Retrieved 30/06/2015, from 55 56 467 http://www.winsteps.com/winman/reliability.htm. 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 27 of 34 BMJ Open
1 2 3 468 67. Gothwal VK, Wright TA, Lamoureux EL, and Pesudovs K. Rasch Analysis of Visual 4 5 469 Function and Quality of Life Questionnaires. Optom Vis Sci. 2009;86(10):1160 1168. 6 7 470 68. Marella M, Gothwal VK, Pesudovs K, Lamoureux EL. Validation of the Visual 8 9 471 Disability Questionnaire (VDQ) in India. Optom Vis Sci. 2009;86:826 835. 10 11 472 69. Wright B, Linacre JM. Reasonable mean square fit values. Retrieved 29/02/2016, 12 13 473 from www.rasch.org/rmt/rmt83b.htm. 14 15 474 70. DeLongFor ER, DeLong peer DM, Clarke Pearson review DL. Comparing only the areas under two or 16 17 475 more correlated receiver operating characteristic curves: a nonparametric approach. 18 19 476 Biometrics. 1988;44(3):837 45. 20 21 477 71. Linacre JM. Dimensionality investigation an example. Retrieved 12/11/2014, from 22 23 478 http://www.winsteps.com/winman/multidimensionality.htm. 24 25 479 72. Fenwick EK, O'Hare F, Deverell L, Ayton LN, Luu CD, McSweeney S, Bentley SA, 26 27 480 Guymer RH, Finger RP. Rasch Analysis of the Independent Mobility Questionnaire. 28 29 481 Optom Vis Sci. 2016;93(2):181 187. 30 31 482 73. van Nispen RMA, Hoeijmakers JGJ, de Boer MR, et al. Agreement Between Self 32 http://bmjopen.bmj.com/ 33 Reported Co morbidity of Visually Impaired Older Patients and Reports from their 34 483 35 36 484 General Practitioners. Visual Impair Res. 2008;10(2 3):49 56. 37 38 485 74. Haymes S, Guest D, Heyes A, Johnston A. Mobility of people with retinitis 39 486 pigmentosa as a function of vision and psychological variables. Optom Vis Sci.
40 on September 26, 2021 by guest. Protected copyright. 41 42 487 1996;73(10):621 637. 43 44 488 75. Safran AB, Glaser JS. Statokinetic Dissociation in Lesions of the Anterior Visual 45 46 489 Pathways: A Reappraisal of the Riddoch Phenomenon. Arch Ophthalmol. 47 48 490 1980;98:291 295. 49 50 491 76. Yanagisawa M, Kato, S, Kunimatsu S, Tamura M, Ochiai M. Relationship between 51 52 492 vision related quality of life in Japanese patients and methods for evaluating visual 53 54 493 field. Jpn J Ophthalmol, 2011;55(2):132 137. 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 28 of 34
1 2 3 494 77. Lynn J, Swanson W, Fellman R. Evaluation of automated kinetic perimetry (AKP) 4 5 495 with the Humphrey Field Analyzer. Perimetry Update. 1990;1991:433 452. 6 7 496 78. Moenter VM, Crabb DP, Artes PH. Reclaiming the Periphery: Automated Kinetic 8 9 497 Perimetry for Measuring Peripheral Visual Fields in Patients with Glaucoma. Invest 10 11 498 Ophthal Vis Sci. 2017;58(2):868 875. 12 13 499 79. Black AA, Wood JM, Lovie Kitchin JE. Inferior field loss increases rate of falls in older 14 15 500 adultsFor with glaucoma. peer Optom Vis review Sci. 2011;88(11):1275 82. only 16 17 501 80. Friedman SM, Munoz B, West SK, et al. Falls and fear of falling: which comes first? A 18 19 502 longitudinal prediction model suggests strategies for primary and secondary 20 21 503 prevention. J Am Geriatr Soc. 2002;50:1329 35. 22 23 504 81. Glynn RJ, Seddon JM, Krug JH, Jr, et al. Falls in elderly patients with glaucoma. Arch 24 25 505 Ophthalmol. 1991;109:205–10. 26 27 506 82. Cummings SR, Nevitt MC, Kidd S. Forgetting falls. The limited accuracy of recall of 28 29 507 falls in the elderly. J Am Geriatr Soc. 1988;36:613–16. 30 31
32 http://bmjopen.bmj.com/ 508 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 28 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 29 of 34 BMJ Open
1 2 3 509 4 5 510 Figure 1. Graphical representation of areas under ROC curves for the different visual field 6 511 tests. The questions to which the numbers on the x axis refer are given in Table 1. * 7 512 indicates a visual field assessment with a statistically significant greater area under the curve 8 513 (AUC) than an alternative assessment (p≤0.05). All other comparisons were non significant. 9 10 514 Figure 2. Test durations of each of the five visual field assessments. 11 12 13 14 515 Figure 3. Participants’ ranking of visual field tests acceptability. 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 29 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 30 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 Graphical representation of areas under ROC curves for the different visual field tests. The questions to 27 which the numbers on the x axis refer are given in Table 1. * indicates a visual field assessment with a 28 statistically significant greater area under the curve (AUC) than an alternative assessment (p≤0.05). All 29 other comparisons were non significant. 30 31 542x305mm (300 x 300 DPI)
