June 2008
Volume 49, Issue 6
Free
Retina  |   June 2008
Precision and Reliability of Retinal Thickness Measurements in Foveal and Extrafoveal Areas of Healthy and Diabetic Eyes
Author Affiliations
  • Gijsbrecht J. M. Tangelder
    From the Departments of Ophthalmology and
  • Rob G. L. Van der Heijde
    Physics and Medical Technology, VU University Medical Center, Amsterdam, The Netherlands.
  • Bettine C. P. Polak
    From the Departments of Ophthalmology and
  • Peter J. Ringens
    From the Departments of Ophthalmology and
Investigative Ophthalmology & Visual Science June 2008, Vol.49, 2627-2634. doi:10.1167/iovs.07-0820
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Gijsbrecht J. M. Tangelder, Rob G. L. Van der Heijde, Bettine C. P. Polak, Peter J. Ringens; Precision and Reliability of Retinal Thickness Measurements in Foveal and Extrafoveal Areas of Healthy and Diabetic Eyes. Invest. Ophthalmol. Vis. Sci. 2008;49(6):2627-2634. doi: 10.1167/iovs.07-0820.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To determine the precision and reliability of retinal thickness measurements with an optical coherence tomograph (Stratus OCT 3; Carl Zeiss Meditec, Dublin, CA) and a retinal thickness analyzer (RTA; Talia Technology Ltd., Neve-Ilan, Israel) in foveal, parafoveal, and perifoveal areas.

methods. Three measurements of all areas were performed within 1 hour on the same day with each instrument in the eyes of healthy volunteers and diabetic patients. The latter group was divided into eyes with and without macular edema.

results. Measurement precision, expressed as the 95% limits of agreement (LA95%), was significantly higher (i.e., a lower LA95%, P < 0.01) for the OCT in comparison to the RTA in virtually all areas of the retina. Moreover, measurement reliability, expressed as the intraclass correlation coefficient, was high with the OCT (>0.90) and moderate to low with the RTA (0.26–0.89). A direct influence of macular edema itself on measurement precision of para- and perifoveal areas was found in the OCT measurements.

conclusions. The high measurement precision and reliability of the OCT suggests that this instrument is currently the most suitable technique for detection and follow-up of diabetic macular edema. When macular edema is present, the OCT can reliably detect changes of at least 36 μm at the fovea, 55 μm in parafoveal areas below a thickness of 744 μm, and 42 μm in perifoveal areas below a thickness of 1011 μm.

