Reproducibility of Anterior Chamber Angle Measurements Obtained with Anterior Segment Optical Coherence Tomography

Abstract

purpose. To evaluate the reproducibility of anterior chamber (AC) angle measurements obtained using anterior segment optical coherence tomography (AS-OCT).

methods. Patients with suspected glaucoma and those with glaucoma, ocular hypertension, or anatomically narrow angles were recruited from the glaucoma service at the National University Hospital, Singapore. All subjects underwent imaging of the nasal, temporal, and inferior AC angles with an AS-OCT prototype under standardized dark and light conditions. For short-term reproducibility analysis, a single observer acquired two sets of images followed by a third set of images acquired by a second observer. The interval between sessions was 10 minutes. For long-term reproducibility analysis, a single observer acquired two sets of images at least 24 hours apart. Images were measured using custom software to determine the AC depth (ACD), angle opening distance at 500 μm (AOD₅₀₀), angle recess area at 500 μm (ARA₅₀₀), and trabecular–iris space area at 500 μm (TISA₅₀₀). The intraclass correlation coefficient (ICC) was calculated as a measure of intraobserver and interobserver reproducibility.

results. Twenty eyes of 20 patients were analyzed for short-term reproducibility, and 23 eyes of 23 patients were analyzed for long-term reproducibility. AC depth measurement demonstrated excellent reproducibility (ICC 0.93–1.00) in both dark and light conditions. For the nasal and temporal quadrants, all AC angle parameters demonstrated good to excellent short-term (ICC 0.67–0.90) and long-term (ICC 0.56–0.93) reproducibility in both dark and light conditions. In the inferior quadrant, reproducibility was lower in all categories of analysis and varied from poor to good (ICC 0.31–0.73).

conclusions. AS-OCT allows quantitative assessment of the AC angle. The reproducibility of AC angle measurements was good to excellent for the nasal and temporal quadrants. The lower reproducibility of measurements in the inferior quadrant may be unique to this prototype due to difficulty in acquiring high-quality images of the inferior angle. Further assessment of the commercially available AS-OCT is needed to clarify this finding.

Optical coherence tomography (OCT) ¹ is a light-based imaging modality that provides high-resolution images of the anterior segment in cross section. ² It allows for an objective assessment of the anterior chamber (AC) angle by a completely noncontact approach and is easy to use after minimal training. These characteristics compare favorably to the current gold standard, gonioscopy, which requires highly trained personnel, is subjective, and involves placing a lens on the eye of the patient. 

Since the first description by Izatt et al. in 1994, ² anterior segment OCT (AS-OCT) has undergone several advances, including the use of 1.3-μm-wavelength light ³ and the development of high-speed imaging at this wavelength. ⁴ These modifications have improved the visualization of AC angle structures and enabled real-time imaging of changes in the angle configuration. Several studies have reported on quantitative measurement of the AC angle width using OCT. ^{5 6} With any modality used for biometry, good reproducibility is essential for the measurements to be meaningful. Karandish et al. ⁷ reported excellent reproducibility for angle measurements using a slit-lamp adapted OCT system. In a recently published study comparing ultrasound biomicroscopy with OCT, ⁵ the reproducibility of AC angle measurements with both modalities was found to be comparable. In this study, we investigated the short- and long-term reproducibility of measurements obtained with a prototype high-speed anterior segment OCT system that was used in a cross-sectional study evaluating the role of OCT in screening for primary angle-closure glaucoma in a Singaporean population. ⁸  

Methods

Informed consent was obtained from all patients, and the study was approved by the Institutional Review Boards of the National University Hospital of Singapore and the Singapore Eye Research Institute. The research was performed in accordance with the tenets of the Declaration of Helsinki. The subject population for this study was a subset selected from a larger study of 203 subjects aged 40 years or older that included patients with suspected glaucoma and patients with glaucoma, ocular hypertension, or primary angle closure. Subjects were recruited from the glaucoma service at the National University Hospital, Singapore and the majority (71%) had a clinical diagnosis of treated or untreated primary angle closure. All subjects underwent gonioscopy and AS-OCT imaging. 

