Abstract
purpose. To determine the test–retest variability of the retinal sensitivity of the Nidek MP1 microperimeter in patients with macular disease.
methods. In this prospective study, 50 patients were enrolled with a range of macular diseases. One examiner performed two consecutive microperimetry tests for all patients using the same test strategy. Test–retest variability for mean sensitivity (MS), mean deviation (MD), point-wise sensitivity (PWS), local defect classification (LDC), average sensitivity for the central macula (CMS, 16 loci inside 10°), paracentral macular sensitivity (PMS, 52 loci in the 10 to 20° ring), and dense scotoma size (DSS) were analyzed by calculating the 95% coefficients of repeatability or percentage agreement.
results. Mean (SD) age and visual acuity were 61 (15) years and 0.34 (0.32) logMAR, respectively. The mean difference in MS between tests 1 and 2 was +0.2 dB (SD, 0.9 dB; P = 0.127). The coefficients of repeatability for MS, MD, CMS, and PMS were 1.81, 2.56, 2.13, and 1.93 dB, respectively. The mean (SD) of coefficients of repeatability for PWS across all 68 loci was 5.56 (0.86) dB. Of all test loci in all patients 76% had perfect agreement in LDC, and 94% of patients had a change in DSS of four or fewer test loci.
conclusions. Test–retest variability was lowest for MS and highest for PWS. However, MS does not provide spatial information. The authors recommend the use of CMS and PMS for monitoring macular function and consider a change of greater than 2.56 and 2.31 dB (the upper limit of the 95% confidence interval of their coefficients of repeatability), respectively, to exceed test–retest variability.
Visual acuity is the most widely used measure of macular function. However, in their early stages, many macular diseases do not affect acuity. An alternative approach is to measure paracentral retinal sensitivity or to estimate dense scotoma size (DSS) by using perimetry. Macular perimetry with a stimulus shown on an external screen, such as the 10-2 program of the Humphrey automated perimeter (Carl Zeiss Meditec, Dublin, CA), does not allow accurate co-registration between sensitivity map and fundus image.
1 2 3 4 This limitation was addressed by the development of the scanning laser ophthalmoscope (SLO) microperimeter (Rodenstock, Munich, Germany).
5 6 However, this instrument is no longer in production, and it does not allow automated follow-up examination. The Nidek MP1 microperimeter (Nidek Technologies, Padova, Italy) has several advantages: It is user friendly, incorporates an eye tracker, allows automated follow-up examination at the same retinal loci, and is combined with a color fundus camera for image registration.
7 The popularity of this instrument is reflected by the numerous case studies, case series, and clinical trials using change in microperimetry measures to document the natural history
8 9 10 11 or treatment responses
12 13 14 15 16 17 18 19 20 21 in various retinal and optic nerve diseases.
22
Intrinsic to any psychophysical test is measurement error or test–retest variability. To define disease progression or response to therapy, we need to know the magnitude of test–retest variability in patients with disease. Although the test–retest variability of mean macular and point-wise retinal sensitivity has been investigated in small samples of normal individuals (Hwang JC, et al.
IOVS 2005;46:ARVO E-Abstract 1561) and patients with early age-related macular degeneration (AMD),
23 there are currently no data on measurement error of parameters derived from the Nidek MP1 in patients with a variety of macular diseases. Therefore, the purpose of this study was to investigate the intraexaminer, intrasession test–retest variability of the measurements derived from the Nidek MP1 in patients with a range of macular disease.
Fifty patients with macular disease were enrolled from retinal clinics at Moorfields Eye Hospital. Patients were excluded if they had significant media opacity, prior slit lamp retinal examination or fundus imaging on the same day or were unable to give informed consent. The type of macular disease was not specified as a criterion, so that the most common macular disorders in our retinal clinics were represented. Although patients with previous experience in using the Nidek MP1 were excluded, some patients with prior experience in other types of visual field tests were not.
The study adhered to the tenets of the Declaration of Helsinki for research involving human subjects and was approved by the Moorfields and Whittington Research Ethics Committee on 26 July 2007 (07/H0721/68). Each participant gave written consent before enrollment in the study.
Visual acuity measurement was performed with either Snellen or the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart and converted to their equivalent in logarithm of minimum angle of resolution (logMAR).
