The Quality of Reporting of Diagnostic Accuracy Studies in Glaucoma Using the Heidelberg Retina Tomograph

Abstract

purpose. Scanning laser tomography with the Heidelberg retina tomograph (HRT; Heidelberg Engineering, Heidelberg, Germany) has been proposed as a useful diagnostic test for glaucoma. This study was conducted to evaluate the quality of reporting of published studies using the HRT for diagnosing glaucoma.

methods. A validated Medline and hand search of English-language articles reporting on measures of diagnostic accuracy of the HRT for glaucoma was performed. Two reviewers selected and appraised the papers independently. The Standards for Reporting of Diagnostic Accuracy (STARD) checklist was used to evaluate the quality of each publication.

results. A total of 29 articles were included. Interobserver rating agreement was observed in 83% of items (κ = 0.76). The number of STARD items properly reported ranged from 5 to 18. Less than a third of studies (7/29) explicitly reported more than half of the STARD items. Descriptions of key aspects of the methodology were frequently missing. For example, the design of the study (prospective or retrospective) was reported in 6 of 29 studies, and details of participant sampling (e.g., consecutive or random selection) were described in 5 of 29 publications. The commonest description of diagnostic accuracy was sensitivity and specificity (25/29) followed by area under the ROC curve (13/29), with 9 of 29 publications reporting both.

conclusions. The quality of reporting of diagnostic accuracy tests for glaucoma with HRT is suboptimal. The STARD initiative may be a useful tool for appraising the strengths and weaknesses of diagnostic accuracy studies.

In the field of glaucoma practice, there has recently been a growth in the number of diagnostic tests designed to detect structural and functional damage at relatively early stages of the disease. Image-based diagnostic technologies introduced in the past decade include scanning laser polarimetry, optical coherence tomography, and scanning laser tomography (e.g., the Heidelberg retina tomograph [HRT]; Heidelberg Engineering, Heidelberg, Germany). The HRT provides objective, quantitative measures of the optic disc topography and shows promise for discriminating between glaucomatous and normal eyes. ¹ The topographic image is derived from multiple optical sections at consecutive focal depth planes. Each image consists of numerous pixels, with each pixel corresponding to the retinal height at its location. ¹ The first model was marketed in 1991 and a second version (HRT-II) was made available in 1999. 

Diagnostic accuracy studies are required to validate new diagnostic tests before they are introduced into clinical practice. These studies typically report sensitivity and specificity, likelihood ratios, diagnostic odds ratio, or area under a receiver operating characteristics (ROC) curve as measures of diagnostic performance. This information enables a clinician to make judgments regarding the potential utility of new tests. However, improperly conducted and incompletely reported studies are prone to bias that, in turn, may lead to overly optimistic estimations of the diagnostic value of the test. ² Exaggerated results may lead to premature adoption of diagnostic tests and to incorrect clinical decisions. 

To improve the quality of reporting of diagnostic accuracy studies, the Standards for Reporting of Diagnostic Accuracy (STARD) initiative was recently established. ^{3 4} During a consensus conference in the year 2000, the STARD project group developed a checklist of 25 items and a prototypical flowchart. ^{5 6} Guidelines for reporting of other study types have been widely accepted—for example, CONSORT (Consolidated Standards of Reporting Trials), for randomized clinical trials, ⁷ and QUOROM (Quality of Reporting of Meta-Analyses) and MOOSE (Meta-analysis of Observed Studies in Epidemiology), for systematic reviews. ^{7 8 9 10}  

The purpose of this study was to examine the current standard of reporting of diagnostic accuracy publications using the HRT for the detection of glaucoma. 

Methods

One reviewer (MS) searched MEDLINE (National Center for Biotechnology Inforamton, Bethesda, MD; .ncbi.nlm.nih.gov) with a validated strategy ¹¹ to identify articles on diagnostic accuracy of glaucoma published between January 1966 and January 2005. A PubMed (NCBI) search using medical subject headings and keywords was executed using the following terms: “specificity” or “false negative” or “accuracy” or “diagnostic accuracy” or “sensitivity and specificity,” and “Heidelberg Retina Tomograph” or “confocal scanning laser tomograph” or “optic disc” or “glaucoma. ” The search was subsequently limited to publications in English and studies focusing on human subjects. 