Gonioscopy was performed under dim illumination with a Goldmann two-mirror lens. An angle quadrant was classified as closed on gonioscopy if the iris was in contact with the posterior trabecular meshwork. An individual was classified with angle-closure if one or more of the temporal, inferior, and nasal quadrants of the angle were closed in either eye. 

Imaging of the nasal, temporal, and inferior AC angles was performed with an AS-OCT prototype (Carl Zeiss Meditec, Inc., Dublin, CA) under standardized dark and light conditions. Figure 1is an example of an AS-OCT image of the nasal and temporal angles obtained with this instrument. The prototype used a 1.3-μm-wavelength light with a scan speed of 2000 A-scans per second and a full width-half maximum axial resolution of 17 μm in tissue. The scan depth was 8 mm, and the scan length was 16 mm. An accommodative internal fixation target was used, and the spherical equivalent of the subject’s distance spectacle correction was dialed into the instrument optics so that imaging was performed in a nonaccommodated state. The superior quadrant could not be imaged because the bulky casing and short working distance of the prototype instrument used in this study prevented the operator from lifting the subject’s upper eyelid with a cotton tip applicator or other devices. A commercially available OCT (Visante; Carl Zeiss Meditec) has a greater working distance and may allow the operator to use a device to lift the upper eyelid and image the superior quadrant. The other design difference between the prototype and the OCT was a joystick control for patient positioning in the former versus a motorized chin-rest in the latter. 

For short-term reproducibility analysis, a single observer acquired two sets of images (sessions 1 and 2, respectively), followed by a third set of images (session 3) acquired by a second observer. The subject was repositioned for each session. The interval between each of the three sessions was 10 minutes. The first set of images acquired by the first observer (session 1) was used for analysis of interobserver reproducibility. For long-term reproducibility analysis, a single observer acquired two sets of images (sessions A and B, respectively) with the mean time between sessions being 10.5 weeks (range, 1 day–73 weeks; median, 3 weeks). Each set of images consisted of one image each of the nasal, inferior, and temporal quadrants obtained with the eye in primary gaze. If the lower lid prevented visualization of the inferior angle, the patient was instructed to hold the lid down against the infraorbital rim. 

All images were measured offline by an independent, masked third observer, who used custom software (MathWorks, Natick, MA) ⁹ to determine the AC depth (ACD), angle opening distance at 500 μm (AOD₅₀₀), angle recess area at 500 μm and 750 μm (ARA₅₀₀ and ARA₇₅₀), and trabecular–iris space area at 500 and 750 μm (TISA₅₀₀ and TISA₇₅₀, respectively; Fig. 2 ). Images of the right eye were used; left eye data were included only if data from the right eye were not available. The software was semiautomated. The operator first marked the anterior and posterior corneal surfaces and the anterior iris surface with the mouse so that the image could be corrected for the effects of refraction at the cornea and the fan-shaped scan geometry of the device. The operator then marked the scleral spur, after which the program calculated the aforementioned quantitative parameters. All images of a particular subject were measured at one sitting. The AOD₅₀₀ was defined as the linear distance between the trabecular meshwork and the iris at 500 μm anterior to the scleral spur. ¹⁰ The ARA was defined as the triangular area formed by the AOD₅₀₀ or AOD₇₅₀ (the base), the angle recess (the apex), the iris surface, and the inner corneoscleral wall (sides of triangle). ¹¹ The TISA was defined as the trapezoidal area with the following boundaries: anteriorly, the AOD₅₀₀ or AOD₇₅₀; posteriorly, a line drawn from the scleral spur perpendicular to the plane of the inner scleral wall to the opposing iris; superiorly, the inner corneoscleral wall and inferiorly, the iris surface. ⁵  

The correlation between gonioscopy and angle parameters measured by AS-OCT was calculated using the Pearson correlation coefficient. Analysis of variance was used to calculate the intraclass correlation coefficient (ICC) as a measure of intraobserver and interobserver reproducibility. An ICC of <0.4 indicated poor reproducibility, between 0.4–0.75 indicated fair to good reproducibility, and >0.75 indicated excellent reproducibility. The mean of measurements from paired data sets (sessions 1 and 2 for intraobserver short-term reproducibility, sessions 1 and 3 for interobserver short-term reproducibility, and sessions A and B for intraobserver long-term reproducibility), the difference between measurements from paired data sets and the 95% confidence interval for this difference were also calculated. 