Macular edema is an important cause of decreased visual acuity in patients with diabetes mellitus. At present, the only evidence-based treatment is focal laser photocoagulation. Other forms, such as surgical vitrectomy or intravitreous triamcinolone injection may be beneficial in selected cases. A third group, the antiangiogenesis agents, is currently under investigation. 1 Until recently, the presence and severity of macular edema could only be detected as an increase in retinal thickness as observed by an experienced ophthalmologist through fundus biomicroscopy or stereographic fundus photography. Alternatively, it can also be detected as leakage of fluorescent dye during fluorescein angiography, which is an invasive method of examination that does not yield any information about retinal thickness. The necessity for photocoagulation is determined by the ophthalmologist based on the extent of increased retinal thickness due to edema in nine areas, as defined by the ETDRS (Early Treatment of Diabetic Retinopathy Study). 2 Stereophotography is more reliable for detecting an increase in retinal thickness, but rather laborious and time consuming and therefore not routinely performed in clinical practice, making fundus biomicroscopy the clinical mainstay for detecting retinal thickening due to macular edema. However, because these techniques are subjective, even for an experienced ophthalmologist it can be difficult to assess the presence and extent of macular edema, making both the clinical diagnosis and an evaluation for the necessity of treatment sometimes arduous. 
Recently, two instruments have been developed that quantitatively measure macular thickness: the optical coherence tomographer (Stratus OCT 3; Carl Zeiss Meditec, Dublin, CA) and the retinal thickness analyzer (RTA; Talia Technology Ltd., Neve-Ilan, Israel). So far, it has been shown that both instruments provide reproducible measurements of retinal thickness at the foveal center (the umbo) and fovea (central 300 μm radius). 3 4 5 6 7 8 9 In three studies, investigators have also evaluated reproducibility of measurement of extrafoveal ETDRS areas by the OCT. 10 11 12 However, to our knowledge, no previous reports have evaluated measurement precision for both instruments in eyes of healthy subjects and diabetic patients with or without macular edema. 
The purpose of the present study was to evaluate measurement precision and reliability between both instruments in all ETDRS areas in three groups: healthy subjects and patients with or without macular edema. Furthermore, we chose to express measurement precision as 95% limits of agreement (LA95%,) 13 because this provides the measurement scale for detection and follow-up in individual subjects. 14 Measurement reliability was determined by calculating the intraclass correlation coefficient (ICC). 
Materials and Methods
Study Groups
Measurements of both eyes were performed in healthy volunteers and patients with diabetes mellitus. In both groups, only subjects of 17 to 70 years of age were considered eligible. All 28 eyes of 14 healthy volunteers had an unremarkable ocular history and normal fundus appearance on ophthalmoscopic evaluation. The average age in healthy volunteers was 49 ± 10 years. 
The patient group consisted of 21 patients with type 1 and 20 with type 2 diabetes, with an average age of 51 ± 15 years. The exclusion criteria per eye were other retinal disease before measurement (e.g., retinal detachment or a vascular occlusion), ocular surgery other than cataract extraction, and significant opacification of ocular media precluding reliable fundus examination. Four eyes were excluded for measurement: one because of vitreous hemorrhage, another for prior branch retinal vein occlusion, the third because of previous rhegmatogenous retinal detachment, and the fourth because evaluation and measurement of the posterior pole were unattainable due to mature cataract. Nine eyes of five diabetic patients were pseudophakic. 
All eyes of the diabetic patients were divided into two groups: eyes with and eyes without macular edema. The former was defined as retinal thickening within two disc diameters of the center of fixation, as observed by a retinal specialist through fundus biomicroscopy. The severity of macular edema was subdivided into clinically significant or clinically nonsignificant, based on the criteria proposed by the ETDRS study group. 2 The macular edema group consisted of 21 eyes of 15 patients. Clinically significant edema was observed in 16 eyes and clinically nonsignificant edema in the remaining 5. In this group, 10 eyes showed signs of background diabetic retinopathy, 4 eyes preproliferative diabetic retinopathy, 5 eyes proliferative retinopathy, and 2 eyes background retinopathy after panretinal photocoagulation therapy. The group without macular edema consisted of 57 eyes of 31 patients, in which 19 eyes had no signs of diabetic retinopathy, 22 eyes showed signs of background diabetic retinopathy, 3 eyes had preproliferative diabetic retinopathy, 1 eye had proliferative retinopathy, and 12 eyes had undergone panretinal photocoagulation because of proliferative retinopathy. 
Written informed consent was obtained from all participants. The study adhered to the tenets of the Declaration of Helsinki and was approved by the institutional review board of our hospital. 
Measurement Protocol
Before measurement, both pupils were dilated with phenylephrine 0.5% and tropicamide 5% eyedrops. Each measurement session started with positioning a subject’s head as upright as possible on the chin and head rest of the instrument. All subjects were measured three times with both the OCT and RTA within 1 hour on the same day by the same operator (GJMT). A minute’s pause was observed between measurements. The right eye was scanned first and the left eye second, first with one instrument (e.g., OCT) followed by the second. Both the first and second measurements were performed with each instrument in random order and equal amounts. 
Optical Coherence Tomography
The OCT data were collected with the Stratus 3 (software ver. 2.0; Carl Zeiss Meditec), which works on the principle of low-coherence interferometry. Employing an 840-nm superluminescent diode laser, a linear cross section of the macula (B-scan) is obtained through the center of fixation. We used the Fast Macular Thickness scan protocol provided by the machine software for all measurements. With this scan protocol 6 B-scans of 6-mm length were obtained, traversing the center of fixation in the macula at 30° separate angles. Each B-scan consisted of 128 longitudinal A-scans. The retinal boundaries are automatically identified by the machines analysis software. All scans were evaluated for boundary misalignment by the scan operator (GJMT). No misalignment of these boundaries was observed in any of the performed scans, and therefore all scans were included for analysis. Subsequently, the 6 B-scans obtained were recalculated by the machines software, to generate a topographical map of the macula, automatically divided into nine areas, as defined by the ETDRS-study group: one central area, four parafoveal areas, and four perifoveal areas (designated A1 to A9 in Fig. 1a ). 2 In addition, we calculated the parafoveal average (P1) and perifoveal average (P2) for all four parafoveal areas (A2 to A5) and all four perifoveal areas (A6 to A9). 
Retinal Thickness Analyzer
The principle of the RTA version A is based on a 3-mm slit laser beam of 540-nm wavelength skewed in the direction of the retina. Reflections from the internal limiting membrane and retinal pigment epithelium are recorded by a digital video camera. An optical cross section of the retina is thus acquired as a 3-mm line scan. Five areas are scanned, each scan area consisting of 16 line scans in a 3 × 3-mm area. Scan validity was automatically tested by the instruments software analysis algorithm for image clarity and positioning. Invalid scans occurred in only one eye of a diabetic patient who had undergone a Nd:YAG laser capsulotomy. However, because there were no failures with the OCT in the same eye, it was included for analysis as a missing value for the RTA. All scanned areas in all other eyes were considered to be valid according to the instruments analysis algorithm and included for analysis. The first area is centered on the fovea, the other four in a nonoverlapping square, with reference to the central scan. The scanned areas are subsequently recalculated by the machine software (software ver. 4.11B) into a topographical map of the macula in a 6 × 6-mm square fashion, with a 7 × 7 grid of average retinal thickness values (Fig. 1b) . This 7 × 7 grid was subdivided into nine ETDRS-like areas (designated A1 to A9, similar to the OCT map), for which the average thickness values were recalculated to facilitate comparison with the OCT (Fig. 1b) . The parafoveal (P1) and perifoveal (P2) average were calculated in a similar fashion. 
Scan Failures
Scan failures of ETDRS areas (A1–A9) were arbitrarily defined as significant if more than 50% of the data points in an area were missing. If significant scan failure of an area occurred, its value was inserted as missing (blank) in the statistical analysis. No scan failures were observed with the OCT. As mentioned, one complete scan failure of all areas occurred with the RTA in one eye. No significant scan failures were observed with the RTA in eyes of healthy volunteers. However, incomplete area scans were observed with the RTA in diabetic eyes without macular edema in three perifoveal areas (A6–A9; 0.4%). Scan failures in these eyes did not occur in the parafoveal areas (A2–A5) and only one (0.6%) occurred for average foveal thickness (A1). In eyes with macular edema, scan failures were observed in five perifoveal areas (2.0%) and one parafoveal area (0.4%) and twice for average foveal thickness (3.2%). 
Measurement Precision and LA95%
Measurement precision of both instruments was expressed as 95% limits of agreement for each ETDRS area (LA95%). 13 This value provides the scale in which an instrument can detect changes over time within a subject. 14 The smaller these limits, the higher the measurement precision. The LA95% values were calculated as follows:  
\[\mathrm{LA}_{95\%}\ {=}\ 1.96\ {\cdot}\ \sqrt{2\ {\cdot}\ {\bar{{\sigma}}}^{2}_{\mathrm{within-patient}}}\]
where σ̄within-patient is the average within-patient SD, which was calculated for each patient in the usual way:  
\[{\sigma}_{\mathrm{within-patient}}\ {=}\ \sqrt{{{\sum}_{i{=}1}^{n}}\left(\frac{(x_{\mathrm{i}}\ {-}\ {\bar{x}})^{2}}{n\ {-}\ 1}\right)}\ \mathrm{and}\ n\ {=}\ 3\]
 
