Contrast Sensitivity and Vision-Related Quality of Life Assessment in the Pediatric Low Vision Setting

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Gregory Robert Hopkins, II

Graduate Program in Vision Science

The Ohio State University

2014

Master's Examination Committee:

Angela M. Brown, PhD, Adviser

Roanne E. Flom, OD

Thomas W. Raasch, OD, PhD

Gregory Robert Hopkins, II

2014

Abstract

A new test of contrast sensitivity (CS), the Stripe Card Contrast

Sensitivity (SCCS) test, could serve as a simple and efficient means for estimating the maximum contrast sensitivity value of a given patient without having to use multiple spatial frequency gratings, and without knowing the spatial frequency at which maximum sensitivity occurs. This test could be useful for a wide range of patients with various levels of visual acuity (VA), ages, and diagnoses.

We measured VA [Bailey-Lovie (BL), Teller Acuity Cards (TAC)] and CS [Pelli-

Robson (PR), SCCS, Berkeley Discs (BD)] in counterbalanced order with subjects at the

Ohio State School for the Blind (OSSB). Thus, we tested VA and CS using letter charts

(B-L, P-R), grating cards (TAC, SCCS) and a chart with shapes (BD).

Vision-related quality of life (QoL) surveys [The Impact of Visual Impairment in

Children (IVI_C) and Low Vision Prasad Functional Vision Questionnaire (LVP-FVQ)] were used following vision testing. Additionally, we obtained Michigan Orientation &

Mobility (O&M) Severity Rating Scale (OMSRS) severity of need scores for some participants.

Testing was performed over a two-year period for 51 participants at OSSB. We have organized our work into three experiments: Experiment I was performed in the

2012-13 school year and included 27 participants who were tested monocularly using the

ii patient’s preferred eye. The following year, we returned for repeat testing of 11 participants from the first year (“Experiment IIa”) and additional testing of 24 new participants (“Experiment IIb”). Those assessments were performed on each eye monocularly (where possible) rather than just with the preferred eye. QoL and O&M results were obtained during both years of testing and are detailed in Experiment III.

Vision tests on the better eyes correlated positively and significantly with one another, except for a non-significant correlation between the B-L and SCCS. The IVI_C correlated significantly with all vision tests, except B-L acuity, with better visual function always correlating with higher quality of life. The LVP-FVQ correlated significantly with all metrics employed. The OMSRS scores did not correlate significantly with any of our metrics, except the LVP-FVQ, probably because so few subjects provided data for the

OMSRS.

Both of the grating tests (SCCS and TAC) and the BD indicated better visual performance than the corresponding letter acuity and contrast charts for subjects with reduced vision. For measuring contrast sensitivity in those with reduced vision, the simpler task and bolder patterns of the SCCS and BD may make them more likely to reveal the maximum performance that a given patient can achieve.

iii

Dedication

This document is dedicated to Katya, my wife, and our two daughters: Adelaide and Matilda.

Acknowledgements

Angela Brown has been a brilliant and gentle mentor to me throughout this process and I have been fortunate to have had the opportunity to develop a deeper understanding of vision science as a result of her attention and support.

I am truly fortunate to join a lineage of recognized field leaders by training with

Dr. Roanne Flom. It has been a privilege to have the opportunity to discuss low vision history, practice, and research with Dr. Thomas Raasch.

I must thank the teachers and staff at The Ohio State School for the Blind, particularly Nurse Judith Babka, Principals Marcom and Miller, and orientation & mobility instructors Phil Northup and Mary Swartwout.

I’d also like to acknowledge Ian L. Bailey, OD, DSc(hc), FCOptom, FAAO, professor at the University of California, Berkeley School of Optometry for providing the spark from which this work was lit.

Finally, I’d like to acknowledge the substantial contributions Bradley E.

Dougherty, OD, PhD has made towards the analysis of the patient-reported outcome and quality of life measures in my study. I would also like to thank him for the overall role he has played in development of my career from a third year optometry student up through post-graduate advanced practice fellowship work.

Vita

June 2002 ...... Moeller High School

2006...... Biology, The Ohio State University

2010...... O.D., The Ohio State University

2012 to present ...... Advanced Practice Fellow in Low Vision Rehabilitation, College of Optometry, The Ohio State University

Publications

Hopkins, G.R., & Flom, R.E. (2013, October). Disability Determination: More Within Our Means Now Than Ever. Poster presented at the annual meeting of the American Academy of Optometry, Seattle, WA.

Hopkins, G.R., & Brown, A.M. (2013, May). Contrast Sensitivity Measurement in the Pediatric Low Vision Setting. Poster presented at the annual meeting of Association for Research in Vision and Ophthalmology, Seattle, WA.

Fields of Study

Major Field: Vision Science

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgements ...... v

Vita ...... vi

List of Tables ...... xi

List of Figures ...... xii

List of Frequently Used Abbreviations ...... xv

Introduction ...... 1

Purpose ...... 1

Visual Acuity Measurement ...... 2

Significance of Acuity Measurement ...... 2

Development of Acuity Measurement ...... 2

Grating Acuity Measurement...... 7

Contrast Sensitivity Measurement ...... 10

Definition ...... 10

vii

Development of Contrast Sensitivity Testing ...... 11

Techniques for Contrast Sensitivity Measurement ...... 15

Significance of Contrast Sensitivity Measurement ...... 22

Vision-Related Quality of Life Assessment ...... 23

The IVI_C ...... 25

The LVP-FVQ ...... 26

Orientation and Mobility Assessment ...... 27

The Michigan Orientation and Mobility Severity Rating Scale ...... 28

Experiment Overview...... 29

Ethics...... 31

Recruitment ...... 31

Participant Characteristics ...... 33

Objectives ...... 37

Experiment I...... 38

Study Design ...... 38

Study Methods ...... 38

Letter Acuity Procedure ...... 39

Grating Acuity Procedure ...... 39

Letter Contrast Procedure ...... 40

viii

Stripe Card Contrast Sensitivity Test ...... 41

The Berkeley Discs of Contrast Sensitivity ...... 42

Results for Experiment I ...... 42

Discussion for Experiment I ...... 58

Experiment II – Separate Eye Testing ...... 60

Introduction to Experiment II ...... 60

Methods for Experiment II...... 60

Results for Experiment IIa: Repeat Testing ...... 61

Repeatability between Experiments I and IIa ...... 71

Results for Experiment IIb: New Subjects...... 73

Discussion for Experiment II ...... 83

Experiment III – Quality of Life and Orientation and Mobility ...... 89

Vision-Related Quality of Life ...... 89

Orientation and Mobility...... 100

Discussion for Experiment III ...... 108

Vision-Related Quality of Life ...... 108

Orientation and Mobility...... 114

General Discussion ...... 116

Test Results ...... 116

Other Considerations ...... 117

Stated Objectives ...... 118

References ...... 122

Appendix A: Study Materials ...... 129

List of Tables

Table 1. Complete Participant List ...... 35

Table 2. Experiment I Participants...... 43

Table 3. Experiment I Summary Test Results ...... 44

Table 4. Experiment IIa Participants ...... 62

Table 5. Experiment IIa Summary Test Results ...... 63

Table 6. Experiment IIa Repeatability Statistics ...... 73

Table 7. Experiment IIb Participants ...... 74

Table 8. Experiment IIb Summary Results ...... 75

Table 9. Experiment I & II Better Eye Only ...... 84

Table 10. Experiment II Summary Results ...... 85

Table 11: Test Chart Correlations ...... 88

Table 12. QoL and O&M Correlations ...... 107

List of Figures

Figure 1: The original Snellen and Sloan Charts ...... 4

Figure 2. Bailey-Lovie Chart ...... 6

Figure 3. ETDRS LogMAR Chart ...... 7

Figure 4. Teller Acuity Cards ...... 10

Figure 5. Campbell-Robson CSF Chart ...... 13

Figure 6. Pelli-Robson Chart ...... 18

Figure 7. The Stripe Card Contrast Sensitivity ...... 21

Figure 8. The Berkeley Discs of Contrast Sensitivity ...... 22

Figure 9. All OSSB Student Visual Acuities by Chart Report ...... 33

Figure 10. Participant Diagnoses ...... 34

Figure 11. Experiment I Summary Plot Statistics ...... 45

Figure 12. Experiment I B-L vs Chart Report ...... 47

Figure 13. Experiment I Lettered Chart Results by Diagnosis ...... 48

Figure 14. Experiment I Striped Chart Results by Diagnosis ...... 49

Figure 15. Experiment I Shaped Chart Results by Diagnosis ...... 50

Figure 16. : Experiment I Acuity Results ...... 51

Figure 17. Experiment I P-R & SCCS Results ...... 53

Figure 18. Experiment I P-R & SCCS Bins ...... 54

xii

Figure 19. Experiment I P-R & BD Results ...... 55

Figure 20. Experiment I P-R & BD Bins ...... 56

Figure 21. Experiment I SCCS & BD Results ...... 57

Figure 22. Experiment I SCCS & BD Bins ...... 58

Figure 23. Experiment IIa Summary Plot Statistics...... 64

Figure 24. Experiment IIa Acuity Results ...... 65

Figure 25. Experiment IIa P-R & SCCS Results ...... 66

Figure 26. Experiment IIa P-R & SCCS Bins...... 67

Figure 27. Experiment IIa SCCS & BD Results ...... 68

Figure 28. Experiment IIa P-R & BD Bins ...... 69

Figure 29. Experiment IIa SCCS & BD Results ...... 70

Figure 30. Experiment IIa SCCS & BD Bins ...... 71

Figure 31. Experiment IIa Lettered Chart Test-Retest ...... 72

Figure 32. Experiment IIa Striped Chart Test-Retest ...... 72

Figure 33. Experiment IIa Shaped Chart Test-Retest ...... 73

Figure 34. Experiment IIb Summary Plot Statistics ...... 76

Figure 35. Experiment IIb Acuity Results ...... 77

Figure 36. Experiment IIb P-R & SCCS Results ...... 78

Figure 37. Experiment IIb P-R & SCCS Bins ...... 79

Figure 38. Experiment IIb P-R & BD Results ...... 80

Figure 39. Experiment IIb P-R & BD Bins...... 81

Figure 40. Experiment IIb SCCS & BD Results...... 82

xiii

Figure 41. Experiment IIb SCCS & BD Bins ...... 83

Figure 42. Experiment II Summary Plot Statistics ...... 86

Figure 43. IVI_C v B-L Regression ...... 91

Figure 44. IVI_C vs. P-R Regression ...... 92

Figure 45. IVI_C vs. TAC Regression...... 93

Figure 46. IVI_C vs. SCCS Regression ...... 94

Figure 47. IVI_C vs. BD Regression ...... 95

Figure 48. LVP-FVQ vs. B-L Regression ...... 96

Figure 49. LVP-FVQ vs. P-R Regression ...... 97

Figure 50. LVP-FVQ vs. TAC Regression ...... 98

Figure 51. LVP-FVQ vs. SCCS Regression ...... 99

Figure 52. LVP-FVQ vs. BD Regression ...... 100

Figure 53. O&M vs. B-L Regression ...... 101

Figure 54. O&M vs. P-R Regression ...... 102

Figure 55. O&M vs. TAC Regression ...... 103

Figure 56. O&M vs. SCCS Regression ...... 104

Figure 57. O&M vs. BD Regression ...... 104

Figure 58. Average person scores for the IVI_C (left) and the LVP-FVQ (right) ...... 110

Figure 59. Subject-Item map for the IVI_C ...... 112

Figure 60. Subject-Item map for the LVP-FVQ...... 113

Figure 61. Person Score Linear Regression of LVP-FVQ vs. IVI_C ...... 114

Figure 62. Cutoffs for normal SCCS Performance ...... 120

xiv

List of Frequently Used Abbreviations

B-L Bailey-Lovie Chart BD Berkeley Discs of Contrast Sensitivity c/deg Cycles per degree CSF Contrast Sensitivity Function HM Hand Motion Only IVI_C Impact of Visual Impairment in Children Survey LogCS Logarithm of the contrast sensitivity LogMAR Logarithm of the minimum angle of resolution LP Light Perception LVP-FVQ Low Vision Prasad Functional Vision Questionnaire MAR Minimum Angle of Resolution NLP No Light Perception O&M Orientation and Mobility OMSRS The Michigan Orientation and Mobility Severity Rating Scale OSSB The Ohio State School for the Blind P-R Pelli-Robson Chart QoL Quality of Life (related to vision) RL Right and Left eye tested individually SCCS Stripe Card Contrast Sensitivity TAC Teller Acuity Cards

Introduction

Purpose

Measuring the functional ability of patients with ocular disorders is a classic problem in clinical vision science. While objective assessments are certainly important, functional visual assessments often yield the most appropriate management strategies for patients with reduced vision. The intention of this research is to develop and validate examination methodologies that best represent the functional visual abilities of persons with reduced vision. This is important because, even when one’s underlying disorder cannot be treated, care can still be given to maximize one’s success in life. Said another way, “Even though it may be true that nothing more can be done for the eye, it is almost never true that nothing more can be done for the patient” (Tandon, 1994). Traditional methods of visual assessment include visual acuity, contrast sensitivity, color vision, and visual field testing, among others. Eye care practitioners’ most well recognized assessment methodology has customarily been visual acuity. However, it has been known from the early testing days that acuity values do not represent a complete picture of a given patient’s ocular health let alone his or her visual functioning. My purpose is to assist in the development of a reliable and easy-to-use test of pediatric contrast sensitivity. My hope is that by making this test available, we will encourage eye care practitioners to consider including contrast sensitivity measurement as one component of the testing that they perform when examining children with visual impairment.

Visual Acuity Measurement

Significance of Acuity Measurement. Visual acuity measurement has become an integral part of eye care ever since the mid 1800s, when it was first introduced. Visual acuity is usually the first testing procedure performed during any ocular examination.

Generally, visual acuity is a measure of the spatial resolving ability of the human eye, combined with the visual system’s ability to process images as distinct based upon the angle that these objects subtend upon the eye. The assessment of visual acuity is useful in many ways. Some examples include, but are not limited to: monitoring the refractive status, health and stability of a given patient’s eyes, determining a patient’s legal blindness status, determining his or her ability to qualify for driving privileges, and candidacy for cataract extraction (or other medical workup).

Development of Acuity Measurement. In the mid 1800s, early medical practitioners developed many different methods of acuity measurement in their efforts to standardize the task (Bennett, 1965). Early normative test results suggested that most observers have a minimum angle of resolution (MAR) that is only slightly smaller than one arc minute. Accordingly, practitioners designed eye charts so that the smallest appreciable detail element would subtend less than one minute of arc from a practicable test distance so that threshold measurements could be obtained. As will be discussed further in the next section of this thesis, a ratio between the MAR and the overall size of the optotype containing this detail element has conventionally been a 1 to 5 ratio.

Considering the smallest letter that a patient can identify, the MAR for his or her visual acuity is given as the reciprocal of the ratio between the height of that barely identifiable

2 letter and a letter corresponding to one arc minute. This proportion is variously expressed by its logarithm (logMAR) or, with numerator and denominator multiplied by 20 feet or 6 meters (as a “Snellen fraction”).

Detection, Localization, Resolution, Recognition or Identification Acuity. There are various ways to go about discovering whether an observer can perceive fine detail.

One approach is to determine the smallest object that a given observer can detect against a uniform background (detection task). A variation on this method is to force an observer to choose whether or not a small object is present in one defined area versus a blank space (localization task). Two alternative forced-choice experiments are often designed with either two spatial or two temporal intervals, with the stimulus being presented in one of those intervals. Another method is to use closely spaced lines or dots as visual stimuli, and then describe the test distance at which they can be resolved as spatially distinct

(resolution task). If, instead, the acuity task is to describe orientation of an object, then visual acuity can be measured using a wide variety of stimuli such as shapes, gratings or letters (recognition task). Alternatively, one could present an array of identifiable symbols such as shapes, numbers or letters (identification task) (Kramer & Mcdonald,

1986; Owsley, 2003).

The identification acuity approach lends itself very well to clinical practice. A

Dutch ophthalmologist named Hermann Snellen designed the original eye chart shown in

Figure 1 and is responsible for the popularization of unrelated letter identification as the primary method of visual acuity measurement. The use of high-contrast capital letters as optotypes was Snellen’s key innovation. It allowed for the rapid proliferation of visual

3 acuity measurement as a technique used by medical practitioners. For a patient stated to have exactly “20/20” visual acuity using Snellen’s notation system, the smallest letter identifiable by this patient will subtend 5 minutes of arc. The minimum angle of resolution of a letter at threshold size is assumed to have a 1:5 ratio of the letter height, so their MAR is one arc minute. Eye charts with various letter fonts were developed over the years and many of them assumed the MAR to be 1/5th the letter height—even if the stroke widths of those letters were not uniform. Subsequently, Louise Sloan developed a letter series with roughly equal legibility and a 5 x 5 stroke width aspect ratio to solve the problem of inconsistent typefaces. Sloan also advocated for equally spaced optotype arrays that scaled geometrically (Sloan, 1959).

Figure 1: The original Snellen and Sloan Charts Left, the original Snellen chart (from www.Precision-Vision.com); Right, Sloan’s distance acuity charts (from Sloan, 1959)

Bailey-Lovie Acuity Chart. Ian Bailey and Jan Lovie developed the Bailey-Lovie

Acuity Chart to better assess the participants in their studies of Australian visual acuity and contrast sensitivity in the 1970s. They designed their acuity charts following a letter size progression based on steps equal to 0.10 LogMAR. This design feature allows for the chart to be used with uniformity at any practicable test distance in order to best capture a given observer’s threshold acuity value. Many of the design concepts they introduced were incorporated in to subsequent acuity chart designs such as: approximately equally legible optotypes, an equal number of letters on every row, uniform letter and row spacing, a logarithmic size progression that covers a wide range of human vision, and the use letter-by-letter LogMAR scoring (Bailey & Lovie, 1976).

The chart we used in the present study (see Figure 2) has seventy 5x4 sans serif

British standard letters total, with five letters per row. The LogMAR scores along the right hand column of the chart are based on a standard testing distance of 6 meters, but the actual letter sizes (in M units) were included to facilitate testing at any practical distance. The original Bailey-Lovie chart covered a wide range of letter sizes from 3.2 M to 63M. Acuity measurements from 6/3 (20/10) through 1/60 (20/12,000) could be reasonably made with appropriate adjustments in testing distance. The chart we used is also designed to cover more than a ten-fold range of acuity: from 20/12 through 20/250 when viewed from 10 feet away.

Figure 2. Bailey-Lovie Chart (from Tasman, 1992)

Later Developments in Letter Acuity Measurement. The investigators in the

Early Treatment of Diabetic Retinopathy Study (ETDRS) created a backlit light box chart to ensure equal illumination of the 5x5 letters from the Sloan series (Ferris, Kassoff,

Bresnick, & Bailey, 1982). Their chart, shown in Figure 3 below, is the current “gold standard” for clinical research. Future developments in visual acuity testing will likely include a move towards computerized methodologies, which will enable examiners to test with even greater uniformity, precision and efficiency.

Practitioners who select well-designed eye charts benefit in many ways— including the use of efficient scoring notation. Conversion of Snellen “20/__” or other notation to LogMAR values allows for statistical analysis of acuity data sets. Acuity scoring can be performed in a number of ways, but letter-by-letter scoring has been shown to be the most repeatable (Raasch, Bailey, & Bullimore, 1998).

Figure 3. ETDRS LogMAR Chart (from www.Precision-Vision.com)

Grating Acuity Measurement. Measurement of visual acuity in the classical way, i.e., via the identification of carefully selected and arranged letters, requires an observer who is cognitively capable of reading and reporting letters. Other methods have been developed to measure visual acuity when a patient cannot read numbers or letters, or even identify a shape or report its orientation. Approaches such as the observation of optokinetic nystagmus, preferential looking and visual evoked potential are some examples. These techniques often include stripes or other patterns as a measure of resolution acuity or cortical response. Examiners have known since at least the 1960s that the eyes of visual observers are preferentially drawn towards visible patterns versus blank homogenous areas (Frantz, Ordy, & Udelf, 1962). To provide a measure of resolution acuity, examiners developed visual stimuli using striped gratings.

Teller Acuity Cards. Davida Teller developed this test with her collaborators in the mid-1980s for the purpose of rapid estimation of acuity in infants (McDonald et al.,

1985). Prior to this time, testing was done either in a very regimented and time-

7 consuming laboratory setting, or very informally by clinicians using “fix and follow” penlight testing techniques.

