Faculty Bibliography

Although biomedical statistics is part of any scientific curriculum, a review of the current scientific literature indicates that statistical data analysis is an area that frequently needs improvement. To address this, we here cover some of the most common problems in statistical analysis, with an emphasis on an intuitive, tutorial approach rather than a rigorous, proof-based one. The topics covered in this manuscript are whether to enter eyes or patients into the analysis, issues related to multiple testing, pitfalls surrounding the correlation coefficient (causation, insensitivity to patterns, range confounding, unsuitability for method comparisons), and when to use standard deviation (SD) versus standard error of the mean (SEM) 'antennas' on graphs

Shinoda and colleagues hypothesized that patients with cone dystrophy (CD) might suffer from a selective ON-system deficit, based on the local nature of the disease [Shinoda, K, Ohde, H, Inoue, R, Ishida, S, Mashima, Y, & Oguchi, Y (2002). ON-pathway disturbance in two siblings. Acta Ophthalmologica Scandinavica, 80, 219-223]. The purpose of the current study was to test this hypothesis by examining onset and offset responses as a function of eccentricity in a group of patients with CD using long-duration LED stimuli. Nine patients with CD participated in this study (mean age of 36.1 years and visual acuity 20/200). For this study, the following measures were obtained: Humphrey threshold visual fields, standard multifocal ERGs (mfERGs) as well as mfERGs to long duration stimuli recorded using the Retiscan stimulator (Roland Instruments). This display contained 61 scaled hexagons and the LEDs were on for 100ms (180cd/m(2)) and off for 100ms. In addition, standard full-field photopic and flicker ERGs using Ganzfeld stimulation were obtained. For the control subjects, the onset responses were larger than the offset responses at all eccentricities; whereas for the patients, there was overlap between the amplitudes of the onset and offset responses. For the patients, the amplitude ratios (relative to the control data) indicated that the difference between the onset and offset responses was greatest for the central-most ring and this difference decreased with increasing eccentricity. For the onset responses, Humphrey thresholds and mfERG amplitudes, performance was poorest for the center ring and best for the most peripheral ring; for the offset responses, the opposite pattern of results was obtained. The differences in the pattern of results in the long duration mfERG data are consistent with a selective loss of the onset responses in our patient population

PURPOSE: To investigate possible adverse effects of a yellow-tinted intraocular lens (IOL) on scotopic sensitivity and hue discrimination. SETTING: Departments of Ophthalmology, Columbia University and New York University School of Medicine, New York, New York, USA. METHODS: Nine patients with a yellow-tinted IOL in 1 eye and a colorless ultraviolet IOL in the fellow eye and 9 young phakic subjects with and without a yellow-tinted clip-on lens were tested. Hue discrimination was measured with the Farnsworth-Munsell (FM) 100-hue test. Dark-adapted thresholds to 440 nm, 500 nm, and 650 nm lights were measured at 23 locations using a modified Humphrey perimeter, and dark-adapted thresholds to white light were measured at 15 degrees temporal retina. RESULTS: In the 9 patients, there were no significant differences in dark-adapted sensitivities to 440, 500, 650 nm, or white-light stimuli and no differences in FM 100-hue error scores between eyes with yellow-tinted IOLs and those with colorless IOLs. Similarly, in young phakic subjects, there were no significant differences in FM 100-hue error scores or dark-adapted sensitivity to the white light with and without the yellow-tinted clip-on lens. However, with the clip-on lens, mean sensitivities to the 440 nm, 500 nm, and 650 nm stimuli were significantly decreased by 2.7 to 2.8 dB, 0.7 to 1.0 dB, and 0 to 1.2 dB, respectively. CONCLUSION: Results suggest that implantation of a yellow-tinted IOL has non-significant effect on scotopic sensitivity and hue discrimination

Our aim was to determine whether patients with retinitis pigmentosa show differences in L- and M-cone multifocal visual evoked potential (mfVEP) responses that are eccentricity dependent, as has been shown for control subjects. Second, we compared the losses for mfVEPs to losses on achromatic visual field and multifocal electroretinogram (mfERG) measures in the patients. Monocular mfVEPs were recorded to a pattern reversing display that modulated only the L- or M-cones. Also, standard automated achromatic visual fields and mfERGs were obtained. For the control subjects, the ratio of L-cone to M-cone mfVEP amplitudes increased as a function of retinal eccentricity. For the patients, the ratio did not vary with eccentricity. For all measures, responses were least affected for the first ring (central 2.4 degrees ) and most affected for the third ring (11.6 degrees - 44.4 degrees ). For the first ring, mfERG amplitudes were more impaired than were the mfVEPs or the visual field thresholds. For most of the patients, there was local response correspondence among our measures of visual function