The LA95% provides the absolute difference between measurements that must be exceeded to detect statistically significant changes over time due to development or progression of disease. A test for heteroscedasticity (i.e., nonuniform within-patient variance) was performed for all areas in each group of eyes by plotting the σwithin-patient against the within-patient mean and testing for any correlation. In this test, if no significant Pearson’s correlation coefficient (P > 0.05) is found, uniformity of within-patient variance is likely, and a single LA95% is valid. If heteroscedasticity is present however, the interval for which the LA95% value is valid should be specified, with a preferable overestimation of measurement variance within the interval. 
Statistics
Nonparametric statistical tests were chosen for comparison between groups, because of unequal variances. A Kruskal-Wallis test was used for comparison of multiple groups. A Mann-Whitney U test was used for unpaired data and a Wilcoxon test for paired data. Determination of the variance components between and within subject eyes was performed by using a restricted maximum-likelihood estimation. The ICC was calculated in the usual way: ICC = (total variance − within patient variance)/(total variance). All statistical analysis was performed with commercial software (SPSS software ver. 12.0.1; SPSS Inc, Chicago, IL). 
Results
Table 1shows the mean retinal thicknesses (left columns) and standard deviation (right columns) for each study group as measured with both instruments for each ETDRS area separately (A1–A9) and also for the parafoveal (P1) and perifoveal (P2) average. First, retinal thickness values for all areas were, on average, higher when measured with the OCT (P < 0.001 for all areas in all groups). Second, an increase in average retinal thickness was measured with the OCT in all areas (A1–A9, P1 and P2) for eyes with macular edema. In contrast, only average foveal thickness (A1) showed a clear increase in its mean value beyond the normal range when measured with the RTA. Regarding extrafoveal areas (A2–A9), only perifoveal temporal thickness (A7) was found as significantly increased. This difference, however, was very low (165 μm versus 181 μm, respectively) and well within the normal range. Furthermore, all other extrafoveal areas, as well as the parafoveal (P1) and perifoveal average (P2) in eyes with macular edema, were found to be equal to or significantly lower than the same areas in both other groups with the RTA. The standard deviations between the OCT and RTA were similar in eyes of healthy volunteers and diabetic patients without macular edema, the SD was lower for the RTA in comparison to the OCT in diabetic eyes with macular edema. A normal anatomic configuration of retinal thicknesses was observed with both instruments in eyes of healthy volunteers and diabetic patients without macular edema. 
Table 2shows the LA95% for each study group as measured with both instruments. The LA95% for the OCT were significantly lower (i.e., better) compared with the RTA for all areas in each study group (P ≤ 0.05), except for average foveal thickness (A1; P = 0.07), and parafoveal superior thickness (A2; P = 0.20) in eyes with macular edema. However, the parafoveal average (P1) was significantly smaller (P < 0.001). More interestingly, for the OCT a marked increase in LA95% was observed in eyes with macular edema in comparison to the other groups for all areas (P < 0.01). For the RTA, measurement precision in eyes with macular edema was equal in comparison to the other two groups (P ≥ 0.10; Table 2 ; the two rightmost columns). 
Moreover, as is shown in Figure 2 , heteroscedasticity of OCT measurements was found in eyes with macular edema both for parafoveal areas (P < 0.001, Fig. 2a ) and perifoveal areas (P = 0.02, Fig. 2c ). Average foveal thickness (A1) was not significantly associated with heteroscedasticity in this group (P = 0.29). Heteroscedasticity was also present in RTA measurements of perifoveal areas in this group (P < 0.001, Fig. 2d ), but not in parafoveal areas (P = 0.12, Fig. 2b ) or average foveal thickness (P = 0.35). No signs of heteroscedasticity were found in eyes of healthy volunteers and diabetic patients without macular edema with either instrument. Based on the findings in Table 2and Figure 2 , there appears to be a direct influence of macular edema on measurement precision in OCT measurements of extrafoveal areas. This influence must be accounted for when applying the LA95% values. An equation for linear fit, as shown in Figure 2a , allowed us to determine the valid interval for the LA95% provided for eyes with macular edema in Table 2 . The average LA95% for all parafoveal areas (P2) in eyes with macular edema is 29 μm. The corresponding σwithin-patient for this value is 10.5 μm, which intercepts the linear fit in Figure 2aat an average retinal thickness of approximately 463 μm. Hence, for parafoveal retinal thickness of 463 μm or less, the OCT can detect changes of 29 μm or more reliably. Furthermore, virtually all σwithin-patient values shown in Figure 2aare 20 μm or less (corresponding to an LA95% of 55 μm), which intercepts the fit at approximately 744 μm. For perifoveal areas, the OCT can detect changes of 25 μm or more in eyes with macular edema below an average of 573 μm, and 42 μm or more up to 1011 μm thickness. 
The most striking finding in the present study is perhaps the high ICC found for the OCT measurements in comparison to the RTA measurements (Table 3) . The OCT values were between 0.90 and 1.0 for all areas in each study group, which indicates a high measurement reliability. In contrast, the ICCs for the RTA were below 0.90 for all areas in each group. The RTA’s ICC is reasonable (0.89) for average foveal thickness in eyes with macular edema, but moderate for parafoveal (0.67) and poor for perifoveal (0.59) areas. 
Discussion
Measurement precision was determined both for the OCT and RTA in eyes of healthy subjects and diabetic patients, with division of the latter group into eyes with and without macular edema. For each eligible eye all nine ETDRS areas were measured three times within 1 hour. The resultant LA95% values were generally lower for the OCT in comparison to the RTA, indicative that the OCT’s measurement precision is higher. Moreover, the OCT’s measurement reliability, expressed as ICCs, was clearly higher in comparison to the RTA, with the possible exception of average foveal thickness in eyes with macular edema. A significant increase in retinal thicknesses in eyes with macular edema was observed in all areas by the OCT, but only for average foveal thickness by the RTA. It was also observed that partial scan failure occurred frequently in RTA measurements, particularly in perifoveal areas, but not in the OCT measurements. However, despite its high measurement precision and reliability, a marked influence of macular edema on measurement precision was observed in the OCT measurements. 
In contrast to the OCT, the RTA showed increased mean retinal thickness values due to edema only at the foveal center (average foveal thickness or area A1). One extrafoveal area was also tested as significantly increased (A7), but the mean difference observed was small and well within the normal range. Furthermore, the mean value of all other extrafoveal areas, as well as the para- and perifoveal average remained equal to or even lower in the group of eyes with macular edema. This matches with our clinical experience that the RTA often seems unable to detect an increase in retinal thickness adequately because of edema in extrafoveal areas. To illustrate, two examples are depicted of eyes with macular edema in Figure 3 , in which we believe the RTA was unable to detect macular edema adequately in extrafoveal areas. A possible explanation of why the RTA has difficulty in detecting macular edema in extrafoveal areas may be offered by the imaging technique. The skew angle in which light is projected may be too large in extrafoveal areas, resulting in distortion of the recorded images. However, because only clinical biomicroscopy was used to detect the presence and extent of macular edema, a full evaluation of measurement accuracy and possible false positives or negatives is beyond the scope of our present study. 
As shown in Table 2 , the LA95% values for the OCT were clearly lower compared to those of the RTA, especially for para- and perifoveal areas. In contrast, previous reports that provide a direct comparison in measurement precision between both instruments found them to be similar. 9 10 A possible explanation for this difference is the focus of these studies on measurement precision at the foveal center. In our study, the OCT’s LA95% for average foveal thickness (A1) was found to be higher compared with extrafoveal areas (A2–A9), most prominently in patients with macular edema. The RTA’s measurement precision for average foveal thickness did not differ greatly from that in extrafoveal areas. In our opinion, the lower precision of the OCT’s measurement of the foveal center relative to extrafoveal areas is attributable to the scan protocol. Slight eccentricity of fixation has a strong influence on the average thickness found, because the scanned area is small. This, along with the fact that an older version of the OCT, with less precise analysis algorithms, was often used in previous studies, may explain the differences found. 
For the OCT, a comparison of both the LA95% and ICC values between the present study and previous reports on measurement precision and reliability is presented in Table 4 . Two other reports shown in Table 4have evaluated measurement precision for the OCT (Paunescu et al. 11 and Polito et al. 12 ). The LA95% values of these reports are somewhat higher in comparison to our findings, which are more in agreement with the findings of Massin et al., 10 who evaluated measurement precision for a previous version of the OCT. Given the overall high-measurement precision in all these reports, a possible explanation for the differences found between them may be simple over- or underestimation due to small sample sizes, which we feel is also reflected in the variation of LA95% observed between comparative areas in Table 4 . For example, the LA95% for area A7 is 7 μm in our study, 24 μm in the report by Polito et al., 12 10 μm in the report by Paunescu et al., 11 and 4 μm in the report by Massin et al. 10 The ICCs reported by Paunescu et al. 11 are somewhat lower in comparison with our findings, but also are more variable. Apart from larger LA95% values, this difference may also be due to a greater variance in the reported standard deviations for comparable ETDRS areas, particularly for extrafoveal areas. For example, Paunescu et al. report an ICC of 0.57 for area A6, 0.97 for area A7, and 0.81 for area A8, whereas the corresponding LA95% values are 11, 10, and 9 μm, respectively (Table 4) . In contrast, Massin et al. 10 found excellent reliability for a previous version of the OCT technology, more in agreement with our findings. 
For the RTA, the LA95% found for average foveal thickness of healthy subjects (LA95% = 33 μm; Table 2 ) are in agreement with Zeimer et al. 4 and Weinberger et al., 5 who report a σwithin-patient of 11.5 and 10.6 μm, corresponding to LA95%s of 32 and 29 μm for the foveal center, respectively. Furthermore, the LA95% for average foveal thickness in healthy subjects and diabetic patients are also similar to those reported by Oshima et al., 9 who found a σwithin-patient of 16 μm for average foveal thickness in healthy subjects and a σwithin-patient of 27 μm in diabetic patients, with or without macular edema combined, which corresponds to LA95%s of 44 and 75 μm, respectively, with no significant difference in measurement precision between both groups. Hence, the LA95%s for average foveal thickness (area A1) in the present study are in agreement with those in previous reports both for the OCT and RTA. Although the calculated ETDRS areas are not identical, they are comparable, because measurement precision can be expected to behave relatively independent of the average retinal thickness. 
The high ICCs found for the OCT (>0.90; Table 3 ) reflect the instrument’s high reliability, which we believe agrees with the values reported by both Paunescu et al. for the OCT (Stratus 3; Carl Zeiss Meditec) and Massin et al. for previous OCT technology (see also the second paragraph above). 10 11 In contrast, the ICC values found for the RTA are moderate (0.60–0.89) to poor (0.30–0.59) in comparison to the OCT, particularly for extrafoveal areas. Zou et al. 15 report an ICC for the RTA of 0.95 for average foveal thickness in 24 eyes of 24 healthy Chinese subjects, which is in contrast with our findings (0.50; Table 3 ). This difference is mainly due to the smaller total variance (i.e., square SD) of our study in comparison to their report (SD of 14 μm vs. 26 μm, respectively). In addition, in their method for calculation of the ICC only two scans were used in a two-way random effects model. The lower number of measurements (i.e., two instead of three) could cause an underestimation of the within-patient variance (i.e., square σwithin-patient) with possible overestimation of the ICC. Furthermore, they report excluding images with poor quality in their general survey, and thus possibly also for their assessment of reliability. The influence of total variance on the ICC is even more clear when in our study the ICC of average foveal thickness (A1) in healthy eyes (0.50) is compared to that in eyes with macular edema (0.89, Table 3 ). Although there is a difference in σwithin-patient values between both groups (12 and 19 μm, respectively) this difference was not found to be statistically significant (P = 0.14; Table 2 ). Hence, the difference in ICC is mainly due to a difference in total variance. This can be seen in Table 1 , where the SD of average foveal thickness (A1) between these two groups shows a marked difference (14 μm vs. 65 μm, P < 0.001), which can be expected if macular edema influences macular thickness. This, in our opinion, underlines the importance of providing measurement precision values as an absolute value of the within patient variance (or σwithin-patient) or a modification thereof (such as the 95% limits of agreement), because the ICC alone can be misleading. To our knowledge, no previous studies have assessed measurement reliability for extrafoveal areas in both instruments. Our results clearly show the OCT to be more reliable and precise for detecting the presence of macular edema before the foveal center is involved and visual acuity is affected, which in our opinion makes it the preferred choice for detection and follow-up of diabetic macular edema. 
However, despite the OCT’s high reliability, a direct influence of macular edema on measurement precision itself (i.e., heteroscedasticity) was found in parafoveal and, to a lesser extent, in perifoveal areas. To our knowledge, this is a novel finding that has not been reported previously. The effect of heteroscedasticity due to edema may be due to the scan protocol used. The location at which the scan lines intersect an area of edema may vary due to ocular movement and positioning. The limited number of three linear scans per area may thus give rise to variation in the average thickness that is measured. Increasing the number of line scans per area may limit this effect. Because heteroscedasticity is present, the OCT’s LA95%s for extrafoveal areas are only valid in a specified measurement range. As explained in the Results section, the OCT can reliably detect changes in parafoveal thickness of 29 μm or more below a mean of 463 μm, and of 55 μm below a mean of 744 μm. In the perifoveal areas, changes of 25 μm or more can be detected up to a mean of 573 μm, and of 42 μm up to 1000 μm. 
In the present study, we provide a direct comparison of measurement precision and measurement reliability of retinal thickness measurements obtained with the OCT and RTA in eyes of healthy subjects and diabetic patients. The latter group was divided into eyes with and those without macular edema. For all studied groups, the OCT showed a higher measurement precision and reliability in comparison to the RTA, which in our opinion, makes it the method of choice for detection and follow-up of diabetic macular edema. However, macular edema itself was found to directly influence the OCT’s measurement precision in extrafoveal areas. The clinician should take this effect into account when evaluating changes in retinal thickness, during patient follow-up. 
 