The Teller Acuity test consists of a deck of sixteen 25.5 x 55.5 cm gray cards, each with a 4 mm round peephole in the middle and a 12 x 12 cm region of vertical black and-white grating centered on one half. The standard testing distance is 55.5 cm, the length of one card. The spatial frequency progression of the Teller Cards starts at about the minimum angle of resolution for a 20/20 optotype. That is, one half of a black/white grating cycle (one black or white stripe) covers one minute of arc from a given distance.

This pattern is printed onto the first card (see Figure 4). A single stripe is always present on the edge of each pattern box to minimize potential edge-detection artifacts. The spatial frequencies shown on subsequent cards are scaled up systematically until many minutes of arc fit within one black and white grating cycle (Dobson & Teller, 1978). Specifically, the spatial frequency of the gratings increases by a factor of two (one octave, i.e., one base-2 log unit) after every other card. Thus, the spatial frequency of each card is higher than its predecessor by a factor of √2. Additionally, one of the cards is left blank to serve as a control, and another has wide stripes of 0.23 cycles per cm (0.236 cy/deg, or

20/2,540 at 55.5 cm) covering an entire half to serve as a “low vision card.” By presenting each of these two cards at least once, the examiner can attune herself or himself to the particular strong “looking behavior” versus “non-looking behavior” of the subject under examination

The “grating acuity” of the participant can be estimated by observing the looking behavior of the participant or by simply asking a capable observer to point. When a card

8 is presented containing stripes so fine as to no longer be resolvable, the participant would either display non-looking behavior or report that they cannot ascertain whether the stripes are present on the right or left hand side. If an examiner shows cards in octave steps until the observer fails to find the grating, the next logical card to display would be the one whose stripes are ½ an octave wider. The cards progress by doubling the size of a grating cycle from initially spanning only 2 minutes of arc all the way up to covering 128 minutes of arc. This final card approximates a visual acuity of 0.32 cycles per cm

(20/1,280 at 55.5 cm). This size progression, combined with preferential looking techniques that were refined in the mid-1980s, allow the examiner to obtain spatial visual information with good efficiency and validity for less-sophisticated patients (McDonald et al., 1985).

Even though measurements obtained with Teller cards can be converted into

Snellen notation, “grating acuity” measurements made by this technique are not directly comparable to optotype identification acuity tasks. Functional visual estimations for a given task are best performed with measurement styles that most closely relate to the task. For example, a prudent examiner interested in reading ability would use word or sentence acuity cards. Conversely, word reading acuity would not be expected to describe a patient’s orientation and mobility skills. The particular utility of grating acuity may be that, once baseline measurements have been obtained, they allow for tracking the visual development for less-sophisticated patients over time.

Figure 4. Teller Acuity Cards (from www.Precision-Vision.com)

Contrast Sensitivity Measurement

Definition. In the context of monochromatic luminance assessments (i.e., not considering equiluminant color contrast), the term “contrast” is meant to quantify how the luminance of a given point makes it discrete from the average luminance of the adjacent area. Spatial contrast sensitivity can be broadly defined as one’s ability to perceive luminance variations across visual space. Contrast itself is expressed as a percentage and can be taken from a visual scene in three main ways: as the magnitude of luminance variation between points in a visual scene and the average luminance of that scene (root-mean-square contrast), as the difference between the brightest and darkest parts of a repeating pattern divided by the sum of those extreme luminance values

(Michelson contrast), or as the brightness increment between a small object and its background divided by the average background luminance (Weber contrast).

Calculation of contrast via the Michelson formula lends itself particularly well to quantifying the contrast of periodic stimuli (like gratings). In the case of stripes with a

50:50 duty cycle, Michelson and Weber contrast values are mathematically equivalent,

퐼 −퐼 Δ퐼 and are written formulaically as 퐶 = 푚푎푥 푚푖푛 and 퐶 = , respectively. Restating 퐼푚푎푥+퐼푚푖푛 퐼푎푣푒

퐼 −퐼 퐼 +퐼 Weber Contrast components as: ∆퐼 = 푚푎푥 푚푖푛 and 퐼 = 푚푎푥 푚푖푛 allows for the 2 푎푣푒 2

Δ퐼 퐼 −퐼 following rearrangement: = 퐶 = 푚푎푥 푚푖푛. 퐼푎푣푒 퐼푚푎푥+퐼푚푖푛

Calculation of contrast for small isolated symbols on a large background is generally done using Weber’s approach, but this situation is found less-often in real world settings.

A pattern or scene where all the dark elements emit zero luminance has bright elements at 100% contrast, and visual stimuli with fainter shades of gray represent different levels of contrast up to the point where no pattern exists and the scene is entirely uniform (0% contrast). Clinically, contrast sensitivity values are given in log10 units as an expression of the reciprocal of the contrast threshold, the lowest percentage contrast that a patient is able to perceive.

Development of Contrast Sensitivity Testing. The analysis of spatial vision

(i.e., the perception of borders, lines and edges) beyond the single assessment of high contrast visual acuity has been a topic of close investigation since the 1960s. To perform this analysis, scientists first needed to understand the optical properties of the human eye.

The modulation transfer function is one method for quantifying the clarity of an optical system. Physicists in the 1960s were generating sine waves with cathode ray tubes to measure optical modulation transfer functions for cameras and televisions. Sine waves make for a convenient testing target because a sinusoidal input will always result in a

11 sinusoidal output through a linear optical system. Physiological optics researchers modified this modulation transfer approach to measure the modulation transfer function of the human visual system. Once the optical transfer component of the eye was generally established, the next step was to determine the neural component of visual processing.

Psychophysicists applied cathode ray display technology to generate sine wave gratings for contrast detection measurements at different spatial frequencies. It was then possible to obtain s-shaped psychometric functions, where the probability of sine wave detection was shown as a function of the contrast value. For example, the threshold value for contrast detection at a given spatial frequency may be derived from the psychometric curve. The threshold is the amount of contrast underlying the point on the curve with the steepest slope. Performing this procedure over various spatial frequencies allows for the threshold results to be combined so that contrast sensitivity, plotted as a function of spatial frequency, forms the contrast sensitivity function (CSF). The CSF can be considered an envelope function made up of spatial frequency tuned channels coexisting within the visual system. The image in Figure 5 was developed by Campbell and Robson in order to serve as a demonstration of the CSF (Shapley & Lam, 1993). The image shows a sine wave of increasing spatial frequency from left to right and decreasing contrast from bottom to top. Any person observing the image should be able to appreciate the band-pass shape of his or her own contrast sensitivity function. Scientists, e.g.,

DeValois and DeValois (1988), have shown that the use of a linear model for human spatial vision is a practicable method. This approach allows investigators to use the CSF

12 to easily predict an observer’s ability to detect other objects or patterns based upon their size and contrast level.

The use of sine waves as visual stimuli can be traced as far back as the 1800s when the physicist Ernest Mach designed a spinning cylindrical apparatus with movable dark strips of paper to produce s-shaped curves along a harmonic progression known as a

Fourier series (Campbell, Howell, & Robson, 1970). Fourier’s work on heat flow in 1822 described how any periodic pattern could be broken down into composite sine waves. His theorem was readily applied to linear systems analysis in other areas such as acoustics.

However, aside from Mach’s early work, very few references to Fourier analysis being applied to measuring the visibility of grating stimuli can be found until the 1950s.

Figure 5. Campbell-Robson CSF Chart (from Izumi Ohzawa)

Campbell and Robson understood that a typical visual scene has very few pure sine waves to behold, so they wanted to know if Fourier analysis would allow examiners to predict an observer’s contrast threshold for more complex visual targets. Campbell and

Robson began by generalizing from sine waves to square, saw-tooth and rectangular wave gratings (Campbell & Robson, 1968). Each wave contains a fundamental frequency and higher harmonic frequencies. For example, a square wave contains a fundamental frequency (f) and combination of odd harmonic frequencies (f, 3f, 5f, 7f, etc.—much like the tone of a clarinet). The amplitude also decreases according to the following pattern for each harmonic: 4/π*((sin(x) + 1/3(sin(3x)) + 1/5(sin(5x)) + …). Continuing with the square wave as an example, Fourier theory predicts a 4/π higher sensitivity relative to a pure sine wave. This is because a square wave contains a higher amplitude fundamental than a sine wave by a factor of a 4/π. Campbell & Robson’s threshold response data supported this prediction, and they also demonstrated that the square wave was distinguishable from the sine wave just when the contrast was high enough for the harmonics to reach threshold. The following year, Blakemore and Campbell demonstrated that if an observer adapts to a square wave, the observer’s contrast sensitivity is relatively lowered along his/her CSF only at the frequencies of the square wave’s fundamental and third harmonic (Blakemore & Campbell, 1969a, 1969b).

Square waves are of particular interest because the detection of straight vertical or horizontal edges is important in daily visual functioning. Around the 1960s, examiners were beginning to understand that the neurology of the human visual cortex leaves us predisposed towards edge detection. Investigation of cat and primate visual cortex cells

14 by Hubel and Wiesel demonstrated the presence of center-surround receptive field organization in these animal models. This receptor array organization enhances edge detection by lateral inhibition (Hubel & Wiesel, 1962). An example of this edge-detection penchant comes from one of Ernest Mach’s early discoveries: the perceptual illusion of

“Mach Bands.” Mach designed a black and white mixing disc that had sectors, similar to

Masson’s spinning discs, but with an area containing a smooth curve of increasing black shading from the center outwards (Ratliff, 1965). He found that humans seemed to perceive edges at the smooth white-to-black gradient created by spinning the disc, as if the sectors had a sharper “step-wise” shading transition.

The concept of lateral inhibition between center-surround ganglion cell receptive fields is useful in describing the band-pass nature or “inverted U-shape” of the CSF.

Investigators struggled initially to explain this phenomenon because the gradual roll-off in contrast sensitivity at low spatial frequencies was not as readily anticipated as the high spatial frequency roll off (which results from optical restrictions and receptor spacing limitations within the human eye). Once technology had advanced to the point at which the general shape and underlying properties of the CSF were well known, practitioners began to apply this knowledge to clinical vision assessment with increasing success.

Techniques for Contrast Sensitivity Measurement. The maximum sensitivity found on the CSF curve for most observers requires contrast presentations of less than

1%. This fact makes designing and administering tests of contrast sensitivity a challenge.

A French scientist, Pierre Bouguer, made initial attempts at measuring the human contrast threshold in 1760. Bouguer designed an experiment in which a wooden rod cast a faint

15 shadow from a distant candle onto an illuminated white screen. The further away the candle was placed, the fainter the shadow, until it faded out of view (Pelli & Bex, 2013).

In 1845, another Frenchman, Antoine Masson, realized that it would be very difficult to accurately print low contrast targets so he designed black and white mixing discs that when spun would appear gray. Varying the size of the black sectors allowed for accurately calibrated contrast assessments. In 1918, George Young attempted to print a contrast sensitivity testing booklet using ink spots that were precisely diluted from page to page (Shapley & Lam, 1993).

All of these early tests were attempts at measurement via detection tasks, in which the observer is to report the faintest detectable stimulus. Such threshold techniques are useful, but like today’s threshold visual field tests, this approach can pose clinical reliability challenges when false positive and false negative responses occur. Localization

(is the target in one spot or another), recognition (which way is the stimulus oriented) and identification (what optotype was seen) tasks lend themselves much more readily to clinical assessments. This is one factor that may explain why the visual acuity and contrast sensitivity measurement tests most frequently used today are letter identification charts.

As stated above, renewed interest around contrast sensitivity testing arose with the development of cathode ray tubes in the 1960s. Finally, the technology existed to generate reliably calibrated stimuli. More clinical tests of contrast sensitivity were developed at that time than ever before. Early examples include: the Arden Plates, the

Cambridge Low Contrast Grating Test and the Pelli-Robson letter contrast sensitivity test

(Arden, 1978; Pelli, Robson, & Wilkins, 1988; Wilkins, Della Sala, Somazzi, & Nimmo-

Smith, 1988).

Pelli-Robson Contrast Chart. Denis Pelli and John Robson developed this chart in the 1980s in order to provide a simple and reliable clinical test of contrast sensitivity that could be adopted by practitioners in order to estimate the maximum contrast sensitivity of a patient (Pelli et al., 1988). Prior to the development of this chart, time- intensive measurement of the complete contrast sensitivity curve using computer- generated sine waves was the standard practice in laboratory studies. Pelli and Robson designed their 60 x 85 cm chart with forty-eight Sloan letters arranged in triads of decreasing contrast value (see Figure 6). The letter triads advance in steps of 0.15 log units from approximately 100% to 0.56% contrast (LogCS 0.00 to 2.25). All the ten letter options are contained within the first three rows and all letters are of equal in size throughout the chart—subtending 2.8 degrees (20/672) from the recommended 1 meter test difference. When viewed from 1 meter, these letters are well above the acuity threshold of most patients. If the chart were held 3 meters from a subject, the spatial frequency of the letters would fall within the range of a normal observer’s contrast sensitivity maximum. Testing at a closer distance ensures that results are not obtained for spatial frequencies that would correspond to the high spatial frequency roll-off region of an observer’s contrast sensitivity curve.

Figure 6. Pelli-Robson Chart (from www.Precision-Vision.com)

Further Developments of Contrast Sensitivity Testing. Following the development of the Pelli-Robson chart, others were designed such as: the Rabin letter contrast sensitivity test, the Vistech Chart, the Melbourne Edge Test, and the Mars letter contrast sensitivity test (Arditi, 2005; Eperjesi, Wolffsohn, Bowden, Napper, &

Rubinstein, 2004; Haymes et al., 2006; Rabin & Wicks, 1996; Reeves, Wood, & Hill,

1991; Wolffsohn, Eperjesi, & Napper, 2005).

For the pediatric population, there are the Hiding Heidi test and the Lea Contrast

Sensitivity booklet among others (Susan J Leat & Wegmann, 2004). A sample of novel tests currently under development include the Stripe Card Contrast Sensitivity, the

Berkeley Discs, the PL-CS Test, the Grating Contrast Sensitivity Test, the iPad Contrast

Sensitivity Test, and the qCSF (Bailey, Chu, Jackson, Minto, & Greer, 2011; Bittner,

Jeter, & Dagnelie, 2011; Dorr, Lesmes, Lu, & Bex, 2013; Kollbaum, 2014; Pokusa, Kran,

& Mayer, 2013).

Stripe Card Contrast Sensitivity Test. Angela Brown, Delwin Lindsey, and I are in the process of refining the design for this novel test of contrast sensitivity. We expect to fill the need for reliable contrast sensitivity testing of non-verbal or otherwise developmentally delayed patients with this test. The Stripe Card Contrast Sensitivity test

(SCCS) is similar in many respects to the Teller Acuity cards. The prototype version we used consists of a deck of 15 gray cards sized 55.5 x 25.5 cm with one side containing a

22 x 20 cm box of horizontal stripes that start 6 cm from the central peephole and extend to the edge of the card at a fixed spatial frequency of 1 cycle per 6.8 cm (0.15 c/deg or

20/4,000 grating acuity from 57 cm). Contrast values for the stripes decrease in 0.15 log10 unit steps from approximately 100% to 1% contrast. This progression of ½ octave steps is similar to the Pelli-Robson chart—i.e., contrast level differs by a factor of two for every- other letter triad on the Pelli-Robson card and for every other card on the SCCS test.

Figure 7 demonstrates two typical cards from the deck: one at full contrast and one at about 30% contrast. In the center of the card is a peephole through which the examiner can observe the patient’s looking behavior. The peephole is especially useful if the patient cannot point their fingers or speak. In this case, the examiner can show the cards to the patient with the stripes on one end, and then flipped to the other. The patient’s eyes should be drawn to one direction for the first presentation and then reliably to the opposite direction when presented with the opportunity for a second look. If the contrast is high enough so that the patient can see the stripes, then his or her eyes will first be

19 drawn towards the pattern when it is presented, and then again in the other way after the pattern has been flipped.

Similar to the Pelli-Robson test, the SCCS test does not set out to completely map the contrast sensitivity function of a given subject. Laboratory studies that obtain threshold values using sine waves of different spatial frequencies are the classical way to obtain the true peak of the contrast sensitivity function (CSF), e.g., Adams’ tests for infants and children (Adams & Courage, 2003). However, these methods are time intensive and currently impractical for clinical application (Lennie & Hemel, 2002).

Unlike the Pelli-Robson test, the SCCS does not attempt to test with optotypes sized at the assumed normal peak of the CSF, which lies between 3-5 c/deg. Rather, the SCCS takes advantage of the spatial harmonic properties of square waves, which activate cells of the human visual cortex even when the fundamental frequency is below that of the cell

(Blakemore & Campbell, 1969a, 1969b; Campbell et al., 1970; Campbell & Robson,

1968). In this way, the SCCS can test at 0.15 c/deg to be sure to avoid the steep higher spatial frequency roll-off portion of the CSF curve, which would otherwise cause the examiner to significantly underestimate the threshold maximum. When testing is performed with square waves at a spatial frequency that would ordinarily be in the gradual low spatial frequency roll-off section of the CSF, higher harmonics present in the stimulus prevent the drop in sensitivity that would be seen in that region otherwise. This is because at low spatial frequencies, detection is mediated by the higher harmonics and not the fundamental frequency as would be the case in higher spatial frequencies

(Campbell et al., 1970). Measuring in the low spatial frequency region with square waves

20 means that our readings will be independent of spatial frequency and scale with the maximum contrast sensitivity of which the subject is capable. For a channel of fixed bandwidth, the amount of harmonic energy within that bandwidth will be constant for the square wave no matter what its spatial frequency is (as long as the fundamental frequency is low enough).

Figure 7. The Stripe Card Contrast Sensitivity

The Berkeley Discs of Contrast Sensitivity. Professor Ian Bailey is in the process of refining the design for this novel test of pediatric contrast sensitivity (pictured in

Figure 8 below). The test consists of three double-sided plastic cards, with each side containing discs of 5 cm in diameter randomly positioned within a 7.5 cm six-cell grid.

Ian Bailey has presented some of his work on this chart alongside the newly released

Berkeley Rudimentary Vision Test at the Association for Research in Vision and

Ophthalmology (“ARVO”) conference in 2011. Testing was performed with this chart on

54 subjects from the California School for the Blind, The Orientation Center for the

Blind, and the San Francisco Lighthouse. Bailey et al. found that when contrast sensitivity was poor, generally better scores were obtained with the Berkeley Discs than 21 with the Mars chart, presumably because of the larger target size and simpler task (Bailey et al., 2011). Measurements from 0.00 (100%) log contrast sensitivity down to 1.95

(1.1%) are possible to the nearest 0.15 log unit. The three discs printed on a given card face are separated by 0.60 log unit (4x or two-octave) steps starting from full-contrast, with the in-between values shifted by a 0.30 log unit (2x or one-octave) step on the reverse side. The discs on the second card are shifted 0.15 log units towards lower contrast from those on the first card. Printing in this manner allows for a clinician to move immediately from a card face on which a patient failed to detect a disc directly to the corresponding front or back card face of the second card to measure to the nearest

0.15 log unit. The first two cards cover a range of log contrast sensitivity values from

0.00 (100%) to 1.65 (2.2%).

Figure 8. The Berkeley Discs of Contrast Sensitivity

Significance of Contrast Sensitivity Measurement. Contrast sensitivity assessments can reveal hidden losses of visual function not captured by visual acuity testing. Diseases such as age-related macular degeneration, diabetes and glaucoma can cause vision losses that acuity measurements may fail to reveal. Visual impairment from 22 contrast sensitivity arises when LogCS values of less than 1.50 are obtained and visual disability is classified as LogCS less than 1.05 (Susan J Leat, Legge, & Bullimore, 1999).

From work performed by Marron in the 1980’s, it turns out that contrast sensitivity has been shown to be better correlated than acuity with orientation and mobility in patients with reduced vision (Marron & Bailey, 1982). Contrast sensitivity measurements are useful for a number of clinical purposes from post-surgical outcome monitoring and disease progression to patient-centered outcomes such as: reading, visual task performance, orientation and mobility, driving ability, facial recognition, and vision- related quality of life (Arden, 1978; Bochsler, Legge, Kallie, & Gage, 2012; Ginsburg,

2003; Lovie-Kitchin, Bevan, & Hein, 2001; Owsley & Sloane, 1987; Owsley, 2003).