Because the yield of search strategies for diagnostic accuracy tests is suboptimal, ¹¹ a hand search of all papers included in the reference list of the short-listed manuscripts was also performed. The articles were included if they reported on measures of diagnostic accuracy of the HRT for glaucoma. We independently assessed the title, abstract, and key words of the 40 identified articles to determine whether they met the inclusion criteria. If there was any doubt, the full text of the article was retrieved and read by both reviewers. Any disparities were resolved by consensus. The process of selection is summarized in Figure 1 . 

The STARD checklist (Table 1)was used to assess the quality of reporting. The current checklist items are arranged under the following headings: (1) title, abstract, and keywords, (2) introduction, (3) methods (11 items), (4) results (11 items), and (5) discussion. Each item was considered to be fully, partially, or not reported (Table 2) . If the item was “not applicable,” it was marked as such. For example, item 21 required reporting of estimates of diagnostic accuracy and measures of statistical uncertainty. If a study reported estimates of accuracy but no measure of uncertainty, it was considered partially fulfilled (Table 2) . Similarly, item 20 (reporting of adverse events associated with the test) was scored as “not applicable” due to the noninvasive nature of the HRT test. We independently evaluated the quality of reporting of each included study. Disagreements were resolved by consensus. Interobserver agreement (κ statistic) and descriptive statistics were calculated on computer (SPSS for Windows, ver. 13.0; SPSS, Chicago, IL). A κ statistic ¹² of 0 is considered poor agreement; 0.01 to 0.2, slight agreement; 0.21 to 0.4, fair agreement; 0.41 to 0.6, moderate agreement; 0.61 to 0.8, substantial agreement; and 0.81 to 1.00, almost perfect agreement. Quantitative comparison of the quality of reporting among the 29 manuscripts was not made because STARD items are qualitative and of variable importance, and so an objective score would not be possible. ³  

Results

Figure 1illustrates the selection process of the manuscripts. One thousand sixty-four publications were initially excluded based on their titles or abstracts, as they were clearly not reporting any evaluation of the diagnostic accuracy of HRT in glaucoma. The search identified 40 manuscripts as potentially suitable for inclusion in the study (Fig. 1)on the basis of the title, abstract, and/or keywords. Subsequently, 11 (Refs. ^{13 14 15 16 17 18 19 20 21 22 23} ) articles were excluded. Reasons for exclusions were (1) not assessing diagnostic accuracy ^{13 14 15 16 17 18 19} (n = 7), (2) describing new reference planes ^{20 21 22} (n = 3), and (3) not using HRT ²³ (n = 1). Twenty-nine (Refs. ^{1 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51} ) manuscripts fulfilled the selection criteria and were included in the study. 

Regarding interobserver rating agreement, the κ statistic was substantial (κ = 0.76). Most disagreement occurred when particular aspects of the manuscript were poorly reported leading to difficulty in scoring certain STARD items (e.g., as partially or not fulfilled). Among the 29 articles, the number of STARD items reported in an article ranged from 5 to 18. Less than a third of the studies (7/29) explicitly reported more than half of the STARD items. 

Reporting of an individual STARD item ranged from 1 of 29 (item 16) to 29 of 29 (items 2 and 25; Table 2 ). The commonest description of diagnostic accuracy was sensitivity and specificity (25/29) followed by area under the ROC curve (13/29), with 9 of 29 publications reporting both. The reporting of each of the items is described in Table 2 . Table 3summarizes the number of STARD items fully, partially, or not fulfilled for each manuscript evaluated after consensus. 

Discussion

With glaucoma being one of the leading causes of blindness in the world ^{52 53} an accurate test to detect the disease would greatly benefit clinicians and patients. ^{54 55} Current glaucoma diagnosis relies on the qualitative evaluation of the optic disc and is dependent on the experience of the observer. ^{56 57} The HRT has the theoretical advantage of providing an objective, user friendly, and quantitative assessment and thus has a promising role in the diagnosis and perhaps screening and monitoring of glaucoma. ^{1 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51}  

Because the introduction of the HRT during the past decade, several publications have reported its diagnostic performance. ^{1 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51} A critical appraisal of the literature may be useful for current and potential users of HRT for diagnosing glaucoma. The use of STARD to evaluate the quality of reporting in diagnostic accuracy studies has been described recently, ^{58 59 60} and a similar approach was used in this study. 