Figure 1.
 
(a) Representation of an OCT macular thickness map of a right eye, which is derived from six linear B-scans through the center of fixation. The map represents a circle of 6-mm diameter that is automatically subdivided into nine ETDRS areas by the system’s analysis software: A1 average foveal thickness, A2 parafoveal superior thickness, A3 parafoveal temporal thickness, A4 parafoveal inferior thickness, A5 parafoveal nasal thickness, A6 perifoveal superior thickness, A7 perifoveal temporal thickness, A8 perifoveal inferior thickness, and A9 perifoveal nasal thickness. Note that in left eyes, the labels A3, A5, A7, and A9 are mirrored. (b) Example of an RTA macular thickness map of a right eye in a 6 × 6-mm square. The map is automatically subdivided in a 7 × 7 grid of retinal thickness values by the system’s analysis software. For comparison with the OCT, this square was subdivided and labeled into nine ETDRS-like areas (A1–A9) in a similar fashion.
Figure 1.
 
(a) Representation of an OCT macular thickness map of a right eye, which is derived from six linear B-scans through the center of fixation. The map represents a circle of 6-mm diameter that is automatically subdivided into nine ETDRS areas by the system’s analysis software: A1 average foveal thickness, A2 parafoveal superior thickness, A3 parafoveal temporal thickness, A4 parafoveal inferior thickness, A5 parafoveal nasal thickness, A6 perifoveal superior thickness, A7 perifoveal temporal thickness, A8 perifoveal inferior thickness, and A9 perifoveal nasal thickness. Note that in left eyes, the labels A3, A5, A7, and A9 are mirrored. (b) Example of an RTA macular thickness map of a right eye in a 6 × 6-mm square. The map is automatically subdivided in a 7 × 7 grid of retinal thickness values by the system’s analysis software. For comparison with the OCT, this square was subdivided and labeled into nine ETDRS-like areas (A1–A9) in a similar fashion.
Table 1.
 