Results

The mean age of patients enrolled in the short-term and long-term reproducibility study groups was 60.8 ± 9.8 and 63.4 ± 10.5 years, respectively. Most of the patients in both study groups had a clinical diagnosis of primary angle-closure glaucoma or anatomically narrow angles (Table 1) . In both groups, the gonioscopic grade of subjects represented the full range of angle widths (Table 2) . The correlation between gonioscopy findings and angle parameters measured by AS-OCT was found to be lower in the inferior quadrant in bright illumination (Table 3) . The number of subjects with a laser peripheral iridotomy was 12 (60%) and 11 (48%) in the short- and long-term reproducibility groups, respectively. 

Short-Term Reproducibility

Twenty eyes (18 right eyes and 2 left eyes) of 20 patients were imaged. All measurements of the nasal and temporal quadrants obtained from images acquired by both observers were included in the analysis. In the inferior quadrant, 11 (13.8%) of 80 measurements obtained from images acquired by the first observer and 8 (21.1%) of 38 measurements obtained from images acquired by the second observer were discarded due to inadequate image quality that caused errors in boundary detection in the angle region by the software program. 

Intraobserver Reproducibility.

AC depth measurement demonstrated excellent reproducibility in both dark (ICC = 0.99) and light conditions (ICC = 1.00). All angle parameters in the nasal and temporal quadrants demonstrated good to excellent reproducibility in both dark and light conditions (ICC range: 0.66–0.90, Table 4 ). In the inferior quadrant, reproducibility was lower and varied from poor to good (ICC range: 0.31–0.59, Table 3 ). The mean of sessions 1 and 2, the difference, and the confidence intervals for the difference between measurements are presented in Table 5 . 

Interobserver Reproducibility.

AC depth measurement demonstrated excellent reproducibility in both dark (ICC = 0.95) and light conditions (ICC = 0.93). All angle parameters in the nasal and temporal quadrants demonstrated good to excellent reproducibility in both dark and light conditions (ICC range: 0.61–0.90, Table 4 ). In the inferior quadrant, reproducibility was lower and varied from poor to good (ICC range: 0.31–0.73, Table 3 ). The mean of sessions 1 and 3, the difference, and the confidence intervals for the difference between measurements are presented in Table 6 . 

Long-Term Reproducibility

Twenty three eyes (17 right eyes and 6 left eyes) were imaged. In the nasal quadrant, all measurements were included in the analysis. Two (2.2%) of 92 measurements of the temporal quadrant and 7 (7.9%) of 89 measurements of the inferior quadrant were discarded because inadequate image quality caused errors in boundary detection in the angle region by the software program. 

AC depth measurement demonstrated excellent reproducibility in both dark (ICC = 0.91) and light conditions (ICC = 0.96). All angle parameters in the nasal and temporal quadrants demonstrated excellent reproducibility (Table 4) . In the temporal quadrant, the reproducibility varied from good to excellent (ICC range, 0.56–0.77). Once again, the reproducibility in the inferior quadrant was lower and ranged from fair to good (ICC range, 0.46–0.60). The mean of sessions A and B, the difference, and the confidence intervals for the difference between measurements are presented in Table 7 . 

Discussion

The reproducibility of quantitative AC measurements by OCT can be influenced by physiological changes in the parameter measured as well as by variations induced by image acquisition and image processing. We were particularly interested in the role of variations induced by the image acquisition process since this device is being evaluated as a screening tool for angle closure, and therefore image acquisition will probably be performed by multiple field operators. We found that the reproducibility of measurements in the nasal and temporal quadrants was good to excellent in all categories of analysis; however, the reproducibility of measurements in the inferior quadrant was lower. The correlation between gonioscopy and OCT angle parameters was also lower in the inferior quadrant. AC depth measurements were less variable than AC angle parameters. 