24 Pupils were dilated with one drop each of tropicamide 1% and phenylephrine 2.5% at least 15 minutes before microperimetry. The better-seeing eye was chosen for the study and the contralateral eye was patched during the test.
Microperimetry was performed by a single experienced examiner (FKC). The examination was conducted in a darkened room. A 2° red cross was used as the fixation target. After a 30-second fixation test, all patients underwent brief training at the beginning of microperimetry allowing familiarization and practice with correct operation of the response trigger and the stimulus target. This was followed immediately by the first microperimetry test. For the test stimulus, the color was white, the size was Goldmann III (26 minutes arc or 0.4°), and the duration was 200 ms. A 68-loci grid covering the central 20° (similar to the 10-2 program of the Humphrey automated perimeter) was manually centered on the fovea
(Fig. 1) . In most of the patients, the foveal landmark was not visible on the infrared fundus image. The algorithm recommended by Sunness et al.,
25 was used for estimating the center of the fovea, which is located two disc diameters temporal to the disc and one third of a disc diameter inferior to the center of the disc. We used a 4-to-2 staircase strategy, as recommended by Convento and Barbaro,
26 to reduce testing time and possible fatigue. A pretest protocol was selected with the initial brightness set at 6 dB for the four paracentral loci, one in each quadrant. The threshold for these 4 test loci was then used as the starting brightness for testing sensitivity for the remaining 15 test loci in each of the corresponding quadrants. This protocol was chosen, as it theoretically reduces testing time. Stimuli were presented at random. The refinement and recheck option was not used.
After the first test, the patients were asked to sit back and rest with eyes closed until they were ready to proceed to the second fixation and microperimetry test (mean ± SD; 3 ± 1 minute). The follow-up protocol, used for the repeat microperimetry, is similar to the first except for the need to align the infrared fundus image from the second test with that of the first test. Automated registration or alignment of the infrared images was performed by selecting two high-contrast landmarks within two separate 2 × 2-mm square regions (e.g., the vascular arcades, disc, or pigmented lesion). The software then automatically detects the same landmarks in the second infrared image. In eyes in which automated registration was not possible due to the poor quality of the infrared fundus image, manual registration was performed by alignment of two reference points (e.g., vessel bifurcations or crossings) on the first infrared image with the corresponding two reference points in the second infrared image.
Commercial software (Stata ver. 9; Stata Corp, College Station, TX) was used to analyze the data. Summary statistics (mean and SD) for demographic and microperimetry performance data were calculated. Paired t-tests were used to compare the duration of test, eye movement speed, MS and MD between tests 1 and 2. P < 0.05 was considered statistically significant.
For continuous variables (MS, MD, PWS, CMS, PMS, and DSS), the relationship between difference and magnitude was first examined by Bland-Altman plots (graph of difference against the mean). As recommended by Bland and Altman,
27 we used the 95% coefficient of repeatability with 95% confidence intervals as a measure of test–retest variability.
Ceiling and floor effects were observed in the variable, PWS. Some patients had recorded PWS of 0 dB on one or both tests, at the same loci. This result may lead to underestimation of test–retest variability due to the floor effect imposed by having the brightest stimulus luminance set at 127 cd/m2 (0 dB threshold). Similarly, some patients had test loci where PWS was 20 dB on one or both tests. This ceiling effect, imposed by setting the lowest luminance at 1.27 cd/m2, may also underestimate the test–retest variability. Floor effects were also observed in CMS and DSS. For CMS, some patients had an average sensitivity of 0 dB in the central macula in 1 or both of the tests. Similarly, some patients had no dense scotoma in one or both of the tests. For parameters whose computation included sensitivities which had been censored (due to floor or ceiling effects), we calculated coefficients of repeatability including and excluding ceiling and floor effects. This sensitivity analysis was conducted to assess the impact of ceiling and floor effects on the coefficients.
Examination of the Bland-Altman plots showed that the differences in PWS and DSS did not follow a normal distribution. Therefore, we also calculated the proportion of all test loci (for PWS) or of all patients (for DSS) with various limit of change in sensitivity or scotoma size, respectively, between tests 1 and 2. For analysis of test–retest variability in the categorical parameter; LDC, we calculated the percentage of perfect agreement, pooled from all patients and all test loci. The unweighted κ coefficient
28 was also used to describe agreement between the four categories of LDC and interpreted as suggested by Landis and Koch.