Descriptives
Table 1.
 
Descriptives
Area Optical Coherence Tomography Kruskal Wallis P
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 57 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Mean SD Mean SD Mean SD
Average foveal thickness A1 193 17 205 26 322 115 <0.001
 Parafoveal
  Superior A2 271 10 269 24 349 106 <0.001
  Temporal A3 260 10 257 26 347 118 <0.001
  Inferior A4 270 10 266 23 325 85 0.01
  Nasal A5 270 9 270 22 334 77 <0.001
  Average P1 268 11 265 24 339 97 <0.001
 Perifoveal
  Superior A6 237 12 234 18 296 95 0.01
  Temporal A7 217 13 217 25 298 102 <0.001
  Inferior A8 226 14 224 16 268 76 0.01
  Nasal A9 250 12 247 16 290 55 0.01
  Average P2 233 18 230 22 288 83 <0.001
Area Retinal Thickness Analyzer Kruskal Wallis P
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 56 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Mean SD Mean SD Mean SD
Average foveal thickness A1 134 14 144 24 210 65 <0.001
 Parafoveal
  Superior A2 195 13 186 22 178 13 0.01
  Temporal A3 187 10 179 22 194 33 0.11
  Inferior A4 196 9 186 21 184 24 <0.01
  Nasal A5 192 11 188 21 191 31 0.24
  Average P1 192 11 185 22 187 27 <0.001
 Perifoveal
  Superior A6 173 12 171 20 168 21 0.70
  Temporal A7 165 12 168 21 181 30 0.03
  Inferior A8 165 12 166 18 163 17 0.88
  Nasal A9 174 14 174 20 170 22 0.92
  Average P2 169 13 170 20 171 24 0.93
Table 2.
 
Measurement Precision
Table 2.
 
Measurement Precision
Area Eyes of Healthy Volunteers Eyes of Diabetic Patients without Macular Edema Eyes of Diabetic Patients with Macular Edema P Kruskal Wallis Test
E = 28 (n = 14) OCT P E = 28 (n = 14) RTA E = 57 (n = 31) OCT P E = 56 (n = 31) RTA E = 21 (n = 15) OCT P E = 21 (n = 15) RTA OCT RTA
Average foveal thickness A1 8 <0.001 33 12 <0.001 43 36 0.07 52 <0.001 0.14
 Parafoveal
  Superior thickness A2 5 <0.001 33 7 <0.001 40 30 0.20 44 <0.001 0.75
  Temporal thickness A3 5 <0.001 35 12 <0.001 39 21 <0.01 49 <0.001 0.27
  Inferior thickness A4 5 <0.001 30 7 <0.001 45 23 <0.01 51 <0.001 0.11
  Nasal thickness A5 5 <0.001 26 11 <0.001 43 39 <0.01 51 <0.001 0.28
  Average P1 5 <0.001 30 10 <0.001 41 29 <0.001 49 <0.001 0.16
 Perifoveal
  Superior thickness A6 6 <0.001 34 7 <0.001 40 23 0.03 34 <0.01 0.96
  Temporal thickness A7 7 <0.001 31 7 <0.001 44 26 <0.01 61 <0.001 0.17
  Inferior thickness A8 7 <0.001 31 8 <0.001 41 20 <0.01 42 <0.01 0.10
  Nasal thickness A9 7 <0.001 34 13 <0.001 41 20 <0.01 44 <0.01 0.59
  Average P2 7 <0.001 33 9 <0.001 40 25 <0.001 46 <0.001 0.74
Figure 2.
 
Graphic representation of a test for heteroscedasticity in eyes of diabetic patients with macular edema. A linear fit of the within-patient SD (σwithin-patient) per area with the mean retinal thickness per area is shown for all parafoveal (a, b, top) and perifoveal (c, d, bottom) areas as measured with both the OCT (a, c, left) and RTA (b, d, right). A linear equation (y = a · x + b) of each fit is shown with its slope (a) and y-intercept (b). A Pearson correlation coefficient test was performed to determine the amount (r 2) and significance (P) of a possible correlation. Significant correlation was present in parafoveal areas when measured with the OCT (P < 0.001; a) with marked presence of heteroscedasticity (r 2 = 0.36). In contrast, no heteroscedasticity was found in parafoveal areas when measurements were performed with the RTA. However, there is a much larger spread in measurement variance (σwithin-patient) in the RTA (compare a with b). A small but significant correlation was also found for perifoveal areas in both instruments.
Figure 2.
 
Graphic representation of a test for heteroscedasticity in eyes of diabetic patients with macular edema. A linear fit of the within-patient SD (σwithin-patient) per area with the mean retinal thickness per area is shown for all parafoveal (a, b, top) and perifoveal (c, d, bottom) areas as measured with both the OCT (a, c, left) and RTA (b, d, right). A linear equation (y = a · x + b) of each fit is shown with its slope (a) and y-intercept (b). A Pearson correlation coefficient test was performed to determine the amount (r 2) and significance (P) of a possible correlation. Significant correlation was present in parafoveal areas when measured with the OCT (P < 0.001; a) with marked presence of heteroscedasticity (r 2 = 0.36). In contrast, no heteroscedasticity was found in parafoveal areas when measurements were performed with the RTA. However, there is a much larger spread in measurement variance (σwithin-patient) in the RTA (compare a with b). A small but significant correlation was also found for perifoveal areas in both instruments.
Table 3.
 
Intraclass Correlation Coefficients
Table 3.
 
Intraclass Correlation Coefficients
Area Optical Coherence Tomography
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 57 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Average foveal thickness A1 0.97 0.97 0.96
 Parafoveal
  Superior A2 0.96 0.99 0.99
  Temporal A3 0.97 0.97 0.996
  Inferior A4 0.96 0.99 0.99
  Nasal A5 0.96 0.97 0.97
  Average P1 0.96 0.98 0.99
 Perifoveal
  Superior A6 0.97 0.98 0.98
  Temporal A7 0.97 0.99 0.99
  Inferior A8 0.97 0.97 0.99
  Nasal A9 0.95 0.92 0.98
  Average P2 0.97 0.97 0.99
Area Retinal Thickness Analyzer
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 56 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Average foveal thickness A1 0.50 0.69 0.89
 Parafoveal
  Superior A2 0.49 0.66 0.24
  Temporal A3 0.26 0.67 0.78
  Inferior A4 0.26 0.58 0.62
  Nasal A5 0.53 0.60 0.71
  Average P1 0.44 0.65 0.67
 Perifoveal
  Superior A6 0.38 0.62 0.65
  Temporal A7 0.47 0.60 0.60
  Inferior A8 0.49 0.54 0.48
  Nasal A9 0.51 0.60 0.62
  Average P2 0.45 0.62 0.59
Figure 3.
 