Regarding the specific aspects of the studies appraised, all articles described the study population (item 3) to some extent. Eighteen of 29 publications fully reported this point, whereas 11 of 29 reported at least one of the four descriptors included in this item: setting and location of the study and inclusion or exclusion criteria. A majority of the publications (22/29) did not describe participant sampling (item 5). Description of participant sampling (e.g., whether patients were prospectively and/or consecutively enrolled) is helpful for readers to judge whether the study was subject to recruitment bias and whether the results would be generalizable. In 22 of 29 publications, there was no indication of the prospective or retrospective nature of the study (item 6). 

Only one publication did not describe the reference standard used (item 7) properly. The influence of variable definitions of the reference standard in studies evaluating the diagnostic accuracy of the HRT has been described recently by Miglior et al., ⁶¹ who highlighted the need for a standard definition of glaucoma in diagnostic accuracy studies. 

As the HRT has undergone several software updates, ^{24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51} it would be prudent to mention when measurements were obtained and the particular version of the HRT used in the study. According to Garway-Heath et al. ⁶² HRT variability differs between software versions and reference planes. Seven of 29 publications clearly described the technical specifications and software involved, including how and when measurements were obtained (item 8). Most publications did not report when the HRT measurements were obtained (22/29), or when the study took place (26/29; item 14). 

Approximately one third of the publications (9/29) fully described the number, training, and expertise of persons involved in the executing and reading of the tests (item 10). Although the reported interobserver variability is low for the HRT, ^{62 63 64 65} knowledge of the numbers, training, and experience of HRT operators would help to estimate the repeatability of the test results in different settings. 

Most (23/29) of the publications failed to report whether the test results were interpreted without knowledge of the reference standard (item 11). Review bias may lead to inflated results of diagnostic accuracy. Description of the demographics of the study population is also essential for readers to judge the possibility of selection bias and the applicability of the study findings to their own clinical practice (item 15). Ten of 29 publications did not report relevant demographic characteristics. 

Most publications (27/29) failed to report the number of participants who had satisfied the criteria for inclusion and subsequently did or did not undergo the index test (item 16). None of the reports included a flow diagram, though one report included a Venn diagram. Regarding the time interval between the index test (HRT) and the reference standard (item 17), in 4 of 29 studies both tests were performed on the same day; 15 of 29 did not report any time intervals, and 10 of 29 reported a range from 2 to 12 months between tests. None of the studies described whether the patients had stable glaucoma and whether patients received any treatment during the time between the index test and the reference standard. It has been shown that both medical and surgical intervention can influence the topography of the optic disc, ^{66 67 68 69} making it a potentially confounding factor to consider. 

Only 5 of 29 publications reported statistical uncertainty (item 21). The reported values of sensitivity, specificity, and ROC of HRT are only estimates. Measures of uncertainty (such as CI) give readers a range within which true values may lie and would help to estimate the precision of the diagnostic test. ⁷⁰ Some patients may not be able to undergo the index test and occasionally results may be uninterpretable (item 22). This information was reported fully in 8 of 29 publications. However, most studies (21/29) did not address the question of indeterminate or uninterpretable results, which would have helped in the evaluation of the usefulness of the test. ⁵  

Recent surveys of diagnostic studies published in ophthalmic journals have shown a poor quality of reporting, ^{60 71} which could result in the early adoption of new tests and erroneous clinical decisions. In this study, the overall quality of reporting of studies investigating the diagnostic accuracy of the HRT for glaucoma was suboptimal. Although all STARD items are valuable, they are not all equally important. Failure to report an item may mislead the reader but does not necessarily invalidate the evidence. In contrast, lack of masking, work-up bias (item 16), not reporting indeterminate results, or inclusion of patients with severe disease will almost always overestimate the diagnostic value of a test. According to Lijmer et al., ² the incomplete description of the study population or of the methods of diagnostic accuracy studies would inflate the diagnostic value of the test by up to 40% and 70%, respectively, when compared with articles that published sufficient detail. 