29 Symmetry of distribution was investigated by using the binomial probability test (two-sided) against the null hypothesis of equal proportion of test loci with improvement and those with decline in LDC.
Microperimetry has become a common way to measure macular function in assessments of natural history and treatment outcome in macular disease. To make sense of changes in sensitivity at each point tested or in the averaged regional or global indices, one needs to know the inherent variability in the testing method. This variability has been investigated in small samples of normal individuals and patients with early AMD.
23 We specifically examined only subjects with macular disease because it is well recognized that the variability of psychophysical tests, such as VA, in diseased states may be greater than that in normal subjects.
30 31
We have shown that in this cohort, the intraexaminer, intrasession coefficients of repeatability were 1.81 and 2.56 dB for MS and MD, respectively. In contrast to global measures, PWS had an average coefficient of repeatability of 4.96 dB. This was increased to 5.56 dB when 38% of test loci affected by ceiling or floor effect were removed. Ceiling and floor effects are limitations of several psychophysical measures such as visual acuity charts and health-related questionnaires.
32 33 Such effects in PWS are due to relatively narrow range of stimulus luminance (1.27–127 cd/m
2) offered by the Nidek MP1. LDC is an alternative way of describing local retinal sensitivity in relation to expected normal values. Although LDC had good agreement, a quarter of the pooled test loci did not agree and showed improvement in test 2. In contrast, the slight improvements in MS and MD, from test 1 to test 2, were not statistically significant. How this relates to perimetry learning and visual adaptation is discussed later in the article. Since global indices (MS and MD) do not provide spatial information which is critical in clinical decision and estimates of test–retest variability of point-wise measures (PWS and LDC) are limited by ceiling and floor effects, we devised two other types of indices to describe macular function.
We used cluster indices analogous to that used in monitoring of glaucomatous field defect progression.
34 35 As the decision to treat macular conditions are usually driven by proximity of the lesion to the foveal center, we divided the test loci into two regions: the central (central 10°) and paracentral (10–20° ring) macula. Although only 16 of 68 test loci were within central macula, the coefficient of repeatability for CMS, 2.04 dB (2.13 dB if floor effects excluded), was similar to MS. We created another parameter, DSS, that may provide additional information on disease progression, especially relevant in clinical trials involving geographic atrophy.
36 We found that only a minority of patients had a change of more than four test loci in DSS and the coefficient of repeatability was seven test loci. The disparity in the estimation of repeatability was due to the three outliers and the lack of normal distribution in the change in DSS, a prerequisite of using the methods described by Bland and Altman
(Fig. 3F) .
27 Although DSS may indicate the extent of the scotoma, it does not provide spatial information. Therefore, we are currently re-examining the data to determine the test–retest variability of dense scotoma localization.
Hwang et al. (
IOVS 2005;46:ARVO E-Abstract 1561) studied test–retest variability in 10 healthy subjects between 22 and 34 years of age using Goldmann size III and 200-ms duration stimuli at 40 preset locations in an array covering the central 20°. Each subject performed three tests within the same session; with steady central fixation in tests 1 and 3 and simulated unsteady fixation in test 2. They found that the averaged difference between tests for the 40 locations ranged between 0.1 and 1.6 dB. The maximum difference in all patients and test loci was 6 dB; 2 of 40 test loci in two subjects between trials 1 and 3 and 3 of 40 test loci in 2 subjects between trials 1 and 2. It is difficult to compare our results with those from the study by Hwang et al. However, we found that the maximum PWS difference was 16 dB compared with 6 dB in their study, which may reflect the expected higher variability in disease. Nevertheless, our data showed that 97% of the point-wise difference was within 6 dB. More recently, Weingessel et al.
23 reported the interexaminer/intrasession and intraexaminer/intersession reliabilities of Nidek MP1 in 30 eyes of 30 normal subjects (15 young subjects and 15 aged over 60 years) and 5 eyes of 5 patients with early AMD. Using a smaller grid of 41 loci within the central 10° and a 4-2-1staircase strategy, they found no significant bias in MS, between the examiners or sessions separated by 1 week. The SD for the difference in MS ranged from 0.5 to 0.9 dB, and the reported 95% limits of agreement ranged from as low as −1.9 dB to as high as +1.9 dB. Coefficients of repeatability and reliability of PWS were not reported. Although their test strategy is slightly different from that used in our study, the results are comparable.