Vision-Related Quality of Life Assessment

Medical examiners classify visual disorders along a continuum where pathology just “outside normal anatomical limits” worsens until an impairment of visual function arises. Increasing levels of visual impairment can cause visual disability or even total handicapping of the individual’s ability to complete the complex visual tasks required for daily living (The World Health Organization, 1980). Aside from the magnitude of visual loss, the activity level and visual goals of a person modify the effect size resultant from vision loss along this continuum. Directly asking patients questions regarding their perceived “functional reserve” for common visual tasks is a popular method for measuring the impact of vision loss on quality of life. Functional reserve can be defined as the difference between a person’s ability and the ability required to perform a given task (Kirby & Basmajian, 1984). A typical approach for this method is to provide

23 examples of specific tasks, and then ask patients how difficult they perceive each task would be for them to complete.

One confounding aspect for the questionnaire approach lies in the “latent factors” related to a person’s visual functioning. As will be explained further in this section, latent factors cannot be directly observed, but only inferred. For most complex tasks performed in daily life, there is no obviously “correct” response to concretely quantify the amount of visual difficulty associated with that task. Therefore, the difficulty of a specific visual task must be obtained using psychometric approaches and statistical models. While direct measurement isn’t possible, it is possible for the items to be arranged by order of relative difficulty for a set of survey respondents.

The other side of the questionnaire approach is the person responding to the questionnaire, and no two people are exactly alike. Each person has a different functional reserve available to him or her for a given task. The latent ability of a given patient cannot be directly measured either, but sorting by perceived ability level based upon responses to a set of survey items is possible (Massof, 2002).

Georg Rasch, a Danish mathematician, developed a set of latent variable measurement models in the 1960s for research in educational test development (Rasch,

1960). It has since been applied in the healthcare setting. The basic premise is that the probability of selecting a given response is equal to the difference between the ability of the person taking the survey and the difficulty of (or ability required for) a given test item. Relative item difficulty and person ability levels should follow a normal distribution in most cases. The probability of obtaining a score on the extreme ends of a

24 normal curve is much lower, so logits (logarithmic odds ratios between subject ability and survey difficulty) are used to allow for the scores to scale evenly. If a person’s overall logit score is positive, then they perceive their ability to be relatively higher than the average ability required across all survey items. Total quality of life scoring and standard error values are based on the results from the performance of all subjects on the entire questionnaire.

A well-designed survey will effectively stratify the participants’ relative ability levels and item difficulty levels (evidenced statistically by good separation indexes) so that differences between respondents can be measured. It is also expected that a histogram distribution of person ability will align well with the corresponding item difficulty histogram for a survey. This kind of comparison is generally performed on a

“subject-item map.” Each item on the survey is checked for “fit statistics” to ensure that all questions are valid. The items analyzed together in a questionnaire should all target the same latent trait—perception of one’s visual ability, in this case—otherwise the measurements cannot be considered together (Massof, 1998).

The IVI_C. Researchers developed the Impact of Visual Impairment on Children

(IVI_C) questionnaire in 2008 by working with focus groups in four Australian states

(Cochrane, Lamoureux, & Keeffe, 2008). The original survey had 30 questions. The authors used Rasch analysis in 2011 to check the quality of the survey and found it to be psychometrically valid for use on children with visual impairment from age 8 to 18

(Cochrane, Marella, Keeffe, & Lamoureux, 2011). The Rasch-modified survey lists 24 questions with five answer choices: “always,” “almost always,” “sometimes,” “almost

25 never,” and “never”. An additional answer choice of “no, for other reasons” allows patients to describe an item that they cannot answer for non-visual reasons. One of the defining features of the IVI_C is that it uses positive phrasing for the majority of the survey questions. Many of the questions follow a pattern similar to: “how confident are you about…” instead of “how difficult is it for you to…” Six of the questions are negatively phrased and spaced within the survey to prevent a response bias. Naturally, the responses to these six negative questions are reverse scored. The survey includes questions regarding social aspects of a child’s school experience in addition to questions pertaining to vision/mobility. The IVI_C has been applied outside of Australia and found to have good transferability. The survey does have a slight bias towards the assessment of students with lower ability levels and is therefore susceptible to a ceiling effect when applied elsewhere (Cochrane et al., 2011).

The LVP-FVQ. Vijaya Gothwal collaborated in 2003 with Jan Lovie-Kitchen and Rishita Nutheti to develop the Low Vision Prasad Functional Vision Questionnaire

(LVP-FVQ) (Gothwal, Lovie-Kitchin, & Nutheti, 2003). The work was performed in

Hyderabad, India, and the survey was developed for research on eye care service-delivery models in rural South India. Rasch analysis was employed from the beginning of survey development. The final nineteen-item questionnaire contains four functional vision domains: 1) distance vision, 2) near vision, 3) color vision and 4) visual field extent. The survey was designed to assess practical problems resultant from pediatric vision loss in developing countries. Unlike some other QoL surveys, it does not include items pertaining to social or emotional experiences (DeCarlo, McGwin, Bixler, Wallander, &

Owsley, 2012). All questions are phrased according to an estimated amount of difficulty for a given complex visual task. The response options are: “no difficulty,” “a little difficulty,” “a moderate amount of difficulty,” “a great deal of difficulty,” and “unable to do.” An additional response option of “not applicable” was also included. The LVP-FVQ survey concludes with a final question that differs from the nineteen preceding questions:

“How do you think your vision is compared with that of your normal-sighted friends? Do you think your vision is  As good as your friend’s  A little bit worse than your friend’s 

Much worse than your friend’s?” This question was designed to assess a patient’s overall rating of their vision to see if the personal ability levels measured via Rasch analysis would match up to this gold standard using a receiver operating characteristic curve.

The LVP-FVQ was revised recently and now includes twenty-three questions, six of which were retained from the original survey, and two survey questions that are actually derived from the IVI_C. The new survey still includes the final global rating of visual impairment question (relative to normally-sighted friends) and now incorporates that question into the Rasch analysis. An attempt was made to introduce mobility and orientation specific questions into the new version of the survey, but Rasch analysis revealed that doing so would affect the one dimensional nature of the survey and adversely affect the measurement validity (Gothwal & Sumalini, 2012).

Orientation and Mobility Assessment

The orientation and mobility training techniques that exist today were developed as a result of the demand generated for these services by the significant number of traumatically blinded veterans returning from World War II. The Academy for

Certification of Vision Rehabilitation and Education Professionals has provided accreditation for orientation and mobility (O&M) specialists for over thirty years. O&M instructors teach individuals with reduced vision how to gain the capacity for confident spatial awareness and safe travel.

Many people have investigated the correlation between reduced vision and orientation and mobility (Black et al., 1997; Geruschat, Turano, & Stahl, 1998; Goodrich

& Ludt, 2003; Kuyk, Elliot, & Fuhr, 1998; Long, Rieser, & Hill, 1990). Sheila West and her colleagues included mobility as a primary outcome measure in their Salisbury Eye

Evaluation (SEE) Project, which they performed to determine the association between visual impairment and everyday task performance (West, Rubin, Broman, & Mun, 2002).

They determined the level of contrast sensitivity reduction that resulted in more than 50% of their study population to perform various tasks at 1 standard deviation below the population mean. They found that a LogCS of 1.35 or worse affected reading speed and facial recognition. Additionally, West et al. stated that a LogCS of 0.90 or worse had a measurable impact on mobility. Different cutoff points for different tasks were anticipated because the level of visual demand for a given task varies. The ability to detect low spatial frequencies in one’s environment is important for navigation.

Preferential looking tasks such as the SCCS and others are able to measure this ability.

For this reason, we will include orientation and mobility assessments in this research

(Susan J Leat & Wegmann, 2004).

The Michigan Orientation and Mobility Severity Rating Scale. The current version of The Michigan Orientation and Mobility Severity Rating Scale (OMSRS) was

28 completed in 2008 by a task force of the Michigan Department of Education Low

Incidence Outreach. Instructors use the OMSRS to approximate the amount of time that a student with visual impairment may require for orientation and mobility training.

Educators find this information is valuable when formulating individualized education plans for their students.

The OMSRS consists of eight categories: 1) Medical level of vision [central and peripheral], 2) Functional level of vision, 3) Use/proficiency of travel tools, 4)

Discrepancy in travel skills between present and projected levels, 5) Independence in travel in current/familiar environments, 6) Spatial/environmental conceptual understanding, 7) Complexity or introduction of new environment, and 8) Opportunities for use of skills outside of school. Each of these categories are scored on a scale of 1-5 using a rubric where a higher score indicates greater severity of need and more time devoted to O&M training. The OMSRS also lists several contributing factors that may be used to adjust the scores given for the categories above (see Appendix A).

Experiment Overview.

Our approach to aid in the care of children and others who struggle with lettered eye charts is to design a new test of contrast sensitivity that complements testing performed with Teller Acuity Cards. Our research was geared towards ensuring that the results from the new test are applicable and valid. We also aim to discover how well these results align with vision-related quality of life as reported on survey questionnaires designed specifically for children with visual impairments.

To validate the Stripe Card Contrast Sensitivity (SCCS), we tested a group of students at The Ohio State School for the Blind (OSSB). Lettered charts used were the

Bailey-Lovie and Pelli-Robson (B-L, P-R) charts. Non-lettered charts included: The

Teller Acuity Cards (TAC), SCCS, and Berkeley Discs of Contrast Sensitivity (BD). A good outcome would be if the SCCS test results were positively correlated with the results from the other contrast tests, and if the TAC test results correlated with the other visual acuity test results. It would also be good if the various vision tests positively correlated with the measures of QoL and O&M. The details of the relationships between the various tests might indicate which tests are better for different patients.

We related the results found with these eye charts to self reports of participant vision-related quality of life (QoL) using two questionnaires: The Impact of Visual

Impairment in Children (IVI_C) and The Low Vision Prasad Functional Vision

Questionnaire (LVP-FVQ). Additionally, we obtained O&M scores for a subset of participants evaluated by their instructors for relation back to the eye chart test results.

The rubric used by these instructors was The Michigan Orientation and Mobility Severity

Rating Scale (OMSRS) and it can be found in the Appendix.

Research performed in the 2012-13 school year included 27 participants who were tested monocularly using the patient’s preferred eye. We will refer to the results of these measurements as “Experiment I” below. Ocular dominance testing was performed using an eye sighting technique if the patient was unable to report a preferred eye. We initially chose the dominant eye for three reasons: 1) functional vision is generally driven by the preferred eye, 2) if performance is not driven exclusively by the better eye, then

30 using a monocular condition should remove any ambiguity regarding the relative contribution of each eye, and 3) testing only one eye streamlines the examination process.

The following year, we returned for repeat testing of 11 participants from the first year (“Experiment IIa”) and additional testing of 24 new participants (“Experiment

IIb”). In an effort to increase the amount of data collected with our five vision tests,

Experiment II assessments were performed on each eye monocularly (where possible) rather than just with the preferred eye. When able to test each eye, the study was initiated using the participant’s right eye first.

We have obtained vision-related quality of life data for all but one subject. We have also obtained orientation and mobility scores from O&M instructors for about half of the subjects for whom the data was potentially available. The results of these non- visual measures fall under Experiment III below.

Ethics

The protocol for the study was approved by the Biomedical Sciences Institutional

Review Board (IRB) of The Ohio State University and followed the tenets of the

Declaration of Helsinki. Full informed consent or parental permission and child assent were obtained before the start of all experimental work and data collection.

Recruitment

The Ohio State School for the Blind is a publicly funded educational facility for students with visual handicaps in grade school up through high school. Students range in age from five to twenty-one years old, with the most common age being fifteen. The student body at OSSB is about 25% under-represented minority, and about 16% of the

31 students there have other disabilities in addition to vision loss. Around half of the students spend the entire week at the school. These residential students leave for home by bus at early dismissal on Friday and then returning on Sunday afternoon. The school does not operate during the summer, but does run summer camps open to all Ohio students with visual impairment interested in attending.

OSSB is also a clinical outreach rotation site for fourth year students at the OSU

College of Optometry. The college furnishes an exam room located within the nurse’s station for our exclusive use. An optometry student practices under the mentorship of a clinical preceptor (the author) on Wednesday mornings for three months. Copies of the examination results are kept at the College of Optometry and at OSSB. Approval was obtained from the university’s IRB for a HIPAA waiver allowing study investigators to view the eye care and medical records kept by the school to determine which students may have measurable vision. At the time of our research project, total enrollment at

OSSB was approximately 115 students and the author determined by chart review that approximately fifty-three (46%) of the students were likely to have sufficient vision for testing (see Figure 9).

5% 23% UNK

20/###

HM 44% LP

22% NLP

Unable 5%

Figure 9. All OSSB Student Visual Acuities by Chart Report

Information packets regarding the study opportunity were assembled and sent via metered mail to the guardians of all fifty-three students with vision recorded as hand motion or better. Sample contents of these information packets can be found in Appendix

A. Briefly, each packet contained cover letters from the school principal and our study group, a parental consent and HIPAA form, a response checklist and a self-addressed envelope for returning signed documents.

Participant Characteristics

Forty-three of our fifty-one participants were students with partial sight, all of whom were examined at the Ohio State School for the Blind (OSSB). Their ages were 5-

21 years old, thirty-three were males and most (forty-two subjects) were Caucasian race.

Eight participants were students in the annual summer camps put on by OSSB with ages from 11—18 years. Four of this group were female, and seven were Caucasian race.

As shown in Figure 10, optic nerve disorders characterized the majority of primary diagnoses in our study sample, at 43% of the total sample. Participants with retinopathy of prematurity were the next most prevalent, comprising 13% of our sample.

Various other disorders, including congenital cataract, cortical blindness, and genetic conditions, made up the rest of our study population

Figure 10. Participant Diagnoses 34

Participants for Experiment I Only ID Age MF Dx Eye(s) O&M 2 8 M Retinopathy of Prematurity OD

4 17 F Optic Nerve Hypoplasia OS

5 12 F Leber's Congenital Amaurosis OS

6 15 F Optic Nerve Hypoplasia OD

8 19 F Optic Atrophy OD

11 15 M High Myopia OD

18 11 F Cone Dystrophy OS X 24 12 F Optic Atrophy OS

34 8 M Congenital Cataract and Aphakia OD X 35 12 M Retinopathy of Prematurity OD

36 18 M Optic Atrophy OS

41 16 M Glaucoma OS X 43* 11 M Microphthalmus OS

46 13 M Cortical Blindness OD

48 15 F Optic Atrophy OD

50 21 M Septo-Optic Dysplasia OS X

Experiment IIa Repeat Participants 1 17 M Optic Nerve Hypoplasia RL X 12 16 M Retinopathy of Prematurity OS X 17 14 F Microphthalmus OS

26 18 M Optic Nerve Hypoplasia RL X 28 12 F Congenital Cataract and Aphakia RL

30 18 F Optic Atrophy RL X 37 18 M Aniridia RL X 45 20 F Optic Atrophy RL X 51 13 M Optic Atrophy RL

52 15 M Optic Atrophy RL

53 16 M Cortical Blindness RL

Continued Table 1. Complete Participant List * Number 43 was only participant unable to provide Quality of Life Data.

Table 1 Continued

Experiment IIb Additional Participants 31 16 M Retinoblastoma OD 54 20 M Retinopathy of Prematurity NA X 55 19 F Leber's Congenital Amaurosis NA X 56 19 M Brain Tumor NA X 57 19 M Optic Nerve Hypoplasia NA X 58 18 F Leber's Congenital Amaurosis RL X 59 20 M Retinopathy of Prematurity NA X 60 18 M Retinopathy of Prematurity RL X 61 13 M Choroideremia RL

62 11 F Blind at age 2 (unexplained) RL

63 10 F Leber's Congenital Amaurosis RL

64 13 M Optic Atrophy OS

65 16 F Marfan's Disease RL

66 17 M High Myopia OD

67 10 M Cortical Blindness OS

68 12 M Aniridia RL

69 14 M Retinopathy of Prematurity RL

70 18 M Retinitis Pigmentosa NA X 71 15 M Best's Disease RL X 72 18 F Glaucoma OS X 73 17 F Optic Atrophy NA X 74 12 M Optic Nerve Hypoplasia NA

75 12 M Retinopathy of Prematurity NA

76 6 M Retinoblastoma NA

Objectives

The work underlying this thesis is intended to:

 Ascertain whether the Stripe Card Contrast Sensitivity (SCCS) Test is easy to

administer and can be successfully used on a wide variety of participants.

 Assess the SCCS test’s capability for validly measuring contrast sensitivity levels

of children with impaired vision.

 Ascertain the relationship between SCCS test results and other functional

measurements such as: visual acuity, letter contrast sensitivity, vision-related

quality of life metrics, and orientation and mobility assessments.

Experiment I

Study Design

To ascertain the effectiveness and validity of the SCCS Test, assessments needed to be performed on young patients with reduced vision because this is the intended clinical population. In order to avoid possible fatigue effects from repeat vision assessment, a pseudo-randomization scheme was necessary. Accordingly, a five item balanced Latin square array guided the order of test presentation for each study encounter. Duration of testing was recorded to the nearest second for each eye on each chart. Room and chart illumination measurements were taken periodically using an iOS application (Megaman LuxMeter). Vision-related quality of life assessments followed vision testing in all cases, and the IVI_C was performed prior to administering the LVP-

FVQ in each case. More will be said about these surveys in another section.

Study Methods

The school nurse (or an aide) escorted eligible subjects to the OSSB examination room, in which most have previously had routine eye care performed. Two 18 watt/950 lumen compact fluorescent flood lamps positioned just above and behind the subject provided additional lighting in the range of the recommended 85 cd/m2 onto the front surface of all the eye charts. An IRB approved verbal assent or consent form was read aloud to all participants, and their responses were noted. Participants were invited to ask questions or discontinue participation at any time. No potential subjects declined to

38 participate when read the assent/consent materials and only one participant discontinued the study (just before beginning the QoL assessment).

Letter Acuity Procedure. When possible, each child in our study was examined with the same version of a Bailey-Lovie acuity chart. Refraction was not performed prior to the initiation of vision testing. Participants wore their habitual correction. Visual acuity assessments were performed with the Bailey-Lovie chart at 2 meters (or closer when necessary) and scoring was calculated using total number of letters read correctly without any substitutions. Test distance variations were taken into account so that each letter represented a value of 0.02 LogMAR. We employed letter-by-letter scoring in our protocol, as this method has been shown to provide better reliability (Raasch et al., 1998).

We did not permit substitutions, such as accepting the letter “C” for “O,” but we did remind participants that the list of possible letter choices was restricted to those letters that could be found on the top three rows. The stopping rule was enforced whenever a subject could not identify 3/5 of the letters on a given row. While the chart is available at two different levels of contrast, we used only the high contrast version: Chart #4. The use of this chart version was purely arbitrary, but we intended to maintain the use of this version across all eyes tested in the experiment to control for any small amount of variation that could be the result of employing a different optotype set.

Grating Acuity Procedure. We used a set of Teller Acuity Cards manufactured in 1991 by Vistech in order to avoid luminance artifacts present in some of the cards

(Teller Acuity Cards II) made by the other major manufacturer, Stereo Optical.

Normative data exist for the Vistech cards, and they incorporate some design

39 modifications from the initial prototype set of Teller Cards (Mayer et al., 1995). Many of the subjects who participated in our research were able to reliably point towards the stripes when they saw them. Therefore, the use of a peephole was not necessary, but the examiner did make assessments of the patient’s looking behavior throughout the examination. The stopping rule was whenever a patient admitted that they could not see the stripes or the examiner was convinced that they could not. We did not need to present an initial blank card for the purpose of calibrating the examiner to the participant’s looking behavior. In order to screen for potential luminance artifacts, we asked subjects to explain whether or not it was the actual stripes they saw or if, instead, they were detecting a “box” of different luminance than the rest of the card. Adjustments in working distance or card orientation from horizontal to vertical are possible, and a “low vision” test distance of 38 cm has been outlined in the Teller Acuity Card user guide, but we did not find it necessary to employ those techniques at any point. We maintained a 55 cm test distance, recording the cy/deg of the last identifiable card by patient report.

Letter Contrast Procedure. We performed calibration testing with a

SpectraScan photometer on a Pelli-Robson chart and employed only this calibrated chart.