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 31 of 34 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 Test durations of each of the five visual field assessments. 36 37 165x137mm (300 x 300 DPI) 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from BMJ Open Page 32 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 Participants’ ranking of visual field tests acceptability. 26 27 579x304mm (300 x 300 DPI) 28 29 30 31
32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39
40 on September 26, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from 7-10 7-10 NA NA 1 1
11 11 10-11 10-11 11 11 3 3 6 6 6 6
6 6 5 5
10-11 10-11 7-10 7-10 10-11 10-11 Reported on page # Reported on page 2 2 6 6
cross-sectional studies cross-sectional
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
) Indicate ) study’sthe design commonlywith useda in term ortitlethe abstract the ) ) theGive eligibility criteria, and sourcesthe methodsand of selection of participants ) ) statisticalDescribe all methods, thoseincluding confoundingusedforcontrol to ) ) Describe any methods used subgroups examine to interactionsand ) If ) applicable, describe methodsanalytical taking account of sampling strategy ) ) in theProvide abstract informativean and balanced summary of what was donewhat wasand found ) Describe) sensitivity any analyses
For peer review only ) howExplain missing data were addressed a a a b e d c b For eachof interest,For variable sourcesdatagive of and details of methods of assessment (measurement). Describe comparability of of comparability assessment methods isthanthere more if one group ( ( why why
applicable collection collection ( ( ( ( Recommendation STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in be included should that items of Statement—Checklist (v4) 2007 STROBE
6 1 ( 3 specific objectives, State including prespecified any hypotheses 5 the Describe setting, locations, relevantand dates, including periods exposure,of recruitment, follow-up, data and 9 Describe any efforts to address sourcespotential of bias 7 define Clearly outcomes,all exposures,predictors, potential confounders,effect and modifiers. diagnostic Give criteria, if 4 key Present of elements study inthedesign early paper 8* 10 10 how Explain studythe size was arrived at # Item Item Results
Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 thebackground Explain scientific for rationale investigation the and being reported Introduction Title andTitle abstract Section/Topic
Page 33 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-018831 on 20 November 2017. Downloaded from Page 34 of 17-19 17-19 12 12 NA
17-19 17-19 3 3 17-19 17-19
12 12 NA NA 11 11 NA NA 4 4 14 14 NA
http://bmjopen.bmj.com/ BMJ Open on September 26, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report) boundaries category when continuous variables were categorized ) ) unadjustedGive estimates if applicable, and, confounder-adjusted estimates their and precision95% (eg, confidence
For peer )relevant, If consider translating estimatesrelativerisk of into absolute for risk meaningfulperioda time review only c b a similar studies,similar and other relevant evidence confounders magnitude of magnitude potential any bias confirmed eligible,confirmed included in study, the completing follow-up, analysed and interval). Make interval). clear confounderswhich were adjusted they why and for were included which the presentthe which is based article
( ( number Indicate (b) of participants missingwith eachfor data of interest variable Consider (c) diagram use flow of a reasons Give (b) non-participationfor each at stage 18 18 Summarise results key to reference withstudy objectives 16 16 ( 22 22 sourcethe Give of funding and role the of fundersthe thefor studypresent if and, applicable, originalthe for study on 19 19 Discuss of study, the limitations taking into sourcesaccount biasof potential imprecision.or Discussboth direction and 13* 13* Report (a) numbers of individuals each at stage of study—eg numbers eligible, potentially examinedfor eligibility, An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/,at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (externalgeneralisabilitythe of validity) studythe results Interpretation 20 cautious Give interpretationa of overall resultsconsidering limitations, objectives, multiplicity of resultsanalyses, from Limitations Key Key results Discussion Other analyses 17 other Report analyses done—eg analyses of subgroups sensitivity interactions, and and analyses
Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures
Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and
Participants 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60