We attempted to minimize the effect of physiological variations by standardizing illumination and imaging patients in a nonaccommodated state. Two factors influencing the image acquisition process were analyzed: the examiner acquiring the scan and the effect of repeated scan acquisitions. Reproducibility of measurements may also be influenced by the type of measurement algorithm (automated versus semiautomated) and the examiner processing the images. We did not analyze variation due to image processing in our study, since the current software is a fairly cumbersome research version that may not be suitable for clinical use. In addition, it is expected to be modified in the near future toward an ultimate goal of completely automated measurement of images. The reproducibility of angle measurements in the nasal and temporal quadrants was good to excellent, despite the use of a relatively slow, semiautomated measurement process; it can be reasonably expected that the reproducibility would be equal or better with completely automated measurement algorithms. 

Two previous studies have reported on reproducibility of angle measurements with prototype OCT devices different from the one we used. In the study by Karandish et al., ⁷ the OCT device was slit lamp based and had a scan acquisition time of 2 seconds and a longitudinal resolution of 11 μm. The angle parameters measured were the AOD and the AC angle width in degrees (ACA) and the ICC was used as a measure of reproducibility. Measurements were obtained with the OCT scan beam perpendicular to the angle structures, and custom-software was used to measure the images. The study analyzed the effect of variation due to the image-measurement process: Three consecutive images obtained by a single observer were measured five times each by two observers. Both intraobserver and interobserver reproducibility for the AOD were found to be excellent (ICC, 0.99 and 0.95, respectively). Factors that may have led to the slightly decreased reproducibility in our study include the patient population studied (many had very narrow angles in our study, and it is often more difficult to identify the scleral spur in these patients), differences in the device, and different methodologies (first, eyes were analyzed in primary gaze in our study, and hence the angle structures were not perpendicular to the OCT scan beam; and, second, we analyzed variation due to image acquisition and not image measurement). Although it is difficult to separate the effects of image acquisition from that of image measurement, it is more likely that consecutive measurements of a single image are less variable than measurements of two different images. 

The OCT device used in the study by Radhakrishnan et al. ⁵ was slit lamp based with a scan acquisition time of 125 ms and an axial resolution of 8 μm. The angle parameters measured were the same as in the present study, and pooled SD was used as a measure of reproducibility. Measurements were obtained with the OCT scan beam perpendicular to the angle structures and images were measured using the same software as that used in the present study. Variation due to the image-acquisition process was analyzed. Three consecutive images obtained by a single observer were measured by a second independent observer. Intraobserver reproducibility with OCT was found to be comparable to ultrasound biomicroscopy. 

The reproducibility of measurements in the nasal and temporal quadrants in this study was good to excellent for both interobserver and intraobserver short- and long-term reproducibility. The reproducibility of measurements in the inferior quadrant, however, was lower than that in the nasal and temporal quadrants for all categories of analysis. The correlation between gonioscopy findings and inferior quadrant OCT parameters under lighted conditions was also lower. This result can be attributed to characteristics of the particular OCT prototype used in our study. Imaging the inferior AC angle required manipulation of the lower eyelid, which was difficult due to limited space between the instrument and the patient’s face. In addition, it was more difficult to obtain a maximum signal-to-noise ratio while imaging in the vertical meridian, since the raw OCT image displayed on the monitor moved in a counterintuitive perpendicular direction to the OCT scan beam. A poor signal-to-noise ratio may result in suboptimal visualization of anatomic landmarks, such as the scleral spur, resulting in measurement errors. The commercially available anterior segment OCT system (Visante; Carl Zeiss Meditec) does not have these limitations, and imaging of the inferior angle may be better with that instrument. 