Naturally, the purpose of using only the calibrated chart was to control for any variability that may result due to discrepancies between the nominal contrast and the actual measured value. Two luminance values were obtained from the center of a given letter stroke as well as the adjacent white background. These values were averaged separately and then the contrast percentage was calculated using Michelson’s formula. The log contrast values on the chart were close to being linearly related to the nominal log

40 contrast values on the chart. However, the contrast levels on the chart were lower than the nominal ones (perhaps due to gradual fading), so the nominal values underestimate the subject’s true contrast sensitivity by 0.36 ± 0.15 LogCS on average. I fit a trend line to the calibration results with the following formula: y = 1.18x + 0.15 so as to generate corrected P-R values. Here, I will report the nominal values in the tables and figures below because those are the ones that every clinician will have available for clinical use.

The results obtained on the better eye (or second measurement) of each subject remain statistically significant. The comparison between the P-R vs. SCCS is significantly different at the p < .01 level using the calibrated values instead of p < 0.001 for the nominal P-R values. The difference between the results obtained for the Berkeley Discs and the calibrated P-R values continue to be not statistically significant.

We generally tested at the recommended distance of 1 meter, only shifting to closer than 1 meter for a few students with very poor acuity. Scoring was letter-by-letter, and each optotype counted for 0.05 log units (Dougherty, Flom, & Bullimore, 2005). The stopping rule was whenever a subject could not get 2/3 of a letter triad correct.

Stripe Card Contrast Sensitivity Test. Calibration was performed on our SCCS cards in a similar manner as was performed for the Pelli-Robson chart. Our calibrated contrast values were only slightly lower than the nominal ones, requiring a greater sensitivity of 0.013 ± 0.03 LogCS. We calibrated our cards prior to initiation of testing and were able to measure down to 1.65 LogCS during the first year of testing (and down to 2.00 during the second year of testing when additional cards were produced).

Test presentation style and stopping criteria were the same as for Teller cards, with the preferential looking technique was used to determine a subject’s threshold contrast sensitivity. We tested at one card length (57 cm) from the subjects and threshold was determined to be the point at which the subject could no longer find the stripes.

The Berkeley Discs of Contrast Sensitivity. As stated earlier, the first two cards cover a range of log contrast sensitivity values from 0.00 (100%) to 1.65 (2.2%).

Precision-Vision provided us with calibration readings for the two cards supplied, and we obtained our own photometry measurements prior to vision testing. An additional third card was not available for our testing due to production difficulties experienced by the manufacturer (Precision-Vision). Thus, we could not evaluate performance at log contrast sensitivity levels of 1.80 (1.6%) or 1.95 (1.1%).

Berkeley discs were presented at 40 cm with a rubberized wand for subjects to use to point to discs. All card sides were presented even when performance was perfect on the first card. The design of the cards allows for a clinician to move immediately from a card face on which a subject failed to detect a disc directly to the corresponding card face

(front or back) of the second card to measure to the nearest 0.15 log unit. However, we allowed our participants to view both sides of each card even if it were possible to skip ahead. We did not enforce a stopping rule for the Berkeley Discs since we had only a two-card set. Most subjects were able to see discs on all card faces, so we presented every student with the opportunity to view both sides of each card.

Results for Experiment I

Below is a summary table for the results on all students tested in Experiment I.

B-L TAC P-R SCCS BD ID Age S Diagnosis Eye LogMAR LogMAR LogCS LogCS LogCS 01 17 M Optic Nerve Hypoplasia OS 1.28 0.50 0.65 1.65 1.65 02 8 M Retinopathy of Prematurity OD 0.62 0.21 1.95 1.65 1.65 04 17 F Optic Nerve Hypoplasia OS 0.48 0.08 2.10 1.65 1.65 05 12 F Leber's Congenital Amaurosis OS 1.42 1.56 1.10 0.60 1.50 06 15 F Optic Nerve Hypoplasia OD 0.74 0.38 2.00 1.65 1.65 08 19 F Optic Atrophy OD 0.88 0.21 1.45 1.65 1.50 11 15 M High Myopia OD 0.94 0.99 1.35 1.65 1.65 12 16 M Retinopathy of Prematurity OS 0.54 0.68 1.80 1.65 1.50 17 14 F Microphthalmus OS 1.40 1.56 0.85 1.20 1.05 18 11 F Cone Dystrophy OS 0.70 0.81 1.40 1.65 1.65 24 12 F Optic Atrophy OS 1.16 0.50 1.65 1.65 1.35 26 18 M Optic Nerve Hypoplasia OS 1.02 0.21 1.80 1.65 1.65 28 12 F Congenital Cataract OD 1.16 0.50 1.65 1.65 1.65 30 18 F Optic Atrophy OS 0.88 0.38 1.00 1.50 1.65 34 8 M Congenital Cataract OD 0.22 0.21 1.65 1.65 1.50 35 12 M Retinopathy of Prematurity OD 0.82 0.68 1.75 1.65 1.65 36 18 M Optic Atrophy OS 1.60 0.81 0.90 1.65 1.65 37 18 M Aniridia OD 0.66 0.68 1.75 1.65 1.65 41 16 M Glaucoma OS 1.96 2.45 0.10 1.20 0.90 43* 11 M Microphthalmus OS — 1.56 — 0.75 0.00 45† 20 F Optic Atrophy OD — 1.11 — 1.35 0.30 46† 13 M Cortical Blindness OD — 0.84 — 1.35 0.75 48 15 F Optic Atrophy OD 1.20 0.68 1.15 1.65 1.50 50 21 M Septo-Optic Dysplasia OS 1.90 0.68 0.70 1.50 0.30 51 13 M Optic Atrophy OS 1.18 0.38 0.75 1.35 0.75 52 15 M Optic Atrophy OS 2.08 0.50 0.80 1.65 1.35 53 16 M Cortical Blindness OD 2.50 0.21 0.15 1.35 1.20 Table 2. Experiment I Participants n = 27 students (27 eyes tested) *Subject 43 was unable to provide quality of life data, nor was he testable by lettered charts. †These subjects were not testable via lettered charts.

Result values of all test encounters are summarized in Tables 3 and 4 below, along with average testing times per eye. For the SCCS test, 17/24 (71%) of subjects obtained the maximum log contrast sensitivity value that we could measure (1.65). For the BD test, 12/24 (50%) obtained the maximum contrast sensitivity value we could measure (also 1.65). 43

B-L TAC P-R SCCS* BD*

Mean 1.13 (20/274) 0.66 (20/92) 1.27 1.65 1.58 Median

STDev ± 0.56 ± 0.54 ± 0.57 1.50 1.35 75th %

Time (sec) 54 ± 33 97 ± 67 61 ± 42 55 ± 45 31 ± 19 Time (sec)

Table 3. Experiment I Summary Test Results

Acuity and contrast sensitivity are presented in LogMAR and LogCS units, respectively.*Unable to measure values better than 1.65. Note: if the calibrated P-R values are used, the results would be 0.30±0.15 LogCS higher.

Two-tailed t-test comparisons of pairwise findings for the acuity and contrast sensitivity measures are shown below. Each comparison was run independently of the others. It is understood that our data do not likely meet all the assumptions required to run t-test analysis. However, parametric statistics have been attempted nonetheless.

2. 5

2 p < 0.001** 1. 5 1 0. 5 B- L TA C 0 P- R SC C S * B D* -0 . 5 -1 -1 . 5 p < 0.01** -2 p < 0.1 1 -2 . 5 p < 0.15

Figure 11. Experiment I Summary Plot Statistics Test results are shown with contrast values in the negative direction so as to keep the direction of better vision consistent (higher values indicate poorer vision). Average values are shown for the B-L, TAC and P-R charts, with error bars indicating standard deviations. *The SCCS and BD tests were limited to a ceiling of 1.65 logCS, so their bar graphs represent median values and error bars extend up to the 75th percentile contrast value. Note: if the calibrated P-R values are used, the difference between the P-R and SCCS is no longer statistically significant (p < 0.14).

We compared the Bailey-Lovie letter acuity values obtained through my testing with those reported for our subjects in their charts (Figure 12A). These charted acuity values were obtained within 1-2 years of our testing by a different optometrist in concert with a fourth year optometry student. Values were obtained using the same physical

Bailey-Lovie chart that we used, or with a Feinbloom number chart. Our testing was done using letter-by-letter tallying, but the values in the record were obtained using a three out of five correct identification criterion for the last row read. In the mean-difference plot

45 below (Figure 12B), the average of my value and the chart value runs along the abscissa and the difference between the reported value and my measurement is plotted on the y- axis (Bland & Altman, 1986). The standard deviation of our measurement differences from those in the record was ±3 lines. These results were within acceptable tolerances for repeatability of visual acuity measures and therefore show good agreement with the corresponding values reported in the participant’s records (Raasch et al., 1998).

3 Reported Better 2.5

L 2 -

1.5

Measured B Measured 1

0.5 Measured Better

0 0.00 1.00 2.00 3.00 Reported B-L

B 3 2.4 1.8 1.2 0.6

Reported 0 - -0.6 -1.2 -1.8 -2.4 -3

Measured 0 1 2 3 Average B-L Values

Figure 12. Experiment I B-L vs Chart Report Comparison between performance on the Bailey-Lovie chart and the visual acuities reported in the subject’s patient file. A, Direct LogMAR comparison between the two data sets. Major diagonal is the equality line. Data above the equality line indicate better performance noted in the patient’s file. Data below the equality line indicate better performance when measured in our study. B, Data from A presented as a Bland-Altman plot.

I have used a novel format to show individual results for acuity and contrast sensitivity testing for each eye tested. Figure 13 shows results that are clustered by 47 primary diagnosis and sorted by visual acuity within diagnoses. As in Figure 11 above, we graph contrast threshold instead of sensitivity in order to keep the direction of better vision consistent (higher values indicate poorer vision).

Optic Nerve Congenital Optic Cortical ROP Hypoplasia Cataract Atrophy Other Blindness 2.7 2.4 2.1 B-L Acuity 1.8 1.5 1.2 0.9 0.6 0.3 0 ** ** -0.3 ** ** -0.6 -0.9 -1.2 -1.5 -1.8 -2.1 Good LogCS LogMAR Poor LogMAR LogCS Good -2.4 ** Unable -2.7 P-R Contrast Sensitivity

Figure 13. Experiment I Lettered Chart Results by Diagnosis LogMAR visual acuity and Log contrast sensitivity measured using lettered charts. Each red + blue bar represents an individual subject’s visual performance. The upper edge of the blue bar on top is the subject’s LogMAR acuity, and the lower edge of the red bar on the bottom is the subject’s LogCS. The bar as a whole is lower for better performance. Bars are grouped by diagnosis and within diagnoses they are sorted by visual performance on the LogMAR chart. Subsequent charts in this format maintain subject order from this graph.

Optic Nerve Congenital Optic Cortical ROP Hypoplasia Cataract Atrophy Other Blindness 2.7 2.4 2.1 TAC 1.8 1.5 1.2 0.9 0.6 0.3 0 -0.3 -0.6 -0.9 -1.2 -1.5 -1.8 Good LogCS LogMAR Poor LogMAR LogCS Good -2.1 SCCS -2.4 -2.7

Figure 14. Experiment I Striped Chart Results by Diagnosis Log MAR visual acuity and log contrast sensitivity using striped charts. Conventions and subject order are as in Figure 13. Note that the SCCS test had a ceiling of 1.65 LogCS (shaded area).

It is difficult to compare Berkeley disc data directly to Teller Acuity values in the way that the SCCS can be. The Berkeley Discs certainly require a different amount of visual search ability than the Bailey-Lovie and Teller Acuity Cards do. The results for the

Berkeley Discs are shown in Figure 15 against the average of Bailey-Lovie letter acuity and Teller Acuity Card grating acuity, since there is a localization element to the

Berkeley Discs that is not present in the SCCS test.

Optic Nerve Congenital Optic Cortical ROP Hypoplasia Cataract Atrophy Other Blindness 2.7 2.4 2.1 Ave B-L & TAC 1.8 1.5 1.2 0.9 0.6 0.3 0 -0.3 -0.6 -0.9 -1.2 -1.5 -1.8

Good LogCS LogMAR Poor LogMAR LogCS Good -2.1 BD -2.4 -2.7

Figure 15. Experiment I Shaped Chart Results by Diagnosis Experiment I Shaped Chart Results by Diagnosis. Here the average between B-L and TAC are used for the acuity value given that the BD test is neither a letter nor a stripe chart and therefore not directly comparable to either acuity measure employed. Conventions and subject order are the same as for Figure 13. Note that the BD had a ceiling of 1.65 LogCS (shaded area).

The Teller Acuity card values were significantly better than those using the

Bailey-Lovie chart, especially for those subjects with acuity worse than LogMAR 1.0

(20/200) (see Figure 16). However, it was clear that the TAC values and the LogMAR values were statistically significantly correlated with one another. These correlations will be examined in a later section of the thesis.

3 Letters Better 2.5

1.5 Stripes Better

TAC LogMAR TAC R² = 0.128 1

0.5

0 0 1 2 3 B-L LogMAR

B 3 2.4

L 1.8 -

B 1.2 0.6 – 0 -0.6 -1.2 TAC TAC -1.8 -2.4 -3 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3 Average LogMAR B-L & TAC

Figure 16. : Experiment I Acuity Results Comparison between visual acuity measured with the B-L chart and the TAC. A, direct comparison as in Figure 12. B, Bland-Altman plot as in Figure 12.

As seen in Figures 17 & 18, the SCCS test showed a ceiling effect during the first school year of testing, with 65% of subjects scoring the maximum LogCS (1.65) and the 51 lower quartile scoring at least LogCS of 1.35. All contrast sensitivity plots have been produced with negative values in order to maintain directional consistency with acuity plots, where a lower value is indicative of better performance.

-0.15 Letters Better -0.45

-0.75

-1.05 R² = 0.2191

-1.35 SCCS LogCS SCCS

-1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

1.95 R

- 1.35

P 0.75 – 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS P-R & SCCS

Figure 17. Experiment I P-R & SCCS Results Comparisons between contrast sensitivity measured with the SCCS test and the Pelli-Robson chart. A, Direct comparison as in Figure 12. B, Bland-Altman plot as in Figure 12.

8 SCCS P-R

6 NumberSubjects of

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 17. Experime nt I P-R & SCC S Bins

Figure 18. Experiment I P-R & SCCS Bins The number of subjects in each contrast sensitivity group for the Pelli-Robson and Stripe Card tests as shown in Figure 17.

All but one of the subjects with impaired CS on the PR (according to Leat et al.,

PR<1.50) showed a better LogCS on the SCCS test than the PR test (sign test, p< 0.05, 2- tailed; nonparametric test required because of the ceiling effect on the SCCS).

The results from the Berkeley Discs were also typically better than those found with the Pelli-Robson chart as well, but not to the same degree as the SCCS test (as shown in Figure 19).

-0.15 Letters Better -0.45

-0.75

R² = 0.3637 -1.05

BD BD LogCS -1.35

-1.65

-1.95 Discs Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

B 1.95

1.35 R - 0.75 P 0.15 – -0.45

BD -1.05 -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS P-R & BD

Figure 19. Experiment I P-R & BD Results Comparison between contrast sensitivity measured with the Berkeley Discs and the Pelli-Robson chart. A, direct comparisons as in Figure 12. B, Bland-Altman plot as in Figure 12.

BD 6 P-R

NumberSubjects of 4

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 20. Experiment I P-R & BD Bins The number of subjects in each contrast sensitivity group for the Berkeley Discs and the Pelli-Robson chart as shown in Figure 19.

A comparison was also possible between the results of the SCCS and Berkeley

Discs since both tests had a ceiling of 1.65 during the first year of testing. The data corresponding to fifteen of the subjects below plot at the same location on the graph, giving the appearance of less data than actually obtained. For this reason, bin count figures have been generated to illustrate the level of overlap in our results.

-0.15 Discs Better -0.45

-0.75

-1.05

-1.35 SCCS LogCS SCCS R² = 0.1353 -1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 BD LogCS

B 1.95 1.35

BD 0.75 - 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS SCCS & BD

Figure 21. Experiment I SCCS & BD Results Comparison between contrast sensitivity measured by SCCS and BD. A, Direct comparison as in Figure 12. B, Bland-Altman plot as in Figure 12.

8 BD SCCS

6 NumberSubjects of

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 22. Experiment I SCCS & BD Bins The number of subjects at each contrast level when tested with the Berkeley Discs and the SCCS test as shown in figure 21.

Discussion for Experiment I

Both of the grating tests (SCCS and TAC) gave statistically significantly better values than the corresponding letter charts. The average grating acuity value measured on

Teller cards was almost 3x better than the acuity value measured by the Bailey-Lovie chart. Statistical analysis of our results by t-test demonstrated that the two acuity charts gave significantly different results. Similarly, the contrast sensitivity values measured with the SCCS were over 2x better than the Pelli-Robson values. Again, t-test results show that the two charts give significantly different results.

Some of the inconsistency between the Pelli-Robson and SCCS test may be a result of the LogCS 1.65 ceiling limiting the SCCS results. This fact affected our ability 58 to employ a t-test without any reservations. We measured 2.4x the contrast sensitivity despite this inability to measure lower levels of contrast during Experiment I. For measuring contrast sensitivity in those with reduced vision, the simpler task and bolder patterns of the SCCS may make it more likely to reveal the maximum performance that a given patient can achieve.

The Berkeley Discs also demonstrated better results than the Pelli-Robson (a factor of 2x), but not to the same level as the SCCS test (again, 2.4x). This was despite the fact that both the BD and SCCS tests had the same LogCS 1.65 ceiling. Perhaps the localization task for the BD test was more complex than for the SCCS.

There were also inconsistent BD responses by some subjects (report of seen circles in blank areas), which make the test difficult to interpret at times. As you may recall, we allowed our subjects to view every card face instead of skipping some.

Theoretically, the subjects should have seen every disc present on a card containing contrast levels higher than previously detected. However, some of them failed to detect all the discs on those card faces. In our case, we gave the participants credit for the lowest contrast value reported, but this finding does affect the validity of our results.

Experiment II – Separate Eye Testing

Introduction to Experiment II

In this second phase of our study, we sought to examine each eye of our subjects individually, whenever possible. We also opened up participation to students with no measurable vision. These students without measurable vision were included to constrain our vision-related quality of life data analysis, discussed under Experiment III. A second information packet was mailed out to all students at the school in order to encourage repeat assessments for some students, and to allow for the inclusion of students with no measurable vision.

We also sought and obtained permission from the IRB to recruit participants from

OSSB’s summer camp activities. Angela Brown attended the registration event for summer camp at OSSB between the first and second year of testing. She was able to recruit interested participants by answering questions and distributing information packets in person.

Methods for Experiment II

The methods for Experiment II were almost entirely the same as for Experiment I.

One difference was that all assessments were performed on each eye monocularly where possible instead of just on the subject’s preferred eye. When both eyes were tested, the right eye was tested first.

New SCCS cards were developed to extend our measurement range of LogCS values from approximately 1.65 to 2.00. We calibrated the cards again prior to initiating testing and found that our cards continue to be close enough to the nominal values for good use.

The results below display testing performed on subjects allowing them to use each eye to view our charts. We allowed subjects to use their “worse” eye (provided that they had any measurable vision in that eye). Inclusion of testing for each eye, whenever possible, for each subject in Experiment II rather than just the dominant (and presumably better) eye as in Experiment I, was intended to increase the data at very low levels of visual performance for comparison across the five vision tests.

Results for Experiment IIa: Repeat Testing

Table 4 below shows the re-test results for eleven subjects who also participated in the Experiment I. Repeat testing was performed for at least one eye, and an additional eye was measured when possible. The right eye was always tested first.

B-L TAC P-R SCCS BD ID Age S Diagnosis Eye LogMAR LogMAR LogCS LogCS LogCS 01 17 M Optic Nerve Hypoplasia OD 1.58 1.11 0.85 1.95 1.05 01 17 M Optic Nerve Hypoplasia OS 1.32 0.68 1.20 2.00 1.05 12 16 M Retinopathy of Prematurity OS 0.68 0.68 1.90 2.00 1.65 17 14 F Microphthalmus OS 1.36 1.29 1.35 1.90 1.5 26 18 M Optic Nerve Hypoplasia OD 1.60 0.81 0.10 1.65 1.2 26 18 M Optic Nerve Hypoplasia OS 0.96 0.38 2.05 2.00 1.65 28 12 F Congenital Cataract OD 1.28 0.68 1.75 2.00 1.65 28 12 F Congenital Cataract OS 1.30 0.68 1.75 2.00 1.65 30 18 F Optic Atrophy OD 0.96 0.68 1.30 1.65 1.65 30 18 F Optic Atrophy OS 0.94 0.99 1.05 1.65 1.65 37 18 M Aniridia OD 0.96 0.38 1.50 1.65 1.65 37 18 M Aniridia OS 0.92 0.50 1.35 1.65 1.65 45† 20 F Optic Atrophy OD — 1.29 — 0.75 0 45† 20 F Optic Atrophy OS — 1.29 — 1.20 0.9 51 13 M Optic Atrophy OD 1.42 1.11 1.00 1.50 1.65 51 13 M Optic Atrophy OS 1.14 0.50 1.20 1.65 1.2 52 15 M Optic Atrophy OD 3.14 0.81 0.20 1.05 — 52 15 M Optic Atrophy OS 2.30 0.68 0.50 1.50 0.75 53 16 M Cortical Blindness OD 3.00 0.38 0.70 1.90 1.5 53 16 M Cortical Blindness OS 2.84 0.38 0.70 1.90 0.9 Table 4. Experiment IIa Participants n = 11 (20 eyes tested) †These subjects were not testable by lettered charts.