Some aspects were properly reported in most of the articles—for example, describing the reference standard and reporting the distribution of severity of disease are examples. An example of a high-quality manuscript would be the article by Medeiros et al. ³⁴ that reported 18 STARD items and included information about important aspects of any diagnostic accuracy study, such as design of the study, selection of groups and measures of diagnostic uncertainty. We hope that the results of this study and others alike ^{58 59 60} will hasten the improvement of methodological and reporting standards of future diagnostic studies. 

 Submitted for publication September 21, 2005; revised November 27, 2005, and January 14, 2006; accepted March 28, 2006.

 Disclosure: M. Shunmugam, None; A. Azuara-Blanco, 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.

 

Process for selection of the manuscripts.

Table 1.

 

STARD Checklist for the Evaluation of Studies of Diagnostic Accuracy³

Section and Topic	Item		On Page #
Title/abstract/keywords	1	Identify the article as a study of diagnostic accuracy (recommend MeSH heading “sensitivity and specificity”).
Introduction	2	State the research questions or study aims, such as estimating diagnostic accuracy or comparing accuracy between tests or across participant groups.
Methods
Participants	3	Describe the study population: the inclusion and exclusion criteria, setting and locations where the data were collected.
	4	Describe participant recruitment: Was recruitment based on presenting symptoms, results from previous tests, or the fact that the participants had received the index tests or the reference standard?
	5	Describe participant sampling: Was the study population a consecutive series of participants defined by the selection criteria in items 3 and 4? If not, specify how participants were further selected.
	6	Describe data collection: Was data collection planned before the index test and reference standard were performed (prospective study) or after (retrospective study)?
Test methods	7	Describe the reference standard and its rationale.
	8	Describe technical specifications of material and methods involved including how and when measurements were taken, and/or cite references for index tests and reference standard.
	9	Describe definition of and rationale for the units, cut offs and/or categories of the results of the index tests and the reference standard.
	10	Describe the number, training and expertise of the persons executing and reading the index tests and the reference standard.
	11	Describe whether or not the readers of the index tests and reference standard were blind (masked) to the results of the other test and describe any other clinical information available to the readers.
Statistical methods	12	Describe methods for calculating or comparing measures of diagnostic accuracy, and the statistical methods used to quantify uncertainty (e.g., 95% confidence intervals).
	13	Describe methods for calculating test reproducibility, if done.
Results
Participants	14	Report when study was done, including beginning and ending dates of recruitment.
	15	Report clinical and demographic characteristics of the study population (e.g. age, sex, spectrum of presenting symptoms, comorbidity, current treatments, recruitment centers).
	16	Report the number of participants satisfying the criteria for inclusion that did or did not undergo the index tests and/or the reference standard; describe why participants failed to receive either test (a flow diagram is strongly recommended).
Test results	17	Report time interval from the index tests to the reference standard, and any treatment administered between.
	18	Report distribution of severity of disease (define criteria) in those with the target condition; other diagnoses in participants without the target condition.
	19	Report a cross tabulation of the results of the index tests (including indeterminate and missing results) by the results of the reference standard; for continuous results, the distribution of the test results by the results of the reference standard.
	20	Report any adverse events from performing the index tests or the reference standard.
Estimates	21	Report estimates of diagnostic accuracy and measures of statistical uncertainty (e.g., 95% confidence intervals).
	22	Report how indeterminate results, missing responses and outliers of the index tests were handled.
	23	Report estimates of variability of diagnostic accuracy between subgroups of participants, readers or centers, if done.
	24	Report estimates of test reproducibility, if done.
Discussion	25	Discuss the clinical applicability of the study findings.

 Published, with permission, from Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative—Standards for Reporting of Diagnostic Accuracy. Clin Chem. 2003;41:1–6. © AACC.

Table 2.