There is an extensive body of literature regarding factors that affect variability in visual fields of normal subjects and patients with ocular hypertension and glaucoma.
37 38 39 40 The effect of perimetric experience on field test results has been described in automated perimetry when patients perform the test over several sessions.
41 42 In addition, other sources of variability include those related to the patient (age, concentration, understanding, pupil size, comorbidities, extent of local damage, test point locus)
41 43 44 45 46 47 48 49 and the instrument (size and duration of stimulus, test strategy, and starting stimulus luminance level).
50 51 52 In this study, we tried to reduce variability due to learning by providing pretest training to allow familiarization with (1) stimulus target size, location, and luminance as well as (2) the correct operation of the response trigger. We also tried to reduce the fatigue effect by giving each patient the opportunity to rest between the two tests. We also standardized the explanation regarding how to perform microperimetry, and the eyes of all patients were fully dilated before testing. A single examiner performed all microperimetry tests and all testing algorithms (stimulus size, duration, staircase strategy, and grid pattern) were kept the same. Similarly, the same refractive error correction setting was used for both tests.
Despite these attempts to reduce variability, we found small improvement in sensitivity measures between the two tests. Factors that may explain the variability reported can be separated into (1) those related to the patient or the eye and (2) those attributable to the instrumentation or the examiner. Patients’ concentration during testing session may play a role. For patients without prior experience in standard automated perimetry, the training protocol may not provide adequate practice, as it uses only a stimulus with the brightest luminance and therefore does not reflect the actual test where the majority of stimulus luminance is presented near the threshold level. Therefore, improvement in sensitivity from test 1 to 2 may still be due to a learning effect. The contribution of cone and rod responses to high-mesopic perimetric threshold is not well understood. Therefore, we cannot rule out the contribution of varying adaptation during the tests to the improvement in sensitivity and test–retest variability. Under- and overcorrection of the refractive error may lead to fluctuation in the accommodative state and hence perimetric threshold during the examination. Retinal disease may also contribute to variability in threshold, similar to the earlier report of higher variability in eyes with glaucomatous field loss.
48
Factors related to the operation and design of the instrument may also contribute to variability. The Nidek MP1 uses a 25-Hz eye tracker to ensure that the same retinal locus is tested repeatedly within the same test session. The eye tracker relies on a high-contrast target on the IR fundus image captured at the beginning of the test session. The accuracy in tracking by using central landmarks may be different from the use of peripheral landmark due to curvature of fundus and different gaze positions of the eye. Lack of high-contrast fundus features may lead to tracking of similar but different retinal landmarks and therefore, projection of stimuli to the incorrect retinal loci. Another source of variability is the error in image registration when the follow-up testing protocol was used. Previous studies have reported error of up to 2° and 0.5° in image registration (Enoch JM, et al. IOVS 2004;45:ARVO E-Abstract 2772 and Woods RL, et al. IOVS 2007;48:ARVO E-Abstract 144). In our study, alignment was performed in the majority of cases by using two high-contrast landmarks and automated registration. However, due to the poor quality of infrared images in some patients, two large vessel crossings were chosen manually for registration. Depending on where the landmarks are situated, this method may be a less accurate one for image alignment, resulting in shift in the position or change in size of the test grid. As a consequence, sensitivity measurement of test loci adjacent to a dense scotoma or in regions with steep gradients of sensitivity may be less repeatable. During examination, some patients complained that the trigger failed to register their responses. This is because the current software will only recognize a response within 1.5 seconds after the stimulus presentation. Therefore, slow response will increase variability. The 4-2 strategy uses only one reversal to determine the threshold of stimulus perception. Although variability may be minimized by increasing the number of reversals, such an increase can lengthen the test’s duration and cause increased fatigue during the test.