All test measurements are summarized in the table below, along with average testing times per eye.

B-L TAC P-R SCCS* BD†

Mean 1.54 (20/689) 0.71 (20/101) 1.14 1.76 1.50 Median

STDev ± 0.76 ± 0.28 ± 0.56 ± 0.25 1.01 75th %

Time (sec) 65 ± 29 89 ± 67 76 ± 66 79 ± 43 57 ± 31 Time (sec)

Table 5. Experiment IIa Summary Test Results

*Unable to measure values better than 2.00 †Unable to measure values better than 1.65 Time of testing is for each individual eye. Note: if the calibrated P-R values are used, the results would be 0.30±0.15 LogCS higher.

Two-tailed t-test comparisons of findings for the acuity and contrast sensitivity measures are shown below.

Figure 23. Experiment IIa Summary Plot Statistics Conventions are as in Figure 11. The ceiling was extended for the SCCS test, so the average is reported and the standard deviation is now shown for the SCCS error bar. The Berkeley Discs are still presented with median and 75th percentile values. *The SCCS ceiling was LogCS 2.00 and the †Berkeley Discs ceiling was 1.65. Note: if the calibrated values for the P-R are used, the difference between the P-R and the BD scores is no longer statistically significant (p < .18)

Importantly, in Experiment II the SCCS test no longer had a ceiling of LogCS

1.65, but instead, we could measure up to LogCS 2.00. The Berkeley discs test continued to have a ceiling of 1.65.

3 Letters Better 2.5

1.5 Stripes Better TAC LogMAR TAC 1

0.5 R² = 0.0133

0 0 1 2 3 B-L LogMAR

B 3 2.4

L 1.8 - 1.2 B 0.6 – 0 -0.6 -1.2 TAC TAC -1.8 -2.4 -3 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3

Average LogMAR B-L & TAC

Figure 24. Experiment IIa Acuity Results Comparison between acuity testing techniques of Teller Acuity and Bailey-Lovie Acuity for subjects for whom we already have seen better-eye data in Experiment I. This plot includes repeat testing of the better eyes as well as measurement of the worse eye where possible. All but two subjects contribute two eyes. A, Direct comparison as in Figure 12. B, Bland-Altman plot as in Figure 12.

-0.15 Letters Better -0.45

-0.75

-1.05

-1.35 SCCS LogCS SCCS R² = 0.3725

-1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

B 1.95

R 1.35 -

P 0.75

– 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15

Average LogCS P-R & SCCS

Figure 25. Experiment IIa P-R & SCCS Results Comparison between contrast sensitivity values obtained by the SCCS test and the P-R chart for repeat subjects’ better and worse eyes. All but two subjects contribute two eyes. A, Direct comparison as in Figure 12. B, Bland-Altman plot.

5 SCCS 4 P-R

NumberSubjects of 3

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 26. Experiment IIa P-R & SCCS Bins Number of subjects at each contrast sensitivity level when measured with the SCCS and P-R tests. These results are not limited by a ceiling of 1.65 as in Experiment I. Instead, the SCCS has a ceiling of 2.00.

-0.15 Letters Better -0.45

-0.75

-1.05

R² = 0.4349 BD BD LogCS -1.35

-1.65

-1.95 Discs Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

B 1.95

1.35 R - 0.75 P 0.15 – -0.45

BD -1.05 -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15

Average LogCS P-R & BD

Figure 27. Experiment IIa SCCS & BD Results Comparison between contrast sensitivity values measured with the BD test and the P-R chart. The BD test is bounded by a ceiling of 1.65. A, Direct comparison as in Figure 12. B, Bland-Altman plot.

5 BD 4 P-R

NumberSubjects of 3

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 28. Experiment IIa P-R & BD Bins Number of subjects at various levels of contrast sensitivity when measured by BD and P-R tests. The BD test is limited by a ceiling of 1.65.

-0.15 Discs Better -0.45

-0.75

-1.05

-1.35 SCCS LogCS SCCS R² = 0.0068 -1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 BD LogCS

B 1.95 1.35

BD 0.75 – 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15

Average LogCS SCCS & BD

Figure 29. Experiment IIa SCCS & BD Results Comparison between contrast sensitivity measured with SCCS and BD test. Here, the SCCS is bounded by a ceiling of 2.00, but the BD is bounded by a 1.65 ceiling. Many subjects obtained those ceiling values, resulting in their data points plotting to the same spot on the graph. This gives the appearance of less data than was actually obtained. A, Direct comparison as in Figure 12. B, Bland-Altman plot.

5 BD 4 SCCS

NumberSubjects of 3

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 30. Experiment IIa SCCS & BD Bins Number of subjects at each contrast sensitivity level when measured with the BD and SCCS tests.

Repeatability between Experiments I and IIa. Since we had the opportunity to retest eleven of our subjects, we were able to generate the graphs below. Naturally, these subjects were a few months older during the retest than they were during the original assessments, which may introduce some variability in test vs. retest findings. Figures 31-

33 demonstrate the value scored for the first eye tested next to the second eye(s) examined.

3.1 2.8 2.5 B-L Acuity 2.2 1.9 1.6 1.3 1 0.7 0.4 0.1 ** **

LogCS LogMAR Poor LogMAR LogCS ** ** -0.2 -0.5 -0.8 -1.1 -1.4 -1.7 -2 -2.3

Good -2.6 -2.9 P-R Contrast Sensitivity -3.2 ** Unable

Figure 31. Experiment IIa Lettered Chart Test-Retest Retest values for the better eyes of subjects who participated in Experiments I and IIa. Conventions are similar to those in Figure 13. The fist test is shown to the left, and the repeat assessment follows to the right.

3.1 2.8 2.5 TAC Acuity 2.2 1.9 1.6 1.3 1 0.7 0.4 0.1 LogCS LogMAR Poor LogMAR LogCS -0.2 -0.5 -0.8 -1.1 -1.4 -1.7 -2 -2.3

Good -2.6 SCCS Contrast Sensitiviy -2.9 -3.2

Figure 32. Experiment IIa Striped Chart Test-Retest Retest values for the better eyes of subjects who participated in Experiments I and IIa. Conventions are similar to those in Figure 13. The first test is shown to the left, and the repeat assessment follows to the right. The SCCS was limited by a ceiling of 1.65 during Experiment I tests, but by 2.00 for Experiment IIa.

3.1 2.8 Average B-L and TAC Acuity 2.5 2.2 1.9 1.6 1.3 1 0.7 0.4 0.1 LogCS LogMAR Poor LogMAR LogCS -0.2 -0.5 -0.8 -1.1 -1.4 -1.7 -2 -2.3

Good -2.6 BD Contrast Sensitivity -2.9 -3.2

Figure 33. Experiment IIa Shaped Chart Test-Retest Retest values for the BD test shown with the average of each experiment’s acuity measurements as in Figure 15.

T-test comparisons demonstrate that the repeatability of our testing was good with these subjects’ better eyes (Table 6).

B-L TAC P-R SCCS BD

p-value 2.11 1.20 1.90 3.60 0.121

Sig (<0.05) Not different Not different Not different Not different Not different

Table 6. Experiment IIa Repeatability Statistics

Results for Experiment IIb: New Subjects

Eight participants were students in the annual summer camps put on by OSSB.

Their subject ID numbers were #62-69. Ten of the subjects listed in the table below had no measurable vision, but were able to provide quality of life data.

B-L TAC P-R SCCS BD ID Age S Diagnosis Eye LogMAR LogMAR LogCS LogCS LogCS 31 16 M Retinoblastoma OD 0.98 0.68 1.75 2.00 1.65 54† 20 M Retinopathy of Prematurity — — — — — — 55† 19 F Leber's Congenital Amaurosis — — — — — — 56† 19 M Brain Tumor — — — — — — 57† 19 M Optic Nerve Hypoplasia — — — — — — 58 18 F Leber's Congenital Amaurosis OD 2.18 0.50 0.95 1.65 1.65 58 18 F Leber's Congenital Amaurosis OS 2.10 0.68 0.95 1.65 1.65 59† 20 M Retinopathy of Prematurity — — — — — — 60 18 M Retinopathy of Prematurity OD 1.56 0.99 1.00 1.50 0.6 6†0 18 M Retinopathy of Prematurity OS — 1.29 — 0.60 0.6 61 13 M Choroideremia OD 1.54 0.99 0.15 0.90 0.45 61 13 M Choroideremia OS 1.18 0.81 0.25 1.05 0.45 62 11 F Blind at age 2 (unexplained) OD 1.58 — 0.15 0.00 0.00 62 11 F Blind at age 2 (unexplained) OS 1.38 1.38 0.15 0.00 0.00 63 10 F Leber's Congenital Amaurosis OD 1.14 0.81 0.85 1.65 1.65 63 10 F Leber's Congenital Amaurosis OS 1.04 0.99 1.00 1.65 1.5 64 13 M Optic Atrophy OS 1.18 1.11 0.80 0.60 0.45 65 16 F Marfan's Disease OD 0.28 0.38 1.55 1.65 1.65 65 16 F Marfan's Disease OS 0.68 0.50 1.55 1.50 1.65 66 17 M High Myopia OD 0.08 0.21 2.00 1.50 1.65 67 10 M Cortical Blindness OS 1.79 0.99 0.90 1.20 1.35 68 12 M Aniridia OD 1.94 1.38 0.35 1.35 1.05 68 12 M Aniridia OS 1.04 0.68 1.80 2.00 1.65 69 14 M Retinopathy of Prematurity OD 0.98 0.99 1.40 1.90 1.65 69 14 M Retinopathy of Prematurity OS 0.98 0.68 1.55 2.00 1.65 70† 18 M Retinitis Pigmentosa — — — — — — 71 15 M Best's Disease OD 0.86 0.38 1.65 2.00 1.65 71 15 M Best's Disease OS 0.84 0.38 1.65 2.00 1.65 72 18 F Glaucoma OS 1.04 0.99 1.05 1.35 1.05 73† 17 F Optic Atrophy — — — — — — 74† 12 M Optic Nerve Hypoplasia — — — — — — 75† 12 M Retinopathy of Prematurity — — — — — — 76† 6 M Retinoblastoma — — — — — — Table 7. Experiment IIb Participants n = 34 (23 eyes tested) †These subjects were not testable via lettered charts.

Result values of all test encounters with the new subjects are summarized in the table below along with average testing times per eye. All but one student contributed two eyes to these results.

B-L TAC P-R SCCS* BD†

Mean 1.20 (20/314) 0.79 (20/121) 1.07 1.41 1.65 Median

STDev ± 0.53 ± 0.32 ± 0.59 ± 0.59 0.71 75th %

Time (sec) 66 ± 45 86 ± 53 50 ± 26 49 ± 27 60 ± 39 Time (sec)

Table 8. Experiment IIb Summary Results

Two-tailed t-test comparisons of findings for the acuity and contrast sensitivity measures are shown below.

2. 5

p < 0.001** 2 1. 5 1 0. 5 B -L $ TA C 0 P- R SC C S * BD † -0 . 5 -1 -1 . 5 -2 p < 0.01** p < 0.71 -2 . 5 p < 0.001**

Figure 34. Experiment IIb Summary Plot Statistics Conventions are as in Figure 23.

The figures below are test results for new students who were added to the study in the second year of testing and were not previously assessed.

3 Letters Better 2.5

1.5 Stripes Better TAC LogMAR TAC 1

R² = 0.1467 0.5

0 0 1 2 3 B-L LogMAR

B 3 2.4

L 1.8 -

B 1.2 0.6 – 0 -0.6 -1.2 TAC TAC -1.8 -2.4 -3 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3 Average LogMAR B-L & TAC

Figure 35. Experiment IIb Acuity Results Comparison of acuity measured by TAC and B-L for both eyes of new subjects during our second year of testing. All but two subjects contribute two eyes. Conventions are as in Figure 12. A, Direct comparison, B, Bland-Altman plot.

-0.15 Letters Better -0.45

-0.75

-1.05 R² = 0.6142

-1.35 SCCS LogCS SCCS

-1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

1.95 R

- 1.35

P 0.75 – 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS P-R & SCCS

Figure 36. Experiment IIb P-R & SCCS Results Comparison of contrast sensitivity values obtained by SCCS and P-R charts for new subjects. All but two subjects contribute two eyes. Conventions are as in Figure 12. A, Direct comparison. B, Bland-Altman plot.

SCCS 3 P-R

NumberSubjects of 2

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 37. Experiment IIb P-R & SCCS Bins Number of subjects at each contrast sensitivity level for the SCCS and P-R testing of new subjects. The SCCS is bounded by a ceiling of LogCS 2.00.

-0.15 Letters Better -0.45

-0.75 R² = 0.6582 -1.05

BD BD LogCS -1.35

-1.65

-1.95 Discs Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Pelli-Robson LogCS

B 1.95

1.35 R - 0.75 P 0.15 – -0.45

BD -1.05 -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS P-R & BD

Figure 38. Experiment IIb P-R & BD Results Comparison of contrast sensitivity testing by BD and P-R chart for new subjects. The BD test is bounded by a ceiling of 1.65. A, Direct comparison. B, Bland-Altman plot.

BD 6 P-R

NumberSubjects of 4

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 39. Experiment IIb P-R & BD Bins Number of subjects at each contrast sensitivity level for testing of new subjects with the BD and P-R chart. The BD is bounded by a ceiling of 1.65.

-0.15 Discs Better -0.45

-0.75

-1.05 R² = 0.8134

-1.35 SCCS LogCS SCCS

-1.65

-1.95 Stripes Better -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 BD LogCS

B 1.95 1.35 BD 0.75 – 0.15 -0.45 -1.05 SCCS SCCS -1.65 -2.25 -2.25 -1.95 -1.65 -1.35 -1.05 -0.75 -0.45 -0.15 Average LogCS SCCS & BD

Figure 40. Experiment IIb SCCS & BD Results Comparison of contrast sensitivity testing by BD and SCCS for new subjects. The BD is bounded by a 1.65 ceiling and the SCCS test by 2.00. Many subjects obtained the maximum values for each chart leading to their data points plotting to the same location on the graphs. This gives the appearance of less data than actually obtained. A, Direct comparison. B, Bland-Altman Plot.

BD 6 SCCS

NumberSubjects of 4

0 2.1 1.95 1.8 1.65 1.5 1.35 1.2 1.05 0.9 0.75 0.6 0.45 0.3 0.15 0 LogCS

Figure 41. Experiment IIb SCCS & BD Bins Number of new subjects at each contrast sensitivity level for the BD and SCCS tests. The BD is bounded by a ceiling of 1.65, and the SCCS has a LogCS ceiling of 2.00.

Discussion for Experiment II

Some of the data that we have obtained through the efforts of Experiment I can be added to the better eye results from Experiment II. In order to check correlations between the five eye charts, it was necessary to generate a table containing test results for each subject using only one eye—the better one. In the case of students tested only in

Experiment I, the results on the only tested eye appear. In the case of students tested in both Experiment I and Experiment II, the results on the better eye in Experiment II are listed. In the case of students tested only in Experiment II, the results on the better eye are listed.

B-L TAC P-R SCCS BD ID Exp Diagnosis Eye LogMAR LogMAR LogCS LogCS LogCS 01 IIa Optic Nerve Hypoplasia OS 1.32 0.68 1.20 2.00 1.05 02 I Retinopathy of Prematurity OD 0.62 0.21 1.95 1.65 1.65 04 I Optic Nerve Hypoplasia OS 0.48 0.08 2.10 1.65 1.65 05 I Leber's Congenital Amaurosis OS 1.42 1.56 1.10 0.60 1.50 06 I Optic Nerve Hypoplasia OD 0.74 0.38 2.00 1.65 1.65 08 I Optic Atrophy OD 0.88 0.21 1.45 1.65 1.50 11 I High Myopia OD 0.94 0.99 1.35 1.65 1.65 12 IIa Retinopathy of Prematurity OS 0.68 0.68 1.90 2.00 1.65 17 IIa Microphthalmus OS 1.36 1.29 1.35 1.90 1.50 18 I Cone Dystrophy OS 0.7 0.81 1.40 1.65 1.65 24 I Optic Atrophy OS 1.16 0.5 1.65 1.65 1.35 26 IIa Optic Nerve Hypoplasia OS 0.96 0.38 2.05 2.00 1.65 28 IIa Congenital Cataract and Aphakia OD 1.28 0.68 1.75 2.00 1.65 30 IIa Optic Atrophy OS 0.94 0.99 1.05 1.65 1.65 31 IIb Retinoblastoma OD 0.98 0.68 1.75 2.00 1.65 34 I Congenital Cataract and Aphakia OD 0.22 0.21 1.65 1.65 1.50 35 I Retinopathy of Prematurity OD 0.82 0.68 1.75 1.65 1.65 36 I Optic Atrophy OS 1.6 0.81 0.90 1.65 1.65 37 IIa Aniridia OD 0.96 0.38 1.50 1.65 1.65 41 I Glaucoma OS 1.96 2.45 0.10 1.20 0.90 45 IIa Optic Atrophy OD — 1.29 — 0.75 0.00 46 I Cortical Blindness OD — 0.84 — 1.35 0.75 48 I Optic Atrophy OD 1.2 0.68 1.15 1.65 1.50 50 I Septo-Optic Dysplasia OS 1.9 0.68 0.70 1.50 0.30 51 IIa Optic Atrophy OS 1.14 0.5 1.20 1.65 1.20 52 IIa Optic Atrophy OS 2.3 0.68 0.50 1.50 0.75 53 IIa Cortical Blindness OD 3 0.38 0.70 1.90 1.50 58 IIb Leber's Congenital Amaurosis OS 2.1 0.68 0.95 1.65 1.65 60 IIb Retinopathy of Prematurity OD 1.56 0.99 1.00 1.50 0.60 61 IIb Choroideremia OS 1.18 0.81 0.25 1.05 0.45 62 IIb Blind at age 2 (unsure) OS 1.38 1.38 0.15 0.00 0.00 63 IIb Leber's Congenital Amaurosis OS 1.04 0.99 1.00 1.65 1.50 64 IIb Optic Atrophy OS 1.18 1.11 0.80 0.60 0.45 65 IIb Marfan's Disease OD 0.28 0.38 1.55 1.65 1.65 66 IIb High Myopia OD 0.08 0.21 2.00 1.50 1.65 67 IIb Cortical Blindness OS 1.79 0.99 0.90 1.20 1.35 68 IIb Aniridia OS 1.04 0.68 1.80 2.00 1.65 69 IIb Retinopathy of Prematurity OS 0.98 0.68 1.55 2.00 1.65 71 IIb Best’s Disease OD 0.86 0.38 1.65 2.00 1.65 72 IIb Glaucoma OS 1.04 0.99 1.05 1.35 1.05 Table 9. Experiment I & II Better Eye Only n = 36 Subjects (36 eyes) Note: Subject 43 was not included here as he did not provide any QoL data.

This grouping of subjects allows for us to check for correlations across all the tests we employed.

B-L TAC P-R SCCS* BD†

Mean 1.16 (20/287) 0.75 (20/112) 1.30 1.55 1.50 Median

STDev ± 0.58 ± 0.44 ± 0.54 ± 0.44 1.05 75th %

Time (sec) 62 ± 37 86 ± 60 61 ± 38 56 ± 32 47 ± 31 Time (sec)

Table 10. Experiment II Summary Results

2. 5 2 p < 0.05* 1. 5 1 0. 5 B -L TA C 0 P- R SC C S * BD † -0 . 5 -1 -1 . 5

-2 p < 0.001** p < 0.001** -2 . 5 p < 0.001**

Figure 42. Experiment II Summary Plot Statistics Conventions are as in Figure 34. Here only one eye per subject is included—the better eye (or second test if repeat data are available). Subject number 43 is not included in this analysis since he did not provide quality of life data. Note: if the calibrated P-R values are compared to the SCCS, then p < 0.01.