We examined the effects of inter-modal attention and mental arithmetic on Humphrey visual field sensitivity and multifocal visual evoked potential (mfVEP) amplitude. Four normally sighted subjects (ages ranging from 24 to 58 years) participated in this study. Monocular visual field sensitivity was measured under two conditions: (1) standard testing condition and (2) while the subject performed a Paced Auditory Serial Addition Task (PASAT). Monocular mfVEPs were recorded in response to a 60-sector stimulus. The checkerboard pattern in each sector was contrast reversed according to a binary m-sequence. mfVEPs were recorded under two conditions: (1) standard testing conditions and (2) while the subject performed a PASAT. We found that, when compared to the no-task condition, all subjects had locations of significantly reduced Humphrey visual field sensitivities when performing the PASAT. In contrast, there were no significant decreases in mfVEP amplitude in any sector for any of the subjects while performing the PASAT. Our findings indicate that divided attention and ongoing mental processes did not affect the mfVEP. Therefore, the mfVEP provides an objective measure of visual field function that may be useful for some patients with unreliable automated static perimetry results

We examined the reliability of Humphrey visual field thresholds and multifocal electroretinogram (mfERG) amplitudes and timing in a group of patients with Retinitis Pigmentosa (RP). Eight patients with RP and seven control subjects were tested five times: at baseline (visit #0), at three weekly follow-up visits (visits #1 - #3), and at three months (visit #4). For the Humphrey thresholds, differences between dB values on repeat visits were obtained. Differences between log values on repeat visits were calculated for mfERG amplitude and implicit time. We used the standard deviations of these difference scores as a measure of reliability and the means of the difference scores as a measure of progression. We found that the majority of the patients' repeat data were more variable than that of the control subjects for both the Humphrey and mfERG. We found no single factor that predicted the magnitude, or the variance, of the SD of differences scores for the patients. We recommend that each patient's reliability be assessed individually. Ultimately, the choice of an outcome measure must be guided by its reliability, as well as its ability to assess the visual function of interest

The purpose of this study was to investigate atypical multifocal ERG (mfERG) responses for patients with diseases that can affect the photoreceptors. MfERGS were obtained from seven patients with retinitis pigmentosa (RP), three with progressive cone dystrophy (CD) and eight with diabetic retinopathy (DR). Both first- and second-order kernel responses were analyzed. The amplitudes and implicit times of the first-order responses were compared to those obtained from age-similar controls. For the first slice of the second-order response, the root-mean-square (RMS) and the signal-to-noise ratio (SNR) of each response were calculated. Achromatic visual fields were also obtained from each subject. For the three groups of patients, first-order responses with relatively large amplitudes, broad-shaped waveforms and markedly increased implicit times had non-measurable second-order responses. These responses were associated with areas of decreased visual field sensitivity. As RP, CD and DR affect the outer retina, the results are consistent with damage to the outer plexiform layer rather than damage to the inner retina

Multidimensional psychophysical and electrophysical maps of the central retina are essential for assessing the functioning of the diseased retina. In this study, grating acuity, contrast sensitivity, duration for letter identification, multifocal electroretinograms, and Humphrey visual field thresholds were measured at equivalent positions throughout the central 20 degrees . We found that the rates of sensitivity loss were not equivalent for all psychophysical measures. The rate of loss in the duration required for letter identification as a function of eccentricity was the steepest, followed by acuity and contrast sensitivity. The rate of loss in luminance sensitivity as measured in the Humphrey visual field was the shallowest. The pattern of losses also varied across meridians. Specifically, the rate of loss as a function of eccentricity was highest in the vertical meridian and lowest in the horizontal meridian. These maps and the correlations among measures as a function of retinal position serve as a baseline so that we can examine disease effects throughout the retina. In addition, the development of vision rehabilitation programs focused on eccentric viewing training should consider the differential sensitivities of the peripheral retina