In conclusion, we have shown that the test–retest variability is lowest for MS and highest for PWS. However, MS does not provide topographic information and may be less sensitivity for localized change since it is an average of retinal sensitivities across a large number of test loci. On the other hand, the higher variability in PWS may be too noisy to be useful in detecting clinical change. Therefore, we recommend the use of CMS and PMS for monitoring macular function in clinical trials, because their variability is similar to that of MS, but they provide additional clinically useful topographic information. Using the upper limits of the 95% CI for the coefficients of repeatability, we propose that a change in CMS and PMS of greater than 2.56 and 2.31 dB be considered to exceed test–retest variability. However, we caution about the use of these criteria for determining disease progression or response to therapy, as the intersession and interexaminer test–retest variability may be different. We anticipate that broadening of the stimulus luminance range by the manufacturer may minimize the constraints of ceiling and floor effects on the ability to estimate the coefficient of repeatability for PWS. Furthermore, advances in image registration, eye tracker technology, and testing strategy may reduce the test–retest variability of microperimetry.
Supported by The London Project To Cure Blindness.
Submitted for publication September 22, 2008; accepted May 20, 2009.
Disclosure:
F.K. Chen, None;
P.J. Patel, None;
W. Xing, None
C. Bunce, None;
C. Egan, None;
A.T. Tufail, None;
P.J. Coffey, None;
G.S. Rubin, None;
L. Da Cruz, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “
advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
No reprints will be available.
Corresponding author: Fred K. Chen, Vitreoretinal Surgery, Moorfields Eye Hospital, 162 City Road, London EC1V 2PD, UK;
[email protected].
Table 1. Summary of Microperimetry Performance in Tests 1 and 2
Table 1. Summary of Microperimetry Performance in Tests 1 and 2
| Test 1 | Test 2 | Test 2 − Test 1 | P (Paired t-Test) |
| Total test duration, min | 12.6 (3.6) | 11.5 (3.6) | −1.1 (2.1) | 0.000 |
| Eye tracking time, min | 11.4 (2.9) | 10.2 (2.7) | −1.2 (1.8) | 0.000 |
| EM speed, deg/s | 0.26 (0.13) | 0.28 (0.13) | 0.02 (0.10) | 0.142 |
| MS, dB | 10.1 (5.6) | 10.3 (5.5) | 0.2 (0.9) | 0.127 |
| MD, dB | −7.4 (4.2) | −7.1 (4.1) | 0.3 (1.3) | 0.071 |
Table 2. Coefficients of Repeatability for Five Global and Regional Retinal Sensitivity Measures in the Macular Region
Table 2. Coefficients of Repeatability for Five Global and Regional Retinal Sensitivity Measures in the Macular Region
| Parameters from Nidek MP1 | Sample Size | 95% Coefficient of Repeatability (dB) | 95% Confidence Interval of Coefficient of Repeatability (dB) |
| MS | 50 | 1.81 | 1.46–2.17 |
| MD | 50 | 2.56 | 2.06–3.07 |
| CMS | 50 | 2.04 | 1.64–2.44 |
| CMS* | 46 | 2.13 | 1.69–2.56 |
| PMS | 50 | 1.93 | 1.55–2.31 |
| DSS | 50 | 7, † | 5.4–8.0, † |
| DSS* | 32 | 8, † | 6.3–10.3, † |
Table 3. Frequency of Change in Point-wise Sensitivity
Table 3. Frequency of Change in Point-wise Sensitivity
| Change (dB) | Frequency (Count) | Cumulative Percentage |
| No change | 1599 | 47 |
| −1 to −2 and +1 to +2 | 462 and 698 | 81 |
| −3 to −4 and +3 to +4 | 236 and 160 | 93 |
| −5 to −6 and +5 to +6 | 37 and 110 | 97.3 |
| −7 to −8 and +7 to +8 | 32 and 25 | 99 |
| −9 to −10 and +9 to +10 | 6 and 17 | 99.7 |
| −11 to −16 and +11 to +16 | 7 and 8 | 100 |
| Total | 3397 | 100 |
| Test 1 | Test 2 | | | | |
| Normal | Suspect | Relative Scotoma | Dense Scotoma | Total |
| Normal | 506 | 82 | 90 | 0 | 678 |
| Suspect | 99 | 102 | 104 | 0 | 305 |
| Relative scotoma | 97 | 143 | 1356 | 95 | 1691 |
| Dense scotoma | 0 | 0 | 114 | 609 | 723 |
| Total | 702 | 327 | 1664 | 704 | 3397 |
Table 5. Frequency of Change in Dense Scotoma Size
Table 5. Frequency of Change in Dense Scotoma Size
| Change (Test Loci) | Frequency (Count) | Cumulative Percentage |
| No change | 16 | 32 |
| −1 to −2 and +1 to +2 | 13 and 9 | 76 |
| −3 to −4 and +3 to +4 | 4 and 5 | 94 |
| −5 to −6 and +5 to +6 | 0 and 0 | 94 |
| −7 to −8 and +7 to +8 | 0 and 0 | 94 |
| −9 to −10 and +9 to +10 | 1 and 1 | 98 |
| −11 to −16 and +11 to +16 | 1 and 0 | 100 |
| Total | 50 | 100 |
The authors thank the reviewers for their insightful comments.