A comparison between the OCT and RTA scans of two different eyes, illustrating the seeming inability of the RTA to detect accurately the extrafoveal retinal thickening due to edema. (a, left) A comparison between the OCT and RTA scan of the right eye of a subject with temporal macular edema. The increase in retinal thickness of the parafoveal and perifoveal temporal thickness (areas A3 and A7, see Fig. 1 ) was reproducible on all three OCT scans. In contrast, the RTA scans failed to measure the central part of the edematous area (a, marked manually with an X) on all three occasions and was able to measure the borders only to a limited extent. (b, left) The right eye of a subject with central macular edema. In this case, the RTA only showed an increased retinal thickness value for the average foveal thickness (area A1; see Fig. 1 ), whereas the OCT also shows increased retinal thicknesses for all parafoveal areas (A2–A5).
Figure 3.
 
A comparison between the OCT and RTA scans of two different eyes, illustrating the seeming inability of the RTA to detect accurately the extrafoveal retinal thickening due to edema. (a, left) A comparison between the OCT and RTA scan of the right eye of a subject with temporal macular edema. The increase in retinal thickness of the parafoveal and perifoveal temporal thickness (areas A3 and A7, see Fig. 1 ) was reproducible on all three OCT scans. In contrast, the RTA scans failed to measure the central part of the edematous area (a, marked manually with an X) on all three occasions and was able to measure the borders only to a limited extent. (b, left) The right eye of a subject with central macular edema. In this case, the RTA only showed an increased retinal thickness value for the average foveal thickness (area A1; see Fig. 1 ), whereas the OCT also shows increased retinal thicknesses for all parafoveal areas (A2–A5).
Table 4.
 
Comparison of LA95% and ICCs in the Present and Previous Studies on Measurement Precision
Table 4.
 
Comparison of LA95% and ICCs in the Present and Previous Studies on Measurement Precision
Parameter Present Study OCT 3 Stratus Polito et al. 12 OCT 3 Stratus Paunescu et al. 11 OCT 3 Stratus Massin et al. 10 OCT Previous Version
Eyes of Healthy Volunteers
E = 28 (n = 14) E = 10 (n = 10) E = 10 (n = 10) E = 10 (n = 10)
LA95% ICC LA95% ICC LA95% ICC LA95% ICC
Average foveal thickness A1 8 0.97 13 NA 22 0.77 6 0.97
 Parafoveal thickness
  Superior A2 5 0.96 9 NA 18 0.61 5 0.99
  Temporal A3 5 0.97 13 NA 14 0.81 5 0.98
  Inferior A4 5 0.96 16 NA 15 0.60 4 0.94
  Nasal A5 5 0.96 16 NA 14 0.55 5 0.97
  Average P1 5 0.96 13 NA 15 0.64 5 0.97
 Perifoveal thickness
  Superior A6 6 0.97 13 NA 11 0.57 5 0.94
  Temporal A7 7 0.97 24 NA 10 0.97 4 0.97
  Inferior A8 7 0.97 16 NA 9 0.81 4 0.97
  Nasal A9 7 0.95 6 NA 17 0.69 6 0.98
  Average P2 7 0.97 15 NA 11 0.76 5 0.97
Eyes of Diabetic Patients with Macular Edema
E = 21 (n = 15) E = 15 (n = 15) None None
LA95% ICC LA95% ICC LA95% ICC LA95% ICC
Average foveal thickness A1 66 0.99 32 NA NA NA NA NA
 Parafoveal thickness
  Superior A2 30 0.99 30 NA NA NA NA NA
  Temporal A3 21 0.996 29 NA NA NA NA NA
  Inferior A4 23 0.99 19 NA NA NA NA NA
  Nasal A5 39 0.97 25 NA NA NA NA NA
  Average P1 29 0.99 26 NA NA NA NA NA
 Perifoveal thickness
  Superior A6 38 0.98 28 NA NA NA NA NA
  Temporal A7 26 0.99 24 NA NA NA NA NA
  Inferior A8 20 0.99 28 NA NA NA NA NA
  Nasal A9 20 0.98 28 NA NA NA NA NA
  Average P2 27 0.99 27 NA NA NA NA NA
MohamedQ, GilliesMC, WongTY. Management of diabetic retinopathy: a systematic review. JAMA. 2007;298:902–916. [CrossRef] [PubMed]
Treatment techniques and clinical guidelines for photocoagulation of diabetic macular edema. Early Treatment Diabetic Retinopathy Study Report Number 2. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology. 1987;94:761–774. [CrossRef] [PubMed]
HeeMR, PuliafitoCA, WongC, et al. Quantitative assessment of macular edema with optical coherence tomography. Arch Ophthalmol. 1995;113:1019–1029. [CrossRef] [PubMed]
ZeimerR, ShahidiM, MoriM, ZouS, AsraniS. A new method for rapid mapping of the retinal thickness at the posterior pole. Invest Ophthalmol Vis Sci. 1996;37:1994–2001. [PubMed]
WeinbergerD, Axer-SiegelR, LandauD, YassurY. Retinal thickness variation in the diabetic patient measured by the retinal thickness analyser. Br J Ophthalmol. 1998;82:1003–1006. [CrossRef] [PubMed]
FritscheP, van der HeijdeR, Suttorp-SchultenMS, PolakBC. Retinal thickness analysis (RTA): an objective method to assess and quantify the retinal thickness in healthy controls and in diabetics without diabetic retinopathy. Retina. 2002;22:768–771. [CrossRef] [PubMed]
PolitoA, ShahSM, HallerJA, et al. Comparison between retinal thickness analyzer and optical coherence tomography for assessment of foveal thickness in eyes with macular disease. Am J Ophthalmol. 2002;134:240–251. [CrossRef] [PubMed]
KonnoS, AkibaJ, YoshidaA. Retinal thickness measurements with optical coherence tomography and the scanning retinal thickness analyzer. Retina. 2001;21:57–61. [CrossRef] [PubMed]
OshimaY, EmiK, YamanishiS, MotokuraM. Quantitative assessment of macular thickness in normal subjects and patients with diabetic retinopathy by scanning retinal thickness analyser. Br J Ophthalmol. 1999;83:54–61. [CrossRef] [PubMed]
MassinP, VicautE, HaouchineB, ErginayA, PaquesM, GaudricA. Reproducibility of retinal mapping using optical coherence tomography. Arch Ophthalmol. 2001;119:1135–1142. [CrossRef] [PubMed]
PaunescuLA, SchumanJS, PriceLL, et al. Reproducibility of nerve fiber thickness, macular thickness, and optic nerve head measurements using StratusOCT. Invest Ophthalmol Vis Sci. 2004;45:1716–1724. [CrossRef] [PubMed]
PolitoA, Del BorrelloM, IsolaM, ZemellaN, BandelloF. Repeatability and reproducibility of fast macular thickness mapping with stratus optical coherence tomography. Arch Ophthalmol. 2005;123:1330–1337. [CrossRef] [PubMed]
BlandJM, AltmanDG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet. 1995;346:1085–1087. [CrossRef] [PubMed]
TangelderGJ, ReusNJ, LemijHG. Estimating the clinical usefulness of optic disc biometry for detecting glaucomatous change over time. Eye. 2006;20:755–763. [CrossRef] [PubMed]
ZouH, ZhangX, XuX, YuS. Quantitative in vivo retinal thickness measurement in chinese healthy subjects with retinal thickness analyzer. Invest Ophthalmol Vis Sci. 2006;47:341–347. [CrossRef] [PubMed]
Figure 1.
 