AC depth measurements were more reproducible than AC angle width parameters. We believe that there are two factors that may account for this difference. First, there appears to be relatively greater physiologic variation in the AC angle configuration than in AC depth with changes in pupil size. Although we made every effort to minimize this effect by controlling accommodation and ambient illumination, it is not possible to control the pupil size precisely. Second, the need to identify the scleral spur for measurement of AC angle parameters is likely an important source of variability in these measurements. The measurement of AC depth is based on other landmarks (corneal surface and anterior lens surface) that are perpendicular to the OCT scan beam and are better delineated than the scleral spur. The reproducibility of AC depth measurements has been reported with partial coherence interferometry, which is essentially the A-scan equivalent of OCT, and our results are in close agreement with these studies. ^{12 13}  

In conclusion, the OCT prototype used in this study demonstrated excellent interobserver and intersession reproducibility for AC depth measurements and good to excellent interobserver and intersession reproducibility for angle parameters in the nasal and temporal quadrants. Lower reproducibility of measurements in the inferior quadrant is probably due to characteristics of the prototype used in this study, and additional research is needed to assess reproducibility of the commercial OCT devices as they come to market. 

 Supported by National University of Singapore: Grants R-191000007101 and R-191000007112; and technical support and loan of the AS-OCT machine from Carl Zeiss Meditec.

 Submitted for publication September 18, 2006; revised January 27, 2007; accepted May 14, 2007.

 Disclosure: S. Radhakrishnan, Carl Zeiss Meditec (C); J. See, None; S. D. Smith, None; W.P. Nolan, Carl Zeiss Meditec (R); Z. Ce, None; D. S. Friedman, Carl Zeiss Meditec (C, R); D. Huang, Optovue Inc. (C, F, I, P) Carl Zeiss Meditec (F); Y. Li, Carl Zeiss Meditec (F); T. Aung, Carl Zeiss Meditec (F, R); P.T.K. Chew, 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.

 

Figure 1.

 

AS-OCT image of the anterior chamber.

Figure 2.

 

Measurement of quantitative angle parameters using AS-OCT. Illustrated are (A) angle opening distance (AOD), (B) angle recess area (ARA), and (C) trabecular–iris space area (TISA).

Table 1.

 

Glaucoma Diagnosis in Study Subjects Participating in the AS-OCT Reproducibility Study

Clinical Diagnosis	Short-Term Reproducibility	Long-Term Reproducibility
PAC or PACG	9 (45.0%)	10 (43.5%)
Narrow angles	5 (25.0%)	9 (39.1%)
POAG	4 (20.0%)	0
OHTN	1 (5.0%)	0
Normal	1 (5.0%)	4 (17.4%)
Total	20	23

 PAC, primary angle closure; PACG, primary angle closure glaucoma; POAG, primary open-angle glaucoma; OHTN, ocular hypertension.

Table 2.

 

Gonioscopy Findings in Study Subjects Participating in the AS-OCT Reproducibility Study

	Shaffer Grade on Gonioscopy
		4	3	2	1	0
Short-term subjects (n = 20)
Nasal quadrant	4 (20%)	5 (25%)	4 (20%)	4 (20%)	3 (15%)
Temporal quadrant	4 (20%)	8 (40%)	4 (20%)	2 (10%)	2 (10%)
Inferior quadrant	3 (15%)	5 (25%)	2 (10%)	4 (20%)	6 (30%)
Long-term subjects (n = 23)
Nasal quadrant	2 (9%)	5 (22%)	2 (9%)	8 (35%)	6 (26%)
Temporal quadrant	2 (9%)	5 (22%)	5 (22%)	7 (30%)	4 (17%)
Inferior quadrant	2 (9%)	2 (9%)	5 (22%)	3 (13%)	11 (48%)

Table 3.

 

Pearson Correlation Coefficients between Gonioscopy Findings and AS-OCT Angle Parameters

Parameter	Temporal Quadrant			Nasal Quadrant			Inferior Quadrant
		Dark	Light	Dark	Light	Dark	Light
AOD 500	0.62	0.68	0.70	0.76	0.62	0.39
ARA 500	0.55	0.69	0.55	0.70	0.57	0.18
ARA 750	0.62	0.71	0.65	0.74	0.60	0.30
TISA 500	0.55	0.70	0.55	0.71	0.57	0.23
TISA 750	0.63	0.63	0.66	0.74	0.61	0.33

 AOD, angle opening distance (μm); ARA, angle recess area (mm²); TISA, trabecular-iris space area (mm²).