When comparing results across the better eye for all subjects, all tests were positively correlated with one another. These correlations were statistically significant in all but one case—the B-L and the SCCS were not significantly correlated. As can be seen in the table below, the B-L chart was most strongly correlated with the P-R, perhaps because it was only other letter chart. The next strongest, correlation was with the TAC, perhaps due to the fact that they are both acuity tests. Next, the BD showed a significant correlation with the B-L chart, perhaps because both tests include a visual search localization component where the stimulus to be detected or identified must be localized.

For the P-R chart, the strongest correlation was found with the BD test, followed by the B-L chart, TAC and finally the SCCS. The SCCS was most strongly correlated with the BD, followed by P-R, and then TAC.

PRLogCS TACLogMAR SCCSLogCS BDLogCS O&MScore IVI_C LVP-FVQ

Pearson Correlation .692** .401* .149 .389* .540 -.197 -.594** BLLogMAR Sig. (2-tailed) .000 .013 .372 .016 .057 .237 .000 N 38 38 38 38 13 38 33 Pearson Correlation .653** .608** .738** .166 -.481** -.738**

PRLogCS Sig. (2-tailed) .000 .000 .000 .588 .002 .000 N 38 38 38 13 38 33 Pearson Correlation .539** .450** .219 -.367* -.597**

88 TACLogMAR Sig. (2-tailed) .000 .004 .453 .020 .000

N 40 40 14 40 35 Pearson Correlation .708** .165 -.576** -.512**

SCCSLogCS Sig. (2-tailed) .000 .574 .000 .002 N 40 14 40 35 Pearson Correlation .450 -.434** -.584** .106 .005 .000 BDLogCS Sig. (2-tailed)

N 14 50 45 Table 11: Test Chart Correlations

**. Correlation was very significant: at the 0.01 level (2-tailed). *. Correlation was significant: at the 0.05 level (2-tailed). Note: The table has been shaded to help the reader quickly visualize significant correlations (or the absence of correlation). Subject 43 was not included in this analysis since he did not provide any QoL or O&M data. Table 13 Continued

Experiment III – Quality of Life and Orientation and Mobility

Vision-Related Quality of Life

Quality of life surveys were administered following vision testing for each subject. In order to facilitate testing, we modified the surveys so that they would contain appropriate phrasing (e.g., for one LVP-FVQ question: switch the word for a “rupee” coin to a “nickel”). Copies of the verbal scripts actually used are in Appendix A. The first five subjects tested (#01, 02, 05, 12, and 34) did not have the opportunity to complete the

LVP-FVQ, but did complete the IVI_C questionnaire. When both tests were administered, the IVI_C was always administered before the LVP-FVQ. Although two examiners were used, each participant was interviewed by the same examiner for both surveys. Angela Brown administered 38 out of 50 IVI_C surveys (76%) and Greg

Hopkins administered the remainder.

We used a 26 item IVI_C survey, which included two questions that were dropped when the IVI_C underwent Rasch analysis because they only received answers in 4 out of 5 response categories (“Do you get the same information as the other students?” and “…Are you confident about asking for help that you need?”).Our research was initiated prior to the publication of the revised LVP-FVQ II survey, so we did not have the opportunity to apply the new version.

Bradley Dougherty performed Rasch analysis on the results from the two surveys that we used. The analysis generated person scores (and standard errors around those scores) for each subject relative to the others who took the surveys. The Rasch analysis was performed using all five-category responses, and the response option probability curves were not ideal. Further work may require a three or four category structure.

Testing time for the IVI_C was 8 ± 9 minutes. Linear regressions for the IVI_C versus the results for each eye chart are below. Error bars represent the standard error value around the person scores for each participant. There were two subjects (marked with the ∆ symbol), who could not perform letter acuity or contrast tasks, but were testable with other charts. Subjects without measurable vision were spread apart on the x- axis systematically for acuity and contrast plots.

3.00

2.50

2.00

1.50

1.00

0.50

IVI_C IVI_C Person Score R² = 0.0387 0.00

-0.50

-1.00 0 0.5 1 1.5 2 2.5 3 3.5 4 B-L LogMAR Acuity

Figure 43. IVI_C v B-L Regression Data are for the better eye of all subjects with measurable vision. Subjects with no measurable vision are plotted separately using open circles (not included in the regression trend line) by adding 0.25 LogMAR to the maximum value for partially-sighted subjects and then systematically adding 0.01 LogMAR to shift each point along the x-axis in order of subject number. Subjects marked with the ∆ symbol could not perform letter tasks. Note: The B-L Chart was the only vision test not significantly correlated with the IVI_C results.

3.00

2.50

2.00

1.50 R² = 0.2312

1.00

0.50 IVI_C IVI_C Person Score 0.00

-0.50

-1.00 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 P-R LogCS

Figure 44. IVI_C vs. P-R Regression Data are for the better eye of all subjects with measurable vision. Subjects with no measurable vision are plotted separately (not included in the regression trend line) by starting at positive 0.25 LogCS and then systematically shifting subjects along the x-axis by adding 0.0125 logCS in order of subject ID number.

3.00

2.50

2.00

1.50

1.00 R² = 0.1343

0.50 IVI_C IVI_C Person Score 0.00

-0.50

-1.00 0 0.5 1 1.5 2 2.5 3 3.5 4 TAC LogMAR Acuity

Figure 45. IVI_C vs. TAC Regression Data are for all subjects with measurable vision. Two subjects on this plot are included that could not be measured via lettered charts. Their points have been marked with a ∆ symbol. Conventions for this plot are the same as for Figure 43. Note: The TAC test was significantly correlated with the IVI_C only at the 0.05 level (2- tailed).

3.00

2.50

2.00

1.50

1.00

0.50

R² = 0.3313 IVI_C IVI_C Person Score

0.00

-0.50

-1.00 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 SCCS LogCS

Figure 46. IVI_C vs. SCCS Regression Data are for the better eye of all subjects with measurable vision. Two subjects on this plot are included that could not be measured via lettered charts. Their points have been marked with a ∆ symbol. Conventions for this plot are the same as for Figure 44.

3.00

2.50

2.00

1.50

R² = 0.1884 1.00

0.50 IVI_C IVI_C Person Score

0.00

-0.50

-1.00 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 BDLogCS

Figure 47. IVI_C vs. BD Regression Data are for the better eye of all subjects with measurable vision. Two subjects on this plot are included that could not be measured via lettered charts. Their points have been marked with a ∆ symbol. Conventions for this plot are the same as for Figure 44.

Testing time for the LVP-FVQ was 6 ± 2 minutes. Linear regressions for the

LVP-FVQ versus the results for each eye chart are below. Again, there were two subjects

(marked with the ∆ symbol), who could not perform letter acuity or contrast tasks, but were testable with other charts.

R² = 0.3525

FVQ FVQ Person Score -

0 LVP

-1

-2 0 0.5 1 1.5 2 2.5 3 3.5 4

B-L LogMAR Acuity

Figure 48. LVP-FVQ vs. B-L Regression Data are for the better eye of all subjects with measurable vision. Conventions are the same as for Figure 43.

FVQ FVQ Person Score -

LVP 0

R² = 0.5447 -1

-2 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 P-R LogCS

Figure 49. LVP-FVQ vs. P-R Regression Data are for the better eye of all subjects with measurable vision. Conventions are the same as for figure 44.

R² = 0.3569

0 LVP_FVQ LVP_FVQ Person Score

-1

-2 0 0.5 1 1.5 2 2.5 3 3.5 4 TAC LogMAR Acuity

Figure 50. LVP-FVQ vs. TAC Regression Data are for the better eye of all subjects with measurable vision. Conventions are the same as for figure 43.

FVQ FVQ Person Score -

0 LVP

-1 R² = 0.2623

-2 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 SCCS LogCS

Figure 51. LVP-FVQ vs. SCCS Regression Data are for the better eye of all subjects with measurable vision. Conventions are the same as for figure 44.

FVQ FVQ Person Score -

LVP 0

-1 R² = 0.3407

-2 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 BDLogCS

Figure 52. LVP-FVQ vs. BD Regression Data are for the better eye of all subjects with measurable vision. Conventions are the same as for figure 44.

Orientation and Mobility

We were able to obtain orientation and mobility data for nineteen of our subjects, seven of whom had no measurable vision. We obtained the severity-of-need scores for each student and modified the values by subtracting out the score for central visual acuity. All other assessments by the orientation and mobility instructors (including peripheral vision) were maintained in the score.

Linear regression for modified O&M severity-of-need score is displayed versus the results for each eye chart. Those subjects without any vision have their O&M scores displayed separately from those subjects with measurable vision. There was one subject

100

(marked with a ∆ symbol), who could not perform letter acuity or contrast tasks, but was testable with other charts. Subjects without measurable vision were spread apart on the x- axis systematically for acuity and contrast plots. For the acuity plots, subjects with no measurable vision begin at LogMAR 3.0 and then an additional 0.01 LogMAR was added in the order of subject ID number. For the contrast sensitivity plots, subjects with no measurable vision are shown on the positive side of the x-axis starting at zero and adding 0.0125 logCS to each subject in order of ID number. All of the O&M regressions below were appropriately related to the vision testing results, but none of them reach statistical significance.

15 O&M Score 10 R² = 0.2916

0 0 0.5 1 1.5 2 2.5 3 3.5 B-L LogMAR Acuity

Figure 53. O&M vs. B-L Regression Data are for the better eye of subjects with measurable vision. Subjects with no measurable vision are shown separately (not included in the regression trend line) and spread out systematically on the x-axis. One subject who could not complete letter charts is marked with a ∆ symbol.

101

O&M Score R² = 0.0275 10

0 -2.5 -2 -1.5 -1 -0.5 0 0.5 P-R LogCS

Figure 54. O&M vs. P-R Regression Data are for the better eye of subjects with measurable vision. Subjects with no measurable vision are shown separately. One subject who could not complete lettered charts is marked with a ∆ symbol.

102

15 O&M Score 10 R² = 0.0478

0 0 0.5 1 1.5 2 2.5 3 3.5 TAC LogMAR Acuity

Figure 55. O&M vs. TAC Regression Data are for the better eye of subjects with measurable vision. One subject is marked with a ∆ symbol because that subject could not read lettered charts.

103

15 O&MScore 10 R² = 0.0271

0 -2.5 -2 -1.5 -1 -0.5 0 0.5 SCCS LogCS

Figure 56. O&M vs. SCCS Regression Data are for the better eye of subjects with measurable vision. One subject (marked with a ∆ symbol) could not complete lettered charts.

R² = 0.2027 O&M Score 10

0 -2.5 -2 -1.5 -1 -0.5 0 0.5 BDLogCS

Figure 57. O&M vs. BD Regression Data are for the better eye of subjects with measurable vision. One subject is marked with a ∆ symbol. This subject was only able to see the first circle on the BD test (LogCS 0.00).

104

Correlations for vision-related quality of life and orientation & mobility and our vision test results are summarized in Table 15, where only the better eye for subjects with measureable vision are included. Generally, the LVP-FVQ correlated best with our vision testing, followed by the IVI_C and then the adjusted OMSRS scores (which do not include central visual acuity). As expected, the two QoL surveys correlated positively and highly significantly with one another. Taken as a whole, our contrast sensitivity testing correlated with a higher significance level to the QoL metrics than our visual acuity testing did. The IVI_C correlated appropriately with all vision tests, demonstrating that higher ability scores match with lower LogMAR and better LogCS values. The correlation between the IVI_C was highly significant with all charts except the TAC test

(2-tailed 0.05 level only). The B-L acuity chart was not significantly correlated with the

IVI_C. In addition, the LVP-FVQ correlated very significantly and appropriately with all vision testing.

We adjusted the OMSRS data by removing “central acuity” from the total score.

It appears that the relationship between the adjusted O&M scores was appropriate for our vision testing measures, where worse visual performance was indicative of greater need for services and a higher score. Likewise, a higher O&M score related to a lower visual ability level on both of our QoL surveys. However, the adjusted OMSRS data correlated only with the LVP-FVQ, and only at the 0.05 level (2-tailed). Interestingly, if we removed all vision data from the OMSRS, then the only significant correlation was with

B-L, and only at the 0.05 level (2-tailed). Otherwise, the O&M data we have obtained did not correlate significantly with any of our vision tests. Clearly, our small sample size (13 105 subjects for letter charts and 14 for shaped charts) affected our power to detect statistically significant relationships.

106

O&MScore IVI_C LVP- BLLogMAR PRLogCS TACLogMAR SCCSLogCS BDLogCS FVQ

Pearson -.372 .651** -.197 -.481** -.367* -.576** -.434** Correlation IVI_C Sig. (2-tailed) .097 .000 .237 .002 .020 .000 .005 N 21 45 38 38 40 40 40

107 Pearson -.573* .651** -.594** -.738** -.597** -.512** -.584**

Correlation LVP-FVQ Sig. (2-tailed) .013 .000 .000 .000 .000 .002 .000 N 18 45 33 33 35 35 35 Pearson -.372 -.573* .540 .166 .219 .165 .450 Correlation O&MScore Sig. (2-tailed) .097 .013 .057 .588 .453 .574 .106 N 21 18 13 13 14 14 14 Table 12. QoL and O&M Correlations

**. Correlation was very significant: at the 0.01 level (2-tailed). *. Correlation was significant: at the 0.05 level (2-tailed). Note: conventions for this table are the same as for Table 14.

Discussion for Experiment III

Vision-Related Quality of Life. The IVI_C and LVP-FVQ can be administered to students at OSSB, and it appears that scores were positively correlated with measures of vision. However, some of the IVI_C content may not be well suited to students in these specialized settings. Rehabilitation and classroom adaptations at OSSB are quite good. It appears that the LVP-FVQ better targets the patient population at OSSB, as the mean ability score for the survey participants was more closely aligned with the mean item difficulty score of the survey.

It could be that some portion of the disability paradox was in play here: people who are visually impaired place a higher value on vision than those who are not, yet report better quality of life than expected (Albrecht & Devlieger, 1999). Perhaps with the

IVI_C (which contains social questions) we were not truly measuring functional reserve but instead, overall happiness or some other construct. The high results we obtained on the IVI_C could be due to the acclimation to vision loss that the student population at

OSSB exhibits. For instance, the least difficult question on the IVI_C by Rasch analysis was: “Do your teachers understand your special needs?” Most people want to be happy, and if they can find a way to adapt, they will be. The functional nature of the LVP-FVQ may be better suited to assessing these students. Other measures like classroom productivity may be appropriate as well.

Eight of our subjects were not OSSB students, but rather, were at OSSB for summer camps. Their average person scores from the two surveys are shown in Figure 58

108 below. Generally, OSSB students with vision performed the best on the IVI_C. This could be due to the fact that they are well adapted in their school environment. The summer campers out performed OSSB students on the LVP-FVQ, perhaps due to the fact that the LVP-FVQ is more functionally oriented, and these students function in mainstream schools. In both cases, the OSSB students without measurable vision perform the worst, and the difference is more dramatic on the LVP-FVQ. Error bars indicate the standard error of measurement for each group. There were two blind students on the

LVP-FVQ who perceived their abilities to be quite high, resulting in a large range of measurement for that group. One-way ANOVA analysis revealed a significant difference between the subject groups (OSSB Partially-Sighted, Functionally Blind, and Summer

Campers) on the IVI_C only, but the effect was not confirmed by post-hoc analysis.

109

Figure 58. Average person scores for the IVI_C (left) and the LVP-FVQ (right) Data are separated out by vision level and enrolment at OSSB.

The location of our summer school students relative to those other students who took the two surveys can be seen on the subject-item maps in Figures 59 and 60. These figures back up the assertion that summer campers generally perform better on the LVP-

FVQ than OSSB partially-sighted students and worse on the IVI_C.

It is also apparent from Figures 59 and 60 that the subject mean IVI_C is well above the item mean. This confirms the fact that the IVI_C is perhaps more appropriate for visually impaired students at mainstream schools and not at specialized school for the blind settings.

110

Figure 60 demonstrates that the LVP-FVQ is better targeted for the subjects we assessed, but our results do not follow a normal distribution. It is possible that increasing the number of subjects would improve this distribution.

111

Figure 59. Subject-Item map for the IVI_C Demonstrates Rasch-calibrated participant locations (x’s, left) and item locations (right). At the top of the map are participants with higher perceived visual ability and questions thought to be more difficult. This map demonstrates the targeting of the different response levels of each question to the respondents. Summer school students have been marked with boxes. M, Mean; S, 1 SD from the mean; T, 2 SD from the mean.

112

Figure 60. Subject-Item map for the LVP-FVQ. Conventions are the same as for Figure 59.

Linear regression between the Rasch-adjusted person scores for each individual survey participant is demonstrated in Figure 61. It is clear that two functionally blind subjects rated their perceived functional ability to be quite high on the LVP-FVQ, yet fell in the middle of the pack (close to the y-axis) for the IVI_C. Likewise, there was one subject in the functionally blind category on the IVI_C whose perceived visual ability level was higher than the others in that group. His or her point can be found outlying the trend line just above the abscissa.

113

R² = 0.2436

FVQ FVQ PersonScore

- LVP 0

-1

IVI_C Person Score -2 -1 0 1 2 3

Figure 61. Person Score Linear Regression of LVP-FVQ vs. IVI_C Subjects with no measurable vision (or unable to perform lettered charts) are plotted with their standard errors using the same symbol conventions as in Figure 39. All subjects are included in the regression in this case.

Orientation and Mobility. It does appear that, generally, subjects with measurable vision receive a lower severity of need score. Thus, they need less time devoted to orientation and mobility training. However, B-L acuity was the only factor whose correlation reaches a level of statistical significance. This was despite the fact that we adjusted the scores so as not to include B-L acuity. This may be due to the fact that

114 the O&M teachers are aware of the child’s acuity and this awareness colors their impressions for other items on the OMSRS.

We were only able to obtain a relatively small set of data for orientation and mobility severity of need scores. This may affect our power to detect significant statistical correlations.

115

General Discussion

Test Results

Our results in the tables above show that not all participants were able to complete every aspect of the study protocol. Some participants (who were functionally blind) were able to provide only quality of life questionnaire responses. We included those participants for the purposes of balancing our Rasch analysis. We wanted to see if the vision-related quality of life of functionally blind students tested as the same as those with vision. It appears that they did not—students who lack measurable vision generally clustered together at lower ability levels than sighted students. They were also generally allocated more training time for orientation and mobility than their partially-sighted classmates.

All vision tests were significantly correlated with one another except the B-L and

SCCS. This makes sense because the two charts are the least alike in terms of measurement outcome and test methodology. As we hoped, we were able to measure a variety of contrast sensitivity and acuity levels in this student population. However, ceiling effects did come into play for some of our contrast sensitivity tests. Testing with the striped cards (TAC and SCCS) cards consistently produced better results than the lettered charts: B-L and P-R. With the SCCS cards, the majority of our subjects hit the

1.65 LogCS ceiling during our first year of testing. When we raised the ceiling to LogCS

2.00 for our second round of testing, again, the majority of subjects were able to see the

116 stripes on that card. The distribution of results for the P-R chart was much more stratified—subjects who performed worse on the P-R chart gathered together with high performing P-R subjects on the SCCS and BD test result graphs.

The Berkeley disc results were also better than the P-R test results, but were limited by a 1.65 LogCS ceiling effect at all phases of our trial. The simpler task and bolder patterns of the SCCS and the Berkeley Discs may make those tests more likely to reveal the maximum performance that a given patient can achieve, even with the ceiling in place. From the results of testing with a higher ceiling using the SCCS, it appears that at least half of the subjects measured at 1.65 LogCS on the Berkeley Discs might have been capable of 2.00 LogCS. These results are likely to be better transferrable to some activities, such as O&M class, life skills, and gym. Performance in other courses, such as

English, Math, and Science may be better predicted by the results of our letter identification tests.

Other Considerations

We did not screen for VF defects, which could affect a patient’s ability to locate the discs within the grid or find the stimulus on the SCCS test. Based on my experience testing these children, I believe that visual field status was only a factor during the testing of three subjects: #5, #41 and #48.