(a) Representation of an OCT macular thickness map of a right eye, which is derived from six linear B-scans through the center of fixation. The map represents a circle of 6-mm diameter that is automatically subdivided into nine ETDRS areas by the system’s analysis software: A1 average foveal thickness, A2 parafoveal superior thickness, A3 parafoveal temporal thickness, A4 parafoveal inferior thickness, A5 parafoveal nasal thickness, A6 perifoveal superior thickness, A7 perifoveal temporal thickness, A8 perifoveal inferior thickness, and A9 perifoveal nasal thickness. Note that in left eyes, the labels A3, A5, A7, and A9 are mirrored. (b) Example of an RTA macular thickness map of a right eye in a 6 × 6-mm square. The map is automatically subdivided in a 7 × 7 grid of retinal thickness values by the system’s analysis software. For comparison with the OCT, this square was subdivided and labeled into nine ETDRS-like areas (A1–A9) in a similar fashion.
Figure 1.
 
(a) Representation of an OCT macular thickness map of a right eye, which is derived from six linear B-scans through the center of fixation. The map represents a circle of 6-mm diameter that is automatically subdivided into nine ETDRS areas by the system’s analysis software: A1 average foveal thickness, A2 parafoveal superior thickness, A3 parafoveal temporal thickness, A4 parafoveal inferior thickness, A5 parafoveal nasal thickness, A6 perifoveal superior thickness, A7 perifoveal temporal thickness, A8 perifoveal inferior thickness, and A9 perifoveal nasal thickness. Note that in left eyes, the labels A3, A5, A7, and A9 are mirrored. (b) Example of an RTA macular thickness map of a right eye in a 6 × 6-mm square. The map is automatically subdivided in a 7 × 7 grid of retinal thickness values by the system’s analysis software. For comparison with the OCT, this square was subdivided and labeled into nine ETDRS-like areas (A1–A9) in a similar fashion.
Figure 2.
 
Graphic representation of a test for heteroscedasticity in eyes of diabetic patients with macular edema. A linear fit of the within-patient SD (σwithin-patient) per area with the mean retinal thickness per area is shown for all parafoveal (a, b, top) and perifoveal (c, d, bottom) areas as measured with both the OCT (a, c, left) and RTA (b, d, right). A linear equation (y = a · x + b) of each fit is shown with its slope (a) and y-intercept (b). A Pearson correlation coefficient test was performed to determine the amount (r 2) and significance (P) of a possible correlation. Significant correlation was present in parafoveal areas when measured with the OCT (P < 0.001; a) with marked presence of heteroscedasticity (r 2 = 0.36). In contrast, no heteroscedasticity was found in parafoveal areas when measurements were performed with the RTA. However, there is a much larger spread in measurement variance (σwithin-patient) in the RTA (compare a with b). A small but significant correlation was also found for perifoveal areas in both instruments.
Figure 2.
 
Graphic representation of a test for heteroscedasticity in eyes of diabetic patients with macular edema. A linear fit of the within-patient SD (σwithin-patient) per area with the mean retinal thickness per area is shown for all parafoveal (a, b, top) and perifoveal (c, d, bottom) areas as measured with both the OCT (a, c, left) and RTA (b, d, right). A linear equation (y = a · x + b) of each fit is shown with its slope (a) and y-intercept (b). A Pearson correlation coefficient test was performed to determine the amount (r 2) and significance (P) of a possible correlation. Significant correlation was present in parafoveal areas when measured with the OCT (P < 0.001; a) with marked presence of heteroscedasticity (r 2 = 0.36). In contrast, no heteroscedasticity was found in parafoveal areas when measurements were performed with the RTA. However, there is a much larger spread in measurement variance (σwithin-patient) in the RTA (compare a with b). A small but significant correlation was also found for perifoveal areas in both instruments.
Figure 3.
 
A comparison between the OCT and RTA scans of two different eyes, illustrating the seeming inability of the RTA to detect accurately the extrafoveal retinal thickening due to edema. (a, left) A comparison between the OCT and RTA scan of the right eye of a subject with temporal macular edema. The increase in retinal thickness of the parafoveal and perifoveal temporal thickness (areas A3 and A7, see Fig. 1 ) was reproducible on all three OCT scans. In contrast, the RTA scans failed to measure the central part of the edematous area (a, marked manually with an X) on all three occasions and was able to measure the borders only to a limited extent. (b, left) The right eye of a subject with central macular edema. In this case, the RTA only showed an increased retinal thickness value for the average foveal thickness (area A1; see Fig. 1 ), whereas the OCT also shows increased retinal thicknesses for all parafoveal areas (A2–A5).
Figure 3.
 
A comparison between the OCT and RTA scans of two different eyes, illustrating the seeming inability of the RTA to detect accurately the extrafoveal retinal thickening due to edema. (a, left) A comparison between the OCT and RTA scan of the right eye of a subject with temporal macular edema. The increase in retinal thickness of the parafoveal and perifoveal temporal thickness (areas A3 and A7, see Fig. 1 ) was reproducible on all three OCT scans. In contrast, the RTA scans failed to measure the central part of the edematous area (a, marked manually with an X) on all three occasions and was able to measure the borders only to a limited extent. (b, left) The right eye of a subject with central macular edema. In this case, the RTA only showed an increased retinal thickness value for the average foveal thickness (area A1; see Fig. 1 ), whereas the OCT also shows increased retinal thicknesses for all parafoveal areas (A2–A5).
Table 1.
 
Descriptives
Table 1.
 