Table 4.

 

Intraclass Correlation Coefficients for AC Angle Parameters in Study Subjects Participating in the AS-OCT Reproducibility Study

Quadrant	Parameter	Short-Term Reproducibility							Long-Term Reproducibility
				Intraobserver			Interobserver			Intraobserver
				Dark	Light	Dark	Light	Dark	Light
Temporal	AOD 500	0.78	0.81	0.80	0.90	0.77	0.75
	ARA 500	0.75	0.78	0.67	0.85	0.56	0.58
	ARA750	0.77	0.82	0.73	0.87	0.67	0.65
	TISA 500	0.79	0.81	0.78	0.87	0.68	0.69
	TISA 750	0.79	0.84	0.81	0.89	0.74	0.73
Nasal	AOD 500	0.76	0.88	0.75	0.87	0.93	0.84
	ARA 500	0.66	0.84	0.61	0.82	0.89	0.82
	ARA750	0.72	0.88	0.66	0.83	0.88	0.86
	TISA 500	0.72	0.88	0.68	0.86	0.92	0.85
	TISA 750	0.74	0.90	0.71	0.87	0.92	0.87
Inferior	AOD 500	0.51	0.59	0.73	0.67	0.54	0.55
	ARA 500	0.36	0.31	0.68	0.33	0.47	0.39
	ARA750	0.43	0.35	0.68	0.32	0.51	0.44
	TISA 500	0.42	0.47	0.72	0.53	0.58	0.46
	TISA 750	0.47	0.52	0.73	0.55	0.60	0.49

 Abbreviations and units are as described in Table 3 .

Table 5.

 

Intraobserver Short-Term Reproducibility

Quadrant	Parameter	Dark					Light
				Mean	Difference	95% CI	Mean	Difference	95% CI
	AC Depth	2.21	−0.00	−0.03–0.03	2.22	0.01	−0.07–0.10
Temporal	AOD 500	183.3	−4.0	−52.3–44.3	328.4	30.6	−17.2–78.3
	ARA 500	0.057	0.009	−0.019–0.019	0.119	0.012	−0.011–0.034
	ARA750	0.111	0.000	−0.030–0.030	0.214	0.019	−0.013–0.051
	TISA 500	0.052	−0.001	−0.018–0.016	0.107	0.012	−0.006–0.03
	TISA 750	0.106	−0.001	−0.029–0.026	0.202	0.018	−0.009–0.046
Nasal	AOD 500	147.5	7.7	−34.5–49.9	269.7	20.4	−12.8–53.5
	ARA 500	0.042	0.004	−0.014–0.022	0.083	0.007	−0.009–0.023
	ARA750	0.085	0.003	−0.025–0.031	0.164	0.014	−0.008–0.037
	TISA 500	0.037	0.002	−0.013–0.017	0.077	0.008	−0.005–0.020
	TISA 750	0.080	0.001	−0.025–0.026	0.158	0.015	−0.004–0.034
Inferior	AOD 500	162.4	−0.4	−71.1–70.4	299.9	32.8	−24.8–90.4
	ARA 500	0.050	−0.002	−0.037–0.034	0.093	0.010	−0.023–0.044
	ARA750	0.094	−0.008	−0.061–0.045	0.181	0.016	−0.029–0.061
	TISA 500	0.044	−0.002	−0.033–0.028	0.088	0.013	−0.015–0.041
	TISA 750	0.089	−0.009	−0.057–0.039	0.176	0.019	−0.021–0.059

 Reproducibility was assessed by mean, difference and 95% confidence intervals for difference between repeat measurements taken 10 minutes apart. Abbreviations and units are as described in Table 3 .

Table 6.