Some subjects exhibited interesting behavior such as apparent malingering on the

Berkeley Discs and SCCS tests. One subject reported detecting our stimulus with his or her “worse” eye on the blank half of the SCCS test 100% of the time (much higher than chance) starting at the contrast threshold of the fellow and continuing all the way down to

117 the lowest contrast card. Perhaps he or she was unaware of the fact that intentionally selecting the area absent a stimulus reveals the ability to discern stimulus location. On the

Berkeley Discs, several other subjects incorrectly reported discs in areas where none were present. Perhaps this was in an attempt in their minds to try and get as many stimuli identified as possible in an effort to boost their hit rate. It has been shown by other investigators that children are much more prone to tolerating misses when reaching their measurement thresholds, i.e., they’re willing to guess (Susan J Leat & Wegmann, 2004).

It is our opinion that these children were not intending to malinger, but instead were experiencing false-positives throughout the entire test. This is because, given the design of the Berkeley Discs, their threshold stimulus may be present in any cell on the grid.

Stated Objectives

Our goal was to ascertain whether the Stripe Card Contrast Sensitivity (SCCS)

Test was easy to administer and whether it could be successfully used on a wide variety of participants. I believe that the test performed very well in this patient population and would be appropriate to use in pediatric low vision settings whenever grating acuity measurements are indicated.

We also wished to assess the SCCS test’s capability for validly quantifying contrast sensitivity levels of children with impaired vision. It appears that this chart has no trouble detecting the very high levels of contrast sensitivity performance that some children with visual impairment are capable of. This is important because, while contrast sensitivity measurement isn’t more useful than other standard examination procedures at detecting disease, it does add valuable information regarding a patient’s expected level of

118 visual function. It seems that this patient population suffers less contrast impairment than those who have visual impairment due to diseases more prevalent in the elderly such as macular degeneration and diabetic retinopathy. However, we have not studied the SCCS outside of the pediatric population as of yet.

Certainly, measurements from the SCCS test cannot be easily decimalized or broken down in to smaller increments (such as LogCS 0.05) as is possible using letter-by- letter scoring on the Pelli-Robson chart. This may affect the repeatability of the SCCS.

However, the target population for the SCCS is likely one where measurement by means such as letter identification tasks simply isn’t possible. Test-retest variance would be expected to be high in this group of patients for other functional reasons (such as autism or developmental delay) and accuracy to the nearest 0.05 LogCS would not be a consideration for most clinical examiners. It is likely that SCCS test results translate best to daily living tasks for children that rely on low and middle spatial frequencies only such as mobility, household, self-care, sports, and social activities. It may be useful to run the

SCCS test whenever input about a given student’s performance potential on these types of tasks is a consideration for parents or educators.

I have calculated the SCCS contrast sensitivity values necessary for a subject to have a vision-related quality of life score at the level found for the 75th percentile of our functionally blind participants for both the IVI_C and the LVP-FVQ (See Figure 62).

Their contrast sensitivity would have to be approximately LogCS 1.02 for both the IVI_C and the LVP-FVQ.

119

3.00 A 2.50

2.00

1.50

1.00

IVI_C IVI_C Person Score 0.50

0.00 R² = 0.3313 -0.50

-1.00 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 SCCS LogCS

4 B 3

R² = 0.2623

FVQ FVQ Person Score -

LVP 0

-1

-2 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 SCCS LogCS

Figure 62. Cutoffs for normal SCCS Performance The 75th percentile vision-related quality of life score was calculated for the students without measurable vision and extended horizontally across the graph. The intersection point with the trend line was drawn vertically so that the LogCS value on the abscissa demonstrates the expected contrast sensitivity level for partially-sighted students with the same quality of life score. A, cutoff point for the IVI_C. B, cutoff point for the LVP-FVQ.

120

This means that successfully localizing the striped pattern on the SCCS test for the LogCS 0.90 card (or worse) would be a reasonable cutoff indicative of performance that is definitely abnormal. Indeed, this value corresponds well with cutoffs for significantly impaired contrast sensitivity found in the literature and described in earlier sections of this thesis (S J Leat & Woo, 1997; Susan J Leat et al., 1999; West et al.,

2002).

The findings in this section, taken together, demonstrate the Stripe Card Contrast

Sensitivity test to be a useful clinical measure for young, non-verbal or otherwise developmentally delayed patients with visual impairment.

121

References

Adams, R. J., & Courage, M. L. (2003). Can the visual acuity of infants be predicted from a measurement of contrast sensitivity? J Pediatr Ophthalmol Strabismus, 40(1), 35–38. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt= Citation&list_uids=12580270

Albrecht, G. L., & Devlieger, P. J. (1999). The disability paradox: high quality of life against all odds. Social Science & Medicine (1982), 48(8), 977–88. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10390038

Arden, G. B. (1978). The importance of measuring contrast sensitivity in cases of visual disturbance. British Journal of Ophthalmology, 62(4), 198–209. doi:10.1136/bjo.62.4.198

Arditi, A. (2005). Improving the design of the letter contrast sensitivity test. Investigative Ophthalmology & Visual Science, 46(6), 2225–9. doi:10.1167/iovs.04-1198

Bailey, I. L., Chu, M. A., Jackson, A. J., Minto, H., & Greer, R. B. (2011). Assessment of vision in severely visually impaired populations. In ARVO Fort Lauderdale. Fort Lauderdale, FL.

Bailey, I. L., & Lovie, J. (1976). New design principles for visual acuity letter charts. … Journal of Optometry and Physiological Optics. Retrieved from http://europepmc.org/abstract/MED/998716

Bennett, A. G. (1965). Ophthalmic test types. A review of previous work and discussions on some controversial questions. The British Journal of Physiological Optics, 22(4), 238–71. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/5331332

Bittner, A. K., Jeter, P., & Dagnelie, G. (2011). Grating acuity and contrast tests for clinical trials of severe vision loss. Optometry and Vision Science, 88(10), 1153– 1163.

Black, A. A., Lovie-Kitchin, J. E., Woods, R. L., Arnold, N., Byrnes, J., & Murrish, J. (1997). Mobility performance with retinitis pigmentosa. Clinical Experimental Optometry Journal of the Australian Optometrical Association, 80, 1–12. Retrieved from http://dx.doi.org/10.1111/j.1444-0938.1997.tb04841.x 122

Blakemore, C., & Campbell, F. W. (1969a). Adaptation to spatial stimuli. The Journal of Physiology. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/5761934

Blakemore, C., & Campbell, F. W. (1969b). On the existence of neurones in the human visual system selectively sensitive to the orientation and size of retinal images. The Journal of Physiology, 237–260. Retrieved from http://jp.physoc.org/content/203/1/237.short

Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement between two methods of clinical measurement. Lance (Vol. 47, pp. 307–310). Elsevier Ltd. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2868172

Bochsler, T. M., Legge, G. E., Kallie, C. S., & Gage, R. (2012). Seeing steps and ramps with simulated low acuity: impact of texture and locomotion. Optometry and Vision Science, 89(9), E1299–307. doi:10.1097/OPX.0b013e318264f2bd

Campbell, F. W., Howell, E., & Robson, J. G. (1970). The appearance of gratings with and without the fundamental Fourier component. The Journal of Physiology, 193– 201. Retrieved from http://europepmc.org/abstract/MED/5571919

Campbell, F. W., & Robson, J. G. (1968). Application of Fourier analysis to the visibility of gratings. The Journal of Physiology, 197(3), 551. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1351748/

Cochrane, G. M., Lamoureux, E. L., & Keeffe, J. E. (2008). Defining the content for a new quality of life questionnaire for students with low vision (the Impact of Vision Impairment on Children: IVI_C). Ophthalmic Epidemiology, 15(2), 114–20. doi:10.1080/09286580701772029

Cochrane, G. M., Marella, M., Keeffe, J. E., & Lamoureux, E. L. (2011). The Impact of Vision Impairment for Children (IVI_C): validation of a vision-specific pediatric quality-of-life questionnaire using Rasch analysis. Investigative Ophthalmology & Visual Science, 52(3), 1632–40. doi:10.1167/iovs.10-6079

DeCarlo, D. K., McGwin, G., Bixler, M. L., Wallander, J., & Owsley, C. (2012). Impact of pediatric vision impairment on daily life: results of focus groups. Optometry and Vision Science : Official Publication of the American Academy of Optometry, 89(9), 1409–16. doi:10.1097/OPX.0b013e318264f1dc

Dobson, V., & Teller, D. Y. (1978). Visual acuity in human infants: a review and comparison of behavioral and electrophysiological studies. Vision Research, 18(11), 1469–83. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/364823

123

Dorr, M., Lesmes, L. a, Lu, Z.-L., & Bex, P. J. (2013). Rapid and reliable assessment of the contrast sensitivity function on an iPad. Investigative Ophthalmology & Visual Science, 54(12), 7266–73. doi:10.1167/iovs.13-11743

Dougherty, B. E., Flom, R. E., & Bullimore, M. A. (2005). An evaluation of the Mars Letter Contrast Sensitivity Test. Optometry and Vision Science, 82(11), 970–5. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/16317373

Eperjesi, F., Wolffsohn, J., Bowden, J., Napper, G., & Rubinstein, M. (2004). Normative contrast sensitivity values for the back-lit Melbourne Edge Test and the effect of visual impairment. Ophthalmic & Physiological Optics : The Journal of the British College of Ophthalmic Opticians (Optometrists), 24(6), 600–6. doi:10.1111/j.1475- 1313.2004.00248.x

Ferris, F. L., Kassoff, A., Bresnick, G. H., & Bailey, I. L. (1982). New visual acuity charts for clinical research. American Journal of Ophthalmology, 94(1), 91–96. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7091289

Frantz, R. L., Ordy, J. M., & Udelf, M. S. (1962). Maturation of pattern vision in infants during the first six months. Journal of Comparative and Physiological Psychology. doi:10.1037/h0044173

Geruschat, D., Turano, K. A., & Stahl, J. (1998). Traditional Measures of Mobility Performance and Retinitis Pigmentosa. Optometry & Vision Science, 75(7), 525– 537. Retrieved from http://journals.lww.com/optvissci/Abstract/1998/07000/Traditional_Measures_of_M obility_Performance_and.22.aspx

Ginsburg, A. P. (2003). Contrast sensitivity and functional vision. International Ophthalmology Clinics, 43(2), 5–15. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12711899

Goodrich, G. L., & Ludt, R. (2003). Assessing visual detection ability for mobility in individuals with low vision. Visual Impairment Research, 5(2), 57–71. Retrieved from http://informahealthcare.com/doi/abs/10.1076/vimr.5.2.57.26265

Gothwal, V. K., Lovie-Kitchin, J. E., & Nutheti, R. (2003). The Development of the LV Prasad-Functional Vision Questionnaire: A Measure of Functional Vision Performance of Visually Impaired Children. Investigative Ophthalmology & Visual Science, 44(9), 4131–4139. doi:10.1167/iovs.02-1238

Gothwal, V. K., & Sumalini, R. (2012). The Second Version of the LV Prasad-Functional Vision Questionnaire. Optometry & Vision Science, 89(11), 1601–10. doi:10.1097/OPX.0b013e31826ca291 124

Haymes, S. A., Roberts, K. F., Cruess, A. F., Nicolela, M. T., LeBlanc, R. P., Ramsey, M. S., … Artes, P. H. (2006). The letter contrast sensitivity test: clinical evaluation of a new design. Investigative Ophthalmology & Visual Science, 47(6), 2739–45. doi:10.1167/iovs.05-1419

Hubel, D., & Wiesel, T. (1962). Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. The Journal of Physiology, 106–154. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1359523/

Kirby, R., & Basmajian, J. (1984). The nature of disability and handicap. In Medical Rehabilitation (pp. 14–18). Baltimore: Williams & Wilkins.

Kollbaum, P. (2014). Validation of an iPad test of letter contrast sensitivity. Optometry and Vision Science: 91(3), 291–296. Retrieved from http://pdfs.journals.lww.com/optvissci/9000/00000/Validation_of_an_iPad_Test_of _Letter_Contrast.99023.pdf

Kramer, S. G., & Mcdonald, M. A. (1986). Diagnostic and surgical techniques, assessment of visual acuity in toddlers. Survey of Ophthalmology, 31(3), 189–210.

Kuyk, T., Elliot, J., & Fuhr, P. (1998). Visual correlates of mobility in real world settings in older adults with low vision. Optometry & Vision Science, 75(7), 538–547. Retrieved from http://journals.lww.com/optvissci/abstract/1998/07000/visual_correlates_of_mobilit y_in_real_world.23.aspx

Leat, S. J., Legge, G. E., & Bullimore, M. A. (1999). What is low vision? A re-evaluation of definitions. Optometry & Vision Science, 76(4), 198–211. Retrieved from http://journals.lww.com/optvissci/Abstract/1999/04000/What_Is_Low_Vision__A_ Re_evaluation_of_Definitions.23.aspx

Leat, S. J., & Wegmann, D. (2004). Clinical testing of contrast sensitivity in children: age-related norms and validity. Optometry and Vision Science, 81(4), 245–54. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15097766

Leat, S. J., & Woo, G. C. (1997). The validity of current clinical tests of contrast sensitivity and their ability to predict reading speed in low vision. Eye (London, England), 11 ( Pt 6)(519), 893–9. doi:10.1038/eye.1997.228

Lennie, P., & Hemel, S. B. V. (2002). Visual impairments: determining eligibility for social security benefits. National Academy Press. Retrieved from http://books.google.com/books?id=5QbtT8rzJcwC

125

Long, R., Rieser, J., & Hill, E. (1990). Mobility in individuals with moderate visual impairments. Journal of Visual Impairment & Blindness, 84, 111–118.

Lovie-Kitchin, J. E., Bevan, J. D., & Hein, B. (2001). Reading performance in children with low vision. Clinical & Experimental Optometry : Journal of the Australian Optometrical Association, 84(3), 148–154. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12366326

Marron, J. A., & Bailey, I. L. (1982). Visual factors and orientation-mobility performance. American Journal of Optometry and Physiological Optics, 59(5), 413– 426.

Massof, R. W. (1998). A systems model for low vision rehabilitation II. Measurement of vision disabilities. Optometry & Vision Science, 75(5), 349. Retrieved from http://journals.lww.com/optvissci/Abstract/1998/05000/A_Systems_Model_for_Lo w_Vision_Rehabilitation__II_.25.aspx

Massof, R. W. (2002). The measurement of vision disability. Optometry and Vision Science, 79(8), 516–52. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12199545

Mayer, D. L., Beiser, a S., Warner, a F., Pratt, E. M., Raye, K. N., & Lang, J. M. (1995). Monocular acuity norms for the Teller Acuity Cards between ages one month and four years. Investigative Ophthalmology & Visual Science, 36(3), 671–85. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7890497

McDonald, M. A., Dobson, V., Sebris, S. L., Baitch, L., Varner, D., & Teller, D. Y. (1985). The acuity card procedure: a rapid test of infant acuity. Investigative Ophthalmology & Visual Science, 2, 1158–1162. Retrieved from http://www.iovs.org/content/26/8/1158.short

Owsley, C. (2003). Contrast sensitivity. Ophthalmology Clinics of North America, 16(2), 171–177. doi:10.1016/S0896-1549(03)00003-8

Owsley, C., & Sloane, M. E. (1987). Contrast sensitivity, acuity, and the perception of “real-world” targets. The British Journal of Ophthalmology, 71(10), 791–796. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt= Citation&list_uids=3676151

Pelli, D. G., & Bex, P. (2013). Measuring contrast sensitivity. Vision Research, 90, 10–4. doi:10.1016/j.visres.2013.04.015

126

Pelli, D. G., Robson, J. G., & Wilkins, A. J. (1988). The design of a new letter chart for measuring contrast sensitivity. Clin Vis Sci, 2(3), 187–199. Retrieved from http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:THE+DESIGN+ OF+A+NEW+LETTER+CHART+FOR+MEASURING+CONTRAST+SENSITIVI TY#0

Pokusa, J. E., Kran, B. S., & Mayer, L. (2013). A pilot study of a preferential looking contrast sensitivity test for individuals with vision and/or multiple impairments. Seattle, WA.

Raasch, T. W., Bailey, I. L., & Bullimore, M. A. (1998). Repeatability of visual acuity measurement. Optometry and Vision Science Official Publication of the American Academy of Optometry, 75, 342–348. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt= Citation&list_uids=9624699

Rabin, J., & Wicks, J. (1996). Measuring resolution in the contrast domain: the small letter contrast test. Optometry & Vision Science, 73, 398–403. doi:10.1097/00006324-199606000-00007

Rasch, G. (1960). Probabilistic models for some intelligence and attainment tests. Copenhagen.

Ratliff, F. (1965). Mach bands : quantitative studies on neural networks in the retina. San Francisco, CA: Holden-Day.

Reeves, B. C., Wood, J. M., & Hill, A. R. (1991). Vistech VCTS 6500 charts--within- and between-session reliability. Optometry and Vision Science, 68(9), 728–737.

Shapley, R. M., & Lam, D. (1993). Contrast Sensitivity. Cambridge, Mass: MIT Press.

Sloan, L. (1959). New test charts for the measurement of visual acuity at far and near distances. American Journal of Ophthalmology, 48(Dec), 807–813.

Tandon, J. H. (1994). Nothing more can be done...a fable for our times. Ophthalmology, (7), 203–205.

Tasman, W. (1992). Duane’s Foundations of Clinical Ophthalmology. Philadelphia, PA: Lippincott Williams & Wilkins.

The World Health Organization. (1980). International Classification of Impairments, Disabilities, and Handicaps (ICDH). Geneva: World Health Organization.

127

West, S. K., Rubin, G. S., Broman, A. T., & Mun, B. (2002). How does visual impairment affect performance on tasks of everyday life?, 120(June).

Wilkins, A. J., Della Sala, S., Somazzi, L., & Nimmo-Smith, I. (1988). Age-related norms for the Cambridge Low Contrast Gratings, including details concerning their design and use. Clin Vis Sci, 2(3), 201–212. Retrieved from http://www.essex.ac.uk/psychology/overlays/1988-75.pdf

Wolffsohn, J. S., Eperjesi, F., & Napper, G. (2005). Evaluation of Melbourne edge test contrast sensitivity measures in the visually impaired. Ophthalmic & Physiological Optics : The Journal of the British College of Ophthalmic Opticians (Optometrists), 25(4), 371–4. doi:10.1111/j.1475-1313.2005.00282.x

128

Appendix A: Study Materials

129

130

Dear Parent/Guardian,

We would like to share this invitation with you so that your child may join us (perhaps for the second time) in a collaborative research study. As you may already know, the goal of our study will be to develop a new way to measure a person’s ability to sense the contrast between light, gray, and dark. Unlike previous methods of measuring the ability to sense contrast, this method will not require the patient to read the letters of an eye chart.

We would like to ask your child to take part in this study because s/he is partially sighted. If your child agrees to enter the study, we may:  Ask your child some questions to find out how greatly his/her vision affects him/her in everyday life.  Measure his/her vision using regular lettered eye charts and charts with shapes  Measure his/her ability to see different shades of gray using letters and shapes  Give your child a $5 gift card to thank them for their help with our project.

If you give your permission for your child to participate, s/he will have the opportunity to choose not to participate. If your child is uncomfortable answering any of the questions, s/he may just skip the questions s/he does not like. This test does not involve eye drops of any kind. We will be testing one eye at a time, so we will ask your child to wear a “pirate” type eye patch during testing. However, if your child prefers not to wear it, we will not insist. We do not need to touch your child in any way, although we will be happy to help him/her put on the eye patch if s/he needs it.

If you wish to allow your child to participate, please read and sign the permission form, as well as the HIPAA form. These are the standard forms that are used any time human subjects are involved in research projects. Please complete and return all of the provided forms to us.

If you do NOT wish to allow your child to participate, please complete and return the response form enclosed with this letter. On this form, you may request that we will remove your contact information from our records so that we will not contact you again about this project.

If you have any questions or concerns about your child’s participation in this study, please do not hesitate to contact us.

Thank you for your consideration and for potentially allowing your child to participate in our study!

Sincerely yours,

Angela M Brown, PhD Gregory R Hopkins, OD Bradley E Dougherty, OD, MS Professor Senior Research Associate OSU Clinical Attending Optometrist The OSU College of Optometry OSSB Extern Site Preceptor The OSU College of Optometry [email protected] [email protected] [email protected] o: 614-292-4423 o: 614-688-5542 131

PARENTAL PERMISSION IRB Protocol Number: 2011H0350 Biomedical/Cancer IRB Approval date: 4/4/12 Version: 2.0

The Ohio State University Parental Permission For Child’s Participation in Research

Pilot Study Pilot Study validation of the Stripe Card Contrast Sensitivity Test on Study Title: low-vision participants Principal Angela M Brown, PhD Investigator: The Center for Clinical and Translational Science of the National Institutes of Health Sponsor:

National Eye Institute of the National Institutes of Health

 This is a parental permission form for research participation. It contains important information about this study and what to expect if you permit your child to participate. Please consider the information carefully. Feel free to discuss the study with your friends and family and to ask questions before making your decision whether or not to permit your child to participate.