Descriptives
Area Optical Coherence Tomography Kruskal Wallis P
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 57 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Mean SD Mean SD Mean SD
Average foveal thickness A1 193 17 205 26 322 115 <0.001
 Parafoveal
  Superior A2 271 10 269 24 349 106 <0.001
  Temporal A3 260 10 257 26 347 118 <0.001
  Inferior A4 270 10 266 23 325 85 0.01
  Nasal A5 270 9 270 22 334 77 <0.001
  Average P1 268 11 265 24 339 97 <0.001
 Perifoveal
  Superior A6 237 12 234 18 296 95 0.01
  Temporal A7 217 13 217 25 298 102 <0.001
  Inferior A8 226 14 224 16 268 76 0.01
  Nasal A9 250 12 247 16 290 55 0.01
  Average P2 233 18 230 22 288 83 <0.001
Area Retinal Thickness Analyzer Kruskal Wallis P
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 56 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Mean SD Mean SD Mean SD
Average foveal thickness A1 134 14 144 24 210 65 <0.001
 Parafoveal
  Superior A2 195 13 186 22 178 13 0.01
  Temporal A3 187 10 179 22 194 33 0.11
  Inferior A4 196 9 186 21 184 24 <0.01
  Nasal A5 192 11 188 21 191 31 0.24
  Average P1 192 11 185 22 187 27 <0.001
 Perifoveal
  Superior A6 173 12 171 20 168 21 0.70
  Temporal A7 165 12 168 21 181 30 0.03
  Inferior A8 165 12 166 18 163 17 0.88
  Nasal A9 174 14 174 20 170 22 0.92
  Average P2 169 13 170 20 171 24 0.93
Table 2.
 
Measurement Precision
Table 2.
 
Measurement Precision
Area Eyes of Healthy Volunteers Eyes of Diabetic Patients without Macular Edema Eyes of Diabetic Patients with Macular Edema P Kruskal Wallis Test
E = 28 (n = 14) OCT P E = 28 (n = 14) RTA E = 57 (n = 31) OCT P E = 56 (n = 31) RTA E = 21 (n = 15) OCT P E = 21 (n = 15) RTA OCT RTA
Average foveal thickness A1 8 <0.001 33 12 <0.001 43 36 0.07 52 <0.001 0.14
 Parafoveal
  Superior thickness A2 5 <0.001 33 7 <0.001 40 30 0.20 44 <0.001 0.75
  Temporal thickness A3 5 <0.001 35 12 <0.001 39 21 <0.01 49 <0.001 0.27
  Inferior thickness A4 5 <0.001 30 7 <0.001 45 23 <0.01 51 <0.001 0.11
  Nasal thickness A5 5 <0.001 26 11 <0.001 43 39 <0.01 51 <0.001 0.28
  Average P1 5 <0.001 30 10 <0.001 41 29 <0.001 49 <0.001 0.16
 Perifoveal
  Superior thickness A6 6 <0.001 34 7 <0.001 40 23 0.03 34 <0.01 0.96
  Temporal thickness A7 7 <0.001 31 7 <0.001 44 26 <0.01 61 <0.001 0.17
  Inferior thickness A8 7 <0.001 31 8 <0.001 41 20 <0.01 42 <0.01 0.10
  Nasal thickness A9 7 <0.001 34 13 <0.001 41 20 <0.01 44 <0.01 0.59
  Average P2 7 <0.001 33 9 <0.001 40 25 <0.001 46 <0.001 0.74
Table 3.
 
Intraclass Correlation Coefficients
Table 3.
 
Intraclass Correlation Coefficients
Area Optical Coherence Tomography
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 57 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Average foveal thickness A1 0.97 0.97 0.96
 Parafoveal
  Superior A2 0.96 0.99 0.99
  Temporal A3 0.97 0.97 0.996
  Inferior A4 0.96 0.99 0.99
  Nasal A5 0.96 0.97 0.97
  Average P1 0.96 0.98 0.99
 Perifoveal
  Superior A6 0.97 0.98 0.98
  Temporal A7 0.97 0.99 0.99
  Inferior A8 0.97 0.97 0.99
  Nasal A9 0.95 0.92 0.98
  Average P2 0.97 0.97 0.99
Area Retinal Thickness Analyzer
Eyes of Healthy Volunteers E = 28 (n = 14) Eyes of Diabetic Patients without Macular Edema E = 56 (n = 31) Eyes of Diabetic Patients with Macular Edema E = 21 (n = 15)
Average foveal thickness A1 0.50 0.69 0.89
 Parafoveal
  Superior A2 0.49 0.66 0.24
  Temporal A3 0.26 0.67 0.78
  Inferior A4 0.26 0.58 0.62
  Nasal A5 0.53 0.60 0.71
  Average P1 0.44 0.65 0.67
 Perifoveal
  Superior A6 0.38 0.62 0.65
  Temporal A7 0.47 0.60 0.60
  Inferior A8 0.49 0.54 0.48
  Nasal A9 0.51 0.60 0.62
  Average P2 0.45 0.62 0.59
Table 4.
 
Comparison of LA95% and ICCs in the Present and Previous Studies on Measurement Precision
Table 4.
 
Comparison of LA95% and ICCs in the Present and Previous Studies on Measurement Precision
Parameter Present Study OCT 3 Stratus Polito et al. 12 OCT 3 Stratus Paunescu et al. 11 OCT 3 Stratus Massin et al. 10 OCT Previous Version
Eyes of Healthy Volunteers
E = 28 (n = 14) E = 10 (n = 10) E = 10 (n = 10) E = 10 (n = 10)
LA95% ICC LA95% ICC LA95% ICC LA95% ICC
Average foveal thickness A1 8 0.97 13 NA 22 0.77 6 0.97
 Parafoveal thickness
  Superior A2 5 0.96 9 NA 18 0.61 5 0.99
  Temporal A3 5 0.97 13 NA 14 0.81 5 0.98
  Inferior A4 5 0.96 16 NA 15 0.60 4 0.94
  Nasal A5 5 0.96 16 NA 14 0.55 5 0.97
  Average P1 5 0.96 13 NA 15 0.64 5 0.97
 Perifoveal thickness
  Superior A6 6 0.97 13 NA 11 0.57 5 0.94
  Temporal A7 7 0.97 24 NA 10 0.97 4 0.97
  Inferior A8 7 0.97 16 NA 9 0.81 4 0.97
  Nasal A9 7 0.95 6 NA 17 0.69 6 0.98
  Average P2 7 0.97 15 NA 11 0.76 5 0.97
Eyes of Diabetic Patients with Macular Edema
E = 21 (n = 15) E = 15 (n = 15) None None
LA95% ICC LA95% ICC LA95% ICC LA95% ICC
Average foveal thickness A1 66 0.99 32 NA NA NA NA NA
 Parafoveal thickness
  Superior A2 30 0.99 30 NA NA NA NA NA
  Temporal A3 21 0.996 29 NA NA NA NA NA
  Inferior A4 23 0.99 19 NA NA NA NA NA
  Nasal A5 39 0.97 25 NA NA NA NA NA
  Average P1 29 0.99 26 NA NA NA NA NA
 Perifoveal thickness
  Superior A6 38 0.98 28 NA NA NA NA NA
  Temporal A7 26 0.99 24 NA NA NA NA NA
  Inferior A8 20 0.99 28 NA NA NA NA NA
  Nasal A9 20 0.98 28 NA NA NA NA NA
  Average P2 27 0.99 27 NA NA NA NA NA
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×