 

Interobserver Short-Term Reproducibility

Quadrant	Parameter	Dark					Light
				Mean	Difference	95 % CI	Mean	Difference	95 % CI
	AC Depth	2.22	0.03	−0.05–0.08	2.25	−0.02	−0.03–0.00
Temporal	AOD 500	169.2	−18.2	−60.5–24.2	291.5	−6.2	−43.3–30.9
	ARA 500	0.055	0.011	−0.024–0.021	0.101	−0.006	−0.025–0.013
	ARA 750	0.105	−0.006	−0.035–0.024	0.185	−0.011	−0.039–0.016
	TISA 500	0.050	−0.004	−0.022–0.014	0.090	−0.005	−0.020–0.01
	TISA 750	0.100	−0.008	−0.034–0.018	0.173	−0.010	−0.034–0.014
Nasal	AOD 500	159.0	19.2	−24.2–62.6	238.1	−13.3	−48.7–22.0
	ARA 500	0.048	0.010	−0.010–0.029	0.075	−0.002	−0.018–0.015
	ARA 750	0.095	0.013	−0.017–0.044	0.145	−0.007	−0.031–0.017
	TISA 500	0.044	0.008	−0.009–0.025	0.070	−0.000	−0.014–0.015
	TISA 750	0.091	0.012	−0.016–0.039	0.139	−0.005	−0.027–0.017
Inferior	AOD 500	172.3	4.4	−58.0–66.8	268.5	27.8	−24.4–79.9
	ARA 500	0.047	−0.003	−0.028–0.023	0.088	0.015	−0.012–0.051
	ARA 750	0.101	−0.003	−0.043–0.038	0.168	0.025	−0.016–0.066
	TISA 500	0.044	−0.008	−0.025–0.022	0.080	0.017	−0.012–0.045
	TISA 750	0.098	−0.002	−0.039–0.036	0.160	0.023	−0.015–0.061

 Data are assessed by mean, difference and 95% confidence intervals for difference between repeat measurements taken 20 minutes apart by 2 different observers. Abbreviations are as in Table 3 .

Table 7.

 

Intraobserver Long-Term Reproducibility

Quadrant	Parameter	Dark					Light
				Mean	Difference	95 % CI	Mean	Difference	95 % CI
	AC Depth	2.29	−0.04	−0.03–0.10	2.292	0.04	0.00–0.08
Temporal	AOD 500	138.9	21.5	−12.3–55.4	264.0	35.4	−5.7–76.5
	ARA 500	0.054	0.020	0.000–0.039	0.097	0.018	−0.003–0.039
	ARA 750	0.097	0.025	−0.001–0.051	0.172	0.027	−0.002–0.056
	TISA 500	0.045	0.001	−0.002–0.027	0.107	0.012	−0.006–0.029
	TISA 750	0.087	0.016	−0.006–0.038	0.163	0.024	−0.001–0.049
Nasal	AOD 500	119.9	15.2	−5.6–36.0	226.4	6.4	−34.0–46.8
	ARA 500	0.039	0.002	−0.010–0.014	0.076	0.005	−0.016–0.026
	ARA 750	0.074	0.004	−0.013–0.020	0.143	0.010	−0.019–0.038
	TISA 500	0.034	0.001	−0.010–0.012	0.077	0.008	−0.005–0.020
	TISA 750	0.070	0.003	−0.013–0.018	0.131	0.005	−0.019–0.029
Inferior	AOD 500	117.2	40.1	−7.8–88.0	201.6	2.9	−47.7–53.5
	ARA 500	0.026	0.013	−0.003–0.029	0.066	0.008	−0.019–0.036
	ARA 750	0.062	0.023	−0.003–0.049	0.1304	0.0103	−0.031–0.052
	TISA 500	0.024	0.007	−0.003–0.026	0.088	0.013	−0.015–0.041
	TISA 750	0.060	0.022	−0.002–0.046	0.124	0.010	−0.029–0.048

 Data were assessed by mean, difference and 95% confidence intervals for difference between repeat measurements taken at least 1 day apart. Abbreviations and units are as described in Table 3 .