AtchisonDA, Lovie-KitchinJE, SwannPG. Investigation of central visual fields in patients with age-related macular changes. Optom Vis Sci. 1990;67:179–183.
[CrossRef] [PubMed]TolentinoMJ, MillerS, GaudioAR, SandbergMA. Visual field deficits in early age-related macular degeneration. Vision Res. 1994;34:409–413.
[CrossRef] [PubMed]SunnessJS, SchuchardRA, ShenN, RubinGS, DagnelieG, HaselwoodDM. Landmark-driven fundus perimetry using the scanning laser ophthalmoscope. Invest Ophthalmol Vis Sci. 1995;36:1863–1874.
[PubMed]RohrschneiderK, FendrichT, BeckerM, KrastelH, KruseFE, VolckerHE. Static fundus perimetry using the scanning laser ophthalmoscope with an automated threshold strategy. Graefes Arch Clin Exp Ophthalmol. 1995;233:743–749.
[CrossRef] [PubMed]TimberlakeGT, MainsterMA, WebbRH, HughesGW, TrempeCL. Retinal localization of scotomata by scanning laser ophthalmoscopy. Invest Ophthalmol Vis Sci. 1982;22:91–97.
[PubMed]MainsterMA, TimberlakeGT, WebbRH, HughesGW. Scanning laser ophthalmoscopy: clinical applications. Ophthalmology. 1982;89:852–857.
[CrossRef] [PubMed]MidenaE, RadinPP, PilottoE, GhirlandoA, ConventoE, VaranoM. Fixation pattern and macular sensitivity in eyes with subfoveal choroidal neovascularization secondary to age-related macular degeneration: a microperimetry study. Semin Ophthalmol. 2004;19:55–61.
[CrossRef] [PubMed]CarpinetoP, CiancagliniM, Aharrh-GnamaA, et al. Optical coherence tomography and fundus microperimetry imaging of spontaneous closure of traumatic macular hole: a case report. Eur J Ophthalmol. 2005;15:165–169.
[PubMed]SoukaAA, HillenkampJ, GoraF, GabelVP, FrammeC. Correlation between optical coherence tomography and autofluorescence in acute posterior multifocal placoid pigment epitheliopathy. Graefes Arch Clin Exp Ophthalmol. 2006;244:1219–1223.
[CrossRef] [PubMed]BoscarinoMA, JohnsonTM. Microperimetry in multiple evanescent white dot syndrome. Can J Ophthalmol. 2007;42:743–745.
[CrossRef] [PubMed]UnokiN, NishijimaK, SakamotoA, et al. Retinal sensitivity loss and structural disturbance in areas of capillary nonperfusion of eyes with diabetic retinopathy. Am J Ophthalmol. 2007;144:755–760.
[CrossRef] [PubMed]ChalamKV, GuptaSK, ShahGY, AgarwalS. Successful management of melanocytoma-associated choroidal neovascularization with photodynamic therapy. Eur J Ophthalmol. 2006;16:776–778.
[PubMed]Richter-MuekschS, Vecsei-MarlovitsPV, SacuSG, KissCG, WeingesselB, Schmidt-ErfurthU. Functional macular mapping in patients with vitreomacular pathologic features before and after surgery. Am J Ophthalmol. 2007;144(1)23–31.