 

Number and Frequency of the 25 STARD Items Reported in All Studies

Section and Topic	Item	Yes (%)	Partial (%)	No (%)
Title/abstract/keywords	1	28 (97)	1 (3)	0
Introduction	2	29 (100)	0	0
Methods
Participants	3	18 (62)	11 (38)	0
	4	21 (72)	5 (17)	3 (10)
	5	5 (17)	2 (7)	22 (76)
	6	7 (24)	0	22 (76)
Test methods	7	28 (97)	1 (3)	0
	8	7 (24)	22 (76)	0
	9	28 (97)	1 (3)	0
	10	9 (31)	11 (38)	9 (31)
	11	2 (7)	4 (14)	23 (79)
Statistical methods	12	3 (10)	26 (90)	0
	13	2 (7)	14 (48)	13 (45)
Results
Participants	14	2 (7)	1 (3)	26 (90)
	15	19 (66)	10 (34)	0
	16	1 (3)	1 (3)	27 (93)
Test results	17	4 (14)	10 (34)	15 (52)
	18	26 (90)	1 (3)	2 (7)
	19	22 (76)	0	7 (24)
	20	—	—	—
Estimates	21	5 (17)	24 (83)	0
	22	8 (28)	0	21 (72)
	23	10 (34)	—	—
	24	2 (7)	14 (48)	13 (45)
Discussion	25	29 (100)	0	0
Item 3: If this item was not fully reported, but at least one of the four aspects of the study was described (i.e. setting, location, inclusion or exclusion criteria) it was marked as “partial.”
Item 8: If the technical specifications of the HRT were reported but there was no information regarding when the test was performed, the item was scored as “partial.”
Item 10: If details of the expertise and number of people executing and reviewing the tests were partially described, this item was marked as “partial.”
Item 12 and 21: If there were estimates of diagnostic accuracy reported, but no data regarding measures of uncertainty (e.g., confidence intervals), the item was marked as “partial.”
Item 13 and 24: These items were marked “yes” if reproducibility analysis was performed and reported. If the study cited other published references of the reproducibility of the HRT, this item was scored as “partial.”
Item 17: If there was a time interval between the index test (HRT) and the reference standard, but there was no information about the stability of the disease or treatments administered between the tests, this item was scored as “partial.”
Item 20: This item was marked as “not applicable” due to the noninvasive nature of the HRT.
Item 23: Not all studies performed subgroup analyses, and so item 23 was “not applicable” in most of the publications (19/29).

  n = 29.

Table 3.

 

Summary of the Total Number of STARD Items Fully (Yes), Partially (Partial), and Not (No) Fulfilled by Each Study