 Your child’s participation is voluntary. You or your child may refuse participation in this study. If your child takes part in the study, you or your child may decide to leave the study at any time. No matter what decision you make, there will be no penalty to your child and neither you nor your child will lose any of your usual benefits.

 Your decision will not affect your future relationship with The Ohio State University. If you or your child is a student or employee at Ohio State, your decision will not affect your grades or employment status.

 Your child may or may not benefit as a result of participating in this study. Also, as explained below, your child’s participation may result in unintended or harmful effects for him or her that may be minor or may be serious depending on the nature of the research.

 You and your child will be provided with any new information that develops during the study that may affect your decision whether or not to continue to participate. If you permit your child to participate, you will be asked to sign this form and will receive a copy of the form. You are being asked to consider permitting your child to participate in this study for the reasons explained below.

1. Why is this study being done?

This study is being done to develop a new method of measuring visual function in infants and children.

2. How many people will take part in this study?

56 132 PARENTAL PERMISSION IRB Protocol Number: 2011H0350 Biomedical/Cancer IRB Approval date: 4/4/12 Version: 2.0

3. What will happen if my child takes part in this study?

Your child will have five vision tests: two visual acuity tests (one with letters and one with stripes) and three contrast sensitivity tests (one with letters, one with stripes, and one with circles). Your child may also be interviewed using a questionnaire to determine what impact his/her vision has on his/her everyday life.

This study does not involve any eye drops, and we will not be puffing your child’s eyes, or touching his/her eyes in any way.

4. How long will my child be in the study?

For up to three sessions only, lasting no more than a half hour each.

5. Can my child stop being in the study?

Your child may leave the study at any time. If you or your child decides to stop participation in the study, there will be no penalty and neither you nor your child will lose any benefits to which you are otherwise entitled. Your decision will not affect your future relationship with The Ohio State University.

6. What risks, side effects or discomforts can my child expect from being in the study?

These tests are not associated with any known risks, side effects or discomforts.

7. What benefits can my child expect from being in the study?

We do not anticipate that your child will benefit from being in this study.

8. What other choices does my child have if he/she does not take part in the study?

You or your child may choose not to participate without penalty or loss of benefits to which you are otherwise entitled.

9. Will my child’s study-related information be kept private?

Efforts will be made to keep your child’s study-related information confidential. However, there may be circumstances where this information must be released. For example, personal information regarding your child’s participation in this study may be disclosed if required by state law.

Also, your child’s records may be reviewed by the following groups (as applicable to the research):  Office for Human Research Protections or other federal, state, or international regulatory agencies;  U.S. Food and Drug Administration;

133 PARENTAL PERMISSION IRB Protocol Number: 2011H0350 Biomedical/Cancer IRB Approval date: 4/4/12 Version: 2.0

 The Ohio State University Institutional Review Board or Office of Responsible Research Practices;  The sponsor supporting the study, their agents or study monitors; and  Your insurance company (if charges are billed to insurance).

If this study is related to your child’s medical care, your child’s study-related information may be placed in their permanent hospital, clinic, or physician’s office records. Authorized Ohio State University staff not involved in the study may be aware that your child is participating in a research study and have access to your child’s information.

You may also be asked to sign a separate Health Insurance Portability and Accountability Act (HIPAA) research authorization form if the study involves the use of your child’s protected health information.

10. What are the costs of taking part in this study?

There are no costs associated with participating in this study.

11. Will I or my child be paid for taking part in this study?

We will offer your child a gift card of $5.00. By law, payments to subjects are considered taxable income.

12. What happens if my child is injured because he/she took part in this study?

If your child suffers an injury from participating in this study, you should notify the researcher or study doctor immediately, who will determine if your child should obtain medical treatment at The Ohio State University Medical Center.

The cost for this treatment will be billed to you or your medical or hospital insurance. The Ohio State University has no funds set aside for the payment of health care expenses for this study.

13. What are my child’s rights if he/she takes part in this study?

If you and your child choose to participate in the study, you may discontinue participation at any time without penalty or loss of benefits. By signing this form, you do not give up any personal legal rights your child may have as a participant in this study.

You and your child will be provided with any new information that develops during the course of the research that may affect your decision whether or not to continue participation in the study.

You or your child may refuse to participate in this study without penalty or loss of benefits to which you are otherwise entitled.

134 PARENTAL PERMISSION IRB Protocol Number: 2011H0350 Biomedical/Cancer IRB Approval date: 4/4/12 Version: 2.0

An Institutional Review Board responsible for human subjects research at The Ohio State University reviewed this research project and found it to be acceptable, according to applicable state and federal regulations and University policies designed to protect the rights and welfare of participants in research.

14. Who can answer my questions about the study?

For questions, concerns, or complaints about the study you may contact Dr Angela M Brown, Professor, the Ohio State University College of Optometry. 614-292-4423.

For questions about your child’s rights as a participant in this study or to discuss other study-related concerns or complaints with someone who is not part of the research team, you may contact Ms. Sandra Meadows in the Office of Responsible Research Practices at 1-800-678-6251.

If your child is injured as a result of participating in this study or for questions about a study-related injury, you may contact Dr Angela M Brown.

Signing the parental permission form

I have read (or someone has read to me) this form and I am aware that I am being asked to provide permission for my child to participate in a research study. I have had the opportunity to ask questions and have had them answered to my satisfaction. I voluntarily agree to permit my child to participate in this study.

I am not giving up any legal rights by signing this form. I will be given a copy of this form.

Printed name of subject

Printed name of person authorized to provide Signature of person authorized to provide

permission for subject permission for subject

AM/PM

Relationship to the subject Date and time

135 PARENTAL PERMISSION IRB Protocol Number: 2011H0350 Biomedical/Cancer IRB Approval date: 4/4/12 Version: 2.0

Investigator/Research Staff

I have explained the research to the participant or his/her representative before requesting the signature(s) above. There are no blanks in this document. A copy of this form has been given to the participant or his/her representative.

Printed name of person obtaining consent Signature of person obtaining consent

AM/PM

Date and time

136

OSU-OSSB Contrast Sensitivity Study Response Form

Student Name:

If at all possible, please use the enclosed envelope to return these forms to us within 10 days. If we have not heard back from you after that time, you may be contacted via telephone or email regarding this study. Thank you for your consideration regarding our collaborative research study!

The Ohio State University College of Optometry Visual Perception Lab has contacted me to invite my child to participate in a research project:

YES NO

I have completed and will enclose: I prefer that my child NOT participate in this study 1. Parental Permission Form: I will return only this Filled-out/signed/dated top portion of the final page form using the enclosed envelope 2. Personal Health Information in Research Form: Please remove our contact Printed child’s name on first page information from your Initialed/dated bottom corner of first two pages records for the purpose of Signed/filled-out bottom portion of final page. this study

Parent/Guardian Signature: X ______

Comments: ______

137

Verbal Consent Script

Pilot Study validation of the Stripe Card Contrast Sensitivity Test on Low-Vision Participants Angela M Brown, PhD Principal Investigator

Funded by the Center for Clinical and Translational Science and the National Eye Institute.

Hi, [student’s name]. My name is Dr/Mr/Ms [first tester’s name], and Dr/Mr/Ms [second tester’s name] will be helping out today with this research project. If it’s OK with you, we want to test your vision each of your eyes separately, and ask you some questions about your life (if we haven’t already). The goal of our study will be to develop a new way to measure a person’s ability to sense the contrast between light, gray, and dark. The tests of your vision won’t be hard: you’ll just read five different eye charts, two with letters and three with shapes. The questions about your life won’t be hard at all either, since there are no “right” or “wrong” answers to any of them. The information that you share with us will help us to learn how to do a better job testing the vision of other students like yourself. This project will take up to one half-hour of your time and after you finish, we will give you a $5 gift card to say “thank you” for helping us today There is a small risk of a breach of confidentiality, but all efforts will be made to keep everything we learn about you in the strictest confidentiality. We will not link your name to anything that you say in any of our publications. There are no other expected risks of participation. Nothing bad will happen to you if you participate. We won’t be using any eye drops at all and we won’t be puffing you in the eye. We would like for you to put a pirate-patch while viewing the eye charts, but we will not be touching you at all, unless you ask for help with the patch. Participation is voluntary. If you decide not to participate, there will be no penalty or loss of benefits to which you are otherwise entitled. You can, of course, decline to wear the pirate patch or to answer any questions about your life that you are uncomfortable with, as well as to stop participating at any time, without any penalty or loss of benefits to which you are otherwise entitled. If you want to talk to your parents, teachers or friends about this before you decide, just let us know. You can come back another day and we can test you then, if you want. We will only test you if you say it is OK. If you have any additional questions concerning this research or your participation in it, please feel free to contact us or our university research office at any time using the contact sheet that we will provide you with today. Do you have any questions for us right now before we start? Is it OK if we do the eye tests today? Is it OK if we ask you the questions today? If everything is OK, then let’s begin…

Student’s signature: x______Printed name: ______I have read this script to this student, and s/he indicated willingness to participate in this research. First tester’s Signature: x ______Date: ___ /___ /___ First tester’s Printed Name: ______Time: ______AM/PM

138

Verbal Assent Script

Pilot Study validation of the Stripe Card Contrast Sensitivity Test on Low-Vision Participants Angela M Brown, PhD Principal Investigator

Funded by the Center for Clinical and Translational Science and the National Eye Institute.

Hi, [student’s name]. My name is Dr/Mr/Ms [first tester’s name], and Dr/Mr/Ms [second tester’s name] will be helping out today. If it’s OK with you, we want to test your vision in each of your eyes separately, and ask you some questions about your life (if we haven’t already). If you are under the age of 18, your parents have already said it is OK for you to do these tests, but we want to make sure it is OK with you too. We are doing this so we can learn how to do a better job testing the vision of students like yourself. The tests of your vision won’t be hard: you’ll just read five different eye charts, two with letters and three with shapes. We would like for you to put a pirate-patch on the eye you don’t want us to test, but if you don’t want to wear the patch, that would be OK too. If you want us to, we will wear a pirate patch too, just for fun. The questions about your life won’t be hard at all either, and there are no right or wrong answers to any of them. The tests and questions will take no more than a half hour. You can say “No” or stop testing any time you want to. You won’t get into trouble if you say “No” or want to stop. Nothing bad will happen to you if you participate. We won’t be using any drops at all, and we won’t be puffing you in the eye, and we will not be touching you at all, unless you ask for help. We don’t expect anything good to happen either, though. After you finish, we will give you a $5 gift card to say “thank you” for helping us today. If you want to talk to your parents, teachers or friends about this before you decide, just let us know. You can come back another day and we can test you then, if you want. We will only test you if you say it is OK. Is it OK if we do the eye tests? Is it OK if we ask you the questions? Do you have any questions for us before we start?

139

Testing Counterbalance No 1st Test 2nd Test 3rd Test 4th Test 5th Test 01 B-L P-R BD TAC SCCS 02 SCCS TAC BD P-R B-L 03 P-R TAC B-L SCCS BD 04 BD SCCS B-L TAC P-R 05 TAC SCCS P-R BD B-L 06 B-L BD P-R SCCS TAC 07 SCCS BD TAC B-L P-R 08 P-R B-L TAC BD SCCS 09 BD B-L SCCS P-R TAC 10 TAC P-R SCCS B-L BD 11 B-L P-R BD TAC SCCS 12 SCCS TAC BD P-R B-L 13 P-R TAC B-L SCCS BD 14 BD SCCS B-L TAC P-R 15 TAC SCCS P-R BD B-L 16 B-L BD P-R SCCS TAC 17 SCCS BD TAC B-L P-R 18 P-R B-L TAC BD SCCS 19 BD B-L SCCS P-R TAC 20 TAC P-R SCCS B-L BD 21 B-L P-R BD TAC SCCS 22 SCCS TAC BD P-R B-L 23 P-R TAC B-L SCCS BD 24 BD SCCS B-L TAC P-R 25 TAC SCCS P-R BD B-L 26 B-L BD P-R SCCS TAC 27 SCCS BD TAC B-L P-R 28 P-R B-L TAC BD SCCS 29 BD B-L SCCS P-R TAC 30 TAC P-R SCCS B-L BD 31 B-L P-R BD TAC SCCS 32 SCCS TAC BD P-R B-L 33 P-R TAC B-L SCCS BD 34 BD SCCS B-L TAC P-R 35 TAC SCCS P-R BD B-L 36 B-L BD P-R SCCS TAC 37 SCCS BD TAC B-L P-R 38 P-R B-L TAC BD SCCS 39 BD B-L SCCS P-R TAC 40 TAC P-R SCCS B-L BD 41 B-L P-R BD TAC SCCS

140 OSSB Vision Testing Worksheet Subject ID: ____ -______

Date: ______/______/______Testing MM DD YY Order #: | Rx: N; Y | Eye: R; L | ______

141 OSSB Vision Testing Worksheet Subject ID: ____ -______

Date: ______/______/______Testing MM DD YY Order #: | Rx: N; Y | Eye: R; L | ______

142 OSSB Human Subject Payment Receipt Subject ID: ____ -______Protocol #: 2011H0350 Date: ______/______/______MM DD YY

143

The Impact of Visual Impairment in Children (IVI_C) Questionnaire:

 I’m going to read some questions to you.  Please say which answer best describes what you do and feel most of the time.  There are no “right” or “wrong” answers.  Please answer the questions for yourself – we don’t want your family’s answers.  Please answer with one of the responses that I read out to you: Always, Almost Always, Sometimes, Almost Never, Never.  Some things you won’t do. In this case, just answer, ‘No, for other reasons’.

These questions are all about how things are for you BECAUSE OF YOUR EYESIGHT: 01 Do you find it difficult to go down stairs or to step off the sidewalk? 02 Are you confident to make your own way to school? 03 Are you confident to use public transport (such as buses, trains, ferries) by yourself? 04 Are you confident in places you don’t know? 05 Are you confident that you can move around safely in places you don’t know in the daytime? 06 Are you confident that you can move around safely in places you don’t know at night-time? 07 Can you find your friends in the playground at lunch and play time? 08 When you are in a room, can you recognize people you know before they speak to you? 09 Can you take part in games or sports that you want to play with your friends? 10 Do you get the chance to go to activities other than sport (such as social groups)? 11 Has your eyesight stopped you from doing the things that you want to do? 12 Do other students help you when you ask them for help? 13 Do other students help you to join in with them? 14 Do you find it hard to join in with other students? 15 Do you get frustrated? 16 Do other students understand your special needs? 17 Do your teachers understand your special needs? 18 In the classroom, do you get all the same information as other students? 19 Do you get all the information at the same time as the other students? 20 Do you get enough time in school to complete the work set by the teacher? 21 When you are in the classroom, are you confident about asking for help you need? 22 When you ask for help, do people understand how much help you need? 23 Do people tell you that you can’t do the things that you want to do? 24 Do people stop you from doing the things you want to do? 25 Do you get yourself ready for school? 26 Can you recognize coins and paper money when paying for things?

144

The LV Prasad Visual Functioning Questionnaire (LVP_VFQ):

 I’m going to read some questions to you.  Please say which answer best describes what you do and feel most of the time.  There are no “right” or “wrong” answers.  Please answer the questions for yourself - your family is important but we don’t want their answers, we want yours.  I will ask you how much difficulty you have with some typical activities.  Please answer with one of the responses that I read out to you: None, Little, Moderate, Great Deal, Unable.  Some things you won’t do because you are too young or for other reasons. In this case, just answer, ‘Does not apply’.

BECAUSE OF YOUR VISION, how much difficulty do you have with: 01 Making out whether the person you are seeing across the road is a boy or a girl (during the day)? 02 Seeing whether somebody is calling you by waving his or her hand from across the road? 03 Walking alone in the hallway at school without bumping into objects or people? 04 Walking home at night (from anywhere) without assistance when there are streetlights? 05 Copying from the blackboard while sitting on the first seat in your class? 06 Reading the bus numbers? 07 Reading the other details on the bus (such as its destination?) 08 Reading your textbooks at an arm’s length? 09 Writing along a straight line? 10 Finding the next line while reading when you take a break and then resume reading? 11 Locating dropped objects (pen, pencil, and eraser) within the classroom? 12 Threading a needle? 13 Distinguishing between a quarter and a nickel (without touching)? 14 Climbing up or down stairs? 15 Lacing your shoes? 16 Locating a ball while playing in the daylight? 17 Putting toothpaste on your toothbrush? 18 Locating food on your plate while eating?

19 Identifying colors (e.g., while coloring)? 20 How do you think your vision is compared with that of your normal-sighted friends? As good as your friend’s? 2 A little bit worse than your friend’s? 3 Much worse than your friend’s?

145

Orientation & Mobility Severity Rating Scale (O&MSRS) NONE MILD MODERATE SEVERE PROFOUND Severity of Need SCORE 0 1 2 3 4 Distance (1) Level of 20/70 - 20/100 20/100 - 20/200 20/200 - 20/600 20/600 - LP Acuity Vision Light Perception to Nil Peripheral 30˚ - 10˚ field or (Medical) No loss No loss 10˚ field - 1˚ field Field hemianopsia Visual impairment Visual impairment Visual impairment Refer to Guidelines: affects ability to Visual impairment (2) Level of Vision affects ability to affects ability to page 4, Level of Vision travel in a few affects ability to travel (Functional) travel in some travel in most (Functional) #1 selected in all environments environments environments environments Visual skills sufficient Travel tool used Level of travel tool Level of travel tool Level of travel tool use (3) Use/proficiency of for safe, independent only for identification use moderately use severely impacts profoundly impacts travel tool (cane/Alternate travel w/o travel tool or or occasional impacts safe travel safe travel in most safe travel in all Mobility Device) demonstrates mastery instruction in cane in some environments environments of cane skills skills environments

146 (4) Discrepancy in travel Discrepancy in a Discrepancy in Discrepancy in most Discrepancy in all skills between present No discrepancy few selected

some situations situations situations and projected levels situations Needs occasional Needs some Needs some Needs extensive (5) Independence in travel Maintains & refines instruction in current instruction to instruction in new instruction to introduce in current/familiar skills in all current environments to maintain current skills and regular new skills and environments travel environments maintain travel skills in current instruction to maintain current skills ability environments maintain current sills in current environment Conceptual Progress in O&M is Progress in O&M is Progress in O&M is Progress in O&M is understanding (6) Spatial/ Environmental mildly inhibited by moderately inhibited severely inhibited by profoundly inhibited by sufficient for conceptual understanding conceptual by conceptual conceptual conceptual development of travel understanding understanding understanding understanding skills Needs some Needs regular (7) Complexity or Is able to learn new Learns new Needs extensive instruction of skills in instruction in skills introduction of new environment with no environment after instruction in new or new or future for new or future environment specialized instruction brief introduction future environments environments environments Student has Student has (8) Opportunities for use Student has multiple Student has sporadic Student rarely has frequent occasional of skills outside of school opportunities opportunities opportunities opportunities opportunities Adapted from Michigan Department of Education - Low Incidence Outreach - Revised 2012 Need Severity Score 0

CONTRIBUTING FACTORS TO SERVICE DELIVERY Severity of Need Score Frequency add (.5) or subtract (-.5) points for each + or - Total From Page 1 0 - 2 Service not indicated Posture, gait and motor development Other physical or health impairment 2.5 - 4 1 - 5 times / year Nature of eye disease or condition Transition to new school, neighborhood, worksite, etc. 4.5 - 6 3 - 4 times / semester Recent vision loss New, hazardous, complex or difficult environment 1 - 2 times / month 20 - 60 minutes 6.5 - 12 Potential for improvement of travel skills each Age of onset of visual impairment 1 - 2 times / week 30 - 45 minutes 12.5 - 24 Maturity and motivation each Attendance 24.5 - 36 2 or more times/week 30 - 60 Team commitment for follow-up Travel time needed to transport student to area of instruction

affects frequency of instruction RECOMMENDATION OF SERVICES Instruction in low vision aids (monocular, bioptic system) Final Severity of Need

147 Instruction in use of GPS Score Frequency of Service

Other (explain)

Severity of Need Score

Contributing Factors + / -

Final Severity of Need Score