[CrossRef] [PubMed]YodoiY, TsujikawaA, KamedaT, et al. Central retinal sensitivity measured with the micro perimeter 1 after photodynamic therapy for polypoidal choroidal vasculopathy. Am J Ophthalmol. 2007;143:984–994.
[CrossRef] [PubMed]RenschF, SpandauUH, SchlichtenbredeF, von BaltzS, LibondiT, JonasJB, VossmerbaeumerU. Intravitreal bevacizumab for myopic choroidal neovascularization. Ophthalmic Surg Lasers Imaging. 2008;39:182–185.
[CrossRef] [PubMed]BainbridgeJW, SmithAJ, BarkerSS, et al. Effect of gene therapy on visual function in Leber’s congenital amaurosis. N Engl J Med. 2008;358:2231–2239.
[CrossRef] [PubMed]ChenFK, UppalGS, RubinGS, WebsterA, CoffeyPJ, da CruzL. Evidence of retinal function using microperimetry following autologous retinal pigment epithelium-choroid graft in macular dystrophy: outcome in five cases. Invest Ophthalmol Vis Sci. 2008;49(7)3143–3150.
[CrossRef] [PubMed]IssaPC, FingerRP, HolzFG, SchollHP. Eighteen-month follow-up of intravitreal bevacizumab in type 2 idiopathic macular telangiectasia. Br J Ophthalmol. 2008;92:941–945.
[CrossRef] [PubMed]YenerelNM, KucumenB, GorgunE, DincUA. Atypical presentation of multiple evanescent white dot syndrome (MEWDS). Ocul Immunol Inflamm. 2008;16:113–115.
[CrossRef] [PubMed]RadtkeND, AramantRB, PetryHM, GreenPT, PidwellDJ, SeilerMJ. Vision improvement in retinal degeneration patients by implantation of retina together with retinal pigment epithelium. Am J Ophthalmol. 2008;146(2)172–182.
[CrossRef] [PubMed]GrengaP, LupoS, DomanicoD, VingoloEM. Efficacy of intravitreal triamcinolone acetonide in long standing diabetic macular edema: a microperimetry and optical coherence tomography study. Retina. .Published online July 28, 2008.
RohrschneiderK, BultmannS, SpringerC. Use of fundus perimetry (microperimetry) to quantify macular sensitivity. Prog Retin Eye Res. 2008;27(5)536–548.
[CrossRef] [PubMed]WeingesselB, SacuS, Vecsei-MarlovitsPV, WeingesselA, Richter-MuekschS, Schmidt-ErfurthU. Interexaminer and intraexaminer reliability of the microperimeter MP-1. Eye. .Published online August 1, 2008.
HolladayJT. Visual acuity measurements. J Cataract Refract Surg. 2004;30:287–290.
[CrossRef] [PubMed]SunnessJS, BresslerNM, TianY, AlexanderJ, ApplegateCA. Measuring geographic atrophy in advanced age-related macular degeneration. Invest Ophthalmol Vis Sci. 1999;40:1761–1769.
[PubMed]ConventoE, BarbaroG. Technical insights in the interpretation of automatic microperimetry.MidenaE eds. Perimetry and the Fundus: An Introduction to Microperimetry. 2007;229–237.Slack Inc. Thorofare.
BlandJM, AltmanDG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310.
[PubMed]FleissJL. Statistical Methods for Rates and Proportions. 1981;John Wiley & Sons New York.
LandisJR, KochGG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174.
[CrossRef] [PubMed]ElliottDB, SheridanM. The use of accurate visual acuity measurements in clinical anti-cataract formulation trials. Ophthalmic Physiol Opt. 1988;8:397–401.
[CrossRef] [PubMed]PatelPJ, ChenFK, RubinGS, TufailA. Intersession repeatability of visual acuity scores in age-related macular degeneration. Invest Ophthalmol Vis Sci. .Published online February 14, 2009.
BeckRW, MaguireMG, BresslerNM, GlassmanAR, LindbladAS, FerrisFL. Visual acuity as an outcome measure in clinical trials of retinal diseases. Ophthalmology. 2007;114:1804–1809.
[CrossRef] [PubMed]MassofRW. Application of stochastic measurement models to visual function rating scale questionnaires. Ophthalmic Epidemiol. 2005;12:103–124.
[CrossRef] [PubMed]AsmanP, HeijlA, OlssonJ, RootzenH. Spatial analyses of glaucomatous visual fields; a comparison with traditional visual field indices. Acta Ophthalmol (Copenh). 1992;70:679–686.
[PubMed]AsmanP, HeijlA. Weighting according to location in computer-assisted glaucoma visual field analysis. Acta Ophthalmol (Copenh). 1992;70:671–678.
[PubMed]SunnessJS, MargalitE, SrikumaranD, et al. The long-term natural history of geographic atrophy from age-related macular degeneration: enlargement of atrophy and implications for interventional clinical trials. Ophthalmology. 2007;114:271–277.
[CrossRef] [PubMed]LewisRA, JohnsonCA, KeltnerJL, LabermeierPK. Variability of quantitative automated perimetry in normal observers. Ophthalmology. 1986;93:878–881.
[CrossRef] [PubMed]KatzJ, SommerA, WittK. Reliability of visual field results over repeated testing. Ophthalmology. 1991;98:70–75.
[CrossRef] [PubMed]BjerreA, GriggJR, ParryNR, HensonDB. Test–retest variability of multifocal visual evoked potential and SITA standard perimetry in glaucoma. Invest Ophthalmol Vis Sci. 2004;45:4035–4040.
[CrossRef] [PubMed]KimLS, McAnanyJJ, AlexanderKR, FishmanGA. Intersession repeatability of Humphrey perimetry measurements in patients with retinitis pigmentosa. Invest Ophthalmol Vis Sci. 2007;48:4720–4724.
[CrossRef] [PubMed]HeijlA, LindgrenG, OlssonJ. The effect of perimetric experience in normal subjects. Arch Ophthalmol. 1989;107:81–86.
[CrossRef] [PubMed]WildJM, SearleAE, Dengler-HarlesM, O'NeillEC. Long-term follow-up of baseline learning and fatigue effects in the automated perimetry of glaucoma and ocular hypertensive patients. Acta Ophthalmol (Copenh). 1991;69:210–216.
[PubMed]HeijlA, BengtssonB. The effect of perimetric experience in patients with glaucoma. Arch Ophthalmol. 1996;114:19–22.
[CrossRef] [PubMed]WernerEB, AdelsonA, KrupinT. Effect of patient experience on the results of automated perimetry in clinically stable glaucoma patients. Ophthalmology. 1988;95:764–767.
[CrossRef] [PubMed]GuthauserU, FlammerJ. Quantifying visual field damage caused by cataract. Am J Ophthalmol. 1988;106:480–484.
[CrossRef] [PubMed]FlammerJ, DranceSM, FankhauserF, AugustinyL. Differential light threshold in automated static perimetry: factors influencing short-term fluctuation. Arch Ophthalmol. 1984;102:876–879.
[CrossRef] [PubMed]HeijlA, LindgrenG, OlssonJ. Normal variability of static perimetric threshold values across the central visual field. Arch Ophthalmol. 1987;105:1544–1549.
[CrossRef] [PubMed]HeijlA, LindgrenA, LindgrenG. Test–retest variability in glaucomatous visual fields. Am J Ophthalmol. 1989;108:130–135.
[CrossRef] [PubMed]ChauhanBC, DranceSM, LeBlancRP, LiebermanMF, MillsRP, WernerEB. Technique for determining glaucomatous visual field progression by using animation graphics. Am J Ophthalmol. 1994;118:485–491.
[CrossRef] [PubMed]GlassE, SchaumbergerM, LachenmayrBJ. Simulations for FASTPAC and the standard 4-2 dB full-threshold strategy of the Humphrey Field Analyzer. Invest Ophthalmol Vis Sci. 1995;36:1847–1854.
[PubMed]WallM, KutzkoKE, ChauhanBC. Variability in patients with glaucomatous visual field damage is reduced using size V stimuli. Invest Ophthalmol Vis Sci. 1997;38:426–435.
[PubMed]ArtesPH, IwaseA, OhnoY, KitazawaY, ChauhanBC. Properties of perimetric threshold estimates from Full Threshold, SITA Standard, and SITA Fast strategies. Invest Ophthalmol Vis Sci. 2002;43:2654–2659.
[PubMed]