Authors	Title	Journal	Yes	Partial	No
Medeiros FA, Zangwill LM, Bowd C, Weinreb RN	Comparison of the GDx VCC scanning laser polarimeter, HRT II confocal scanning laser ophthalmoscope, and stratus OCT optical coherence tomograph for the detection of glaucoma.	Arch Ophthalmol. 2004	18	3	2
Wollstein G, Garway-Heath DF, Hitchings RA	Identification of early glaucoma cases with the scanning laser ophthalmoscope.	Ophthalmology. 1998	15	3	5
Ford BA, Artes PH, McCormick TA, Nicolela MT, LeBlanc RP, Chauhan BC	Comparison of data analysis tools for detection of glaucoma with the Heidelberg Retina Tomograph.	Ophthalmology. 2003	15	6	3
Kesen MR, Spaeth GL, Henderer JD, Pereira ML, Smith AF, Steinmann WC	The Heidelberg Retina Tomograph versus clinical impression in the diagnosis of glaucoma.	AJO. 2002	15	5	3
Iester M, Mikelberg FS, Drance SM	The effect of optic disc size on diagnostic precision with the Heidelberg retina tomograph.	Ophthalmology. 1997	14	2	8
Iester M, Jonas JB, Mardin CY, Budde WM	Discriminant analysis models for early detection of glaucomatous optic disc changes	BJO. 2000	13	3	8
Sanchez-Galeana C, Bowd C, Blumenthal EZ, Gokhale PA, Zangwill LM, Weinreb RN	Using optical imaging summary data to detect glaucoma.	Ophthalmology. 2001	13	4	7
Miglior S, Guareschi M, Albe E, Gomarasca S, Vavassori M, Orzalesi N	Detection of glaucomatous visual field changes using the Moorfields regression analysis of the Heidelberg retina tomograph.	AJO. 2003	12	7	5
Swindale NV, Stjepanovic G, Chin A, Mikelberg FS	Automated analysis of normal and glaucomatous optic nerve head topography images.	IOVS. 2000	12	4	7
Greaney MJ, Hoffman DC, Garway-Heath DF, Nakla M, Coleman AL, Caprioli J	Comparison of optic nerve imaging methods to distinguish normal eyes from those with glaucoma.	IOVS. 2002	12	3	8
Miglior S, Casula M, Guareschi M, Marchetti I, Iester M, Orzalesi N	Clinical ability of Heidelberg retinal tomograph examination to detect glaucomatous visual field changes.	Ophthalmology. 2001	12	6	6
Bowd C, Chan K, Zangwill LM, Goldbaum MH, Lee TW, Sejnowski TJ, Weinreb RN	Comparing neural networks and linear discriminant functions for glaucoma detection using confocal scanning laser ophthalmoscopy of the optic disc.	IOVS. 2002	11	5	7
Bathija R, Zangwill L, Berry CC, Sample PA, Weinreb RN	Detection of early glaucomatous structural damage with confocal scanning laser tomography.	J Glaucoma. 1998	11	7	6
Wollstein G, Garway-Heath DF, Fontana L, Hitchings RA	Identifying early glaucomatous changes. Comparison between expert clinical assessment of optic disc photographs and confocal scanning ophthalmoscopy.	Ophthalmology. 2000	11	7	5
Zangwill LM, Chan K, Bowd C, Hao J, Lee TW, Weinreb RN, Sejnowski TJ, Goldbaum MH	Heidelberg retina tomograph measurements of the optic disc and parapapillary retina for detecting glaucoma analyzed by machine learning classifiers.	IOVS. 2004	11	4	8
Iester M, Mikelberg FS, Swindale NV, Drance SM	ROC analysis of Heidelberg Retina Tomograph optic disc shape measures in glaucoma.	Can J Ophthalmol. 1997	10	4	9
Uchida H, Brigatti L, Caprioli J	Detection of structural damage from glaucoma with confocal laser image analysis.	IOVS. 1996	10	7	6
Iester M, Mardin CY, Budde WM, Junemann AG, Hayler JK, Jonas JB	Discriminant analysis formulas of optic nerve head parameters measured by confocal scanning laser tomography.	J Glaucoma. 2002	10	7	7
Iester M, De Ferrari R, Zanini M	Topographic analysis to discriminate glaucomatous from normal-optic nerve heads with a confocal scanning laser: new optic disc analysis without any observer input.	Surv Ophthalmol. 1999	10	6	7
Gundersen KG, Heijl A, Bengtsson B	Comparability of three-dimensional optic disc imaging with different techniques. A study with confocal scanning laser tomography and raster tomography.	Acta Ophthalmol Scand. 2000	9	5	9
Mardin CY, Horn FK, Jonas JB, Budde WM	Preperimetric glaucoma diagnosis by confocal scanning laser tomography of the optic disc.	BJO. 1999	9	7	8
Mardin CY, Horn FK	Influence of optic disc size on the sensitivity of the Heidelberg Retina Tomograph.	Graefes Arch Clin Exp Ophthalmol. 1998	9	7	8
Mikelberg FS, Parfitt CM, Swindale NV, et al.	Ability of the Heidelberg Retina Tomograph to detect early glaucomatous visual field loss.	J Glaucoma. 1995	9	7	7
Mardin CY, Hothorn T, Peters A, Junemann AG, Nguyen NX, Lausen B	New glaucoma classification method based on standard Heidelberg-Retina Tomograph parameters by bagging classification trees.	J Glaucoma. 2003	9	5	9
Gundersen KG, Asman P	Comparison of ranked segment analysis (RSA) and cup-to-disc ratio in computer-assisted optic disc evaluation.	Acta Ophthalmol Scand. 2000	8	6	9
Park KH, Caprioli J	Development of a novel reference plane for the Heidelberg retina tomograph with optical coherence tomography measurements.	J. Glaucoma. 2002	8	8	7
Zangwill LM, Bowd C, Berry CC, Williams J, Blumenthal EZ, Sanchez-Galeana CA, Vasile C, Weinreb RN	Discriminating between normal and glaucomatous eyes using the Heidelberg Retina Tomograph, GDx Nerve Fiber Analyzer, and Optical Coherence Tomograph.	Arch Ophthalmol. 2001	7	7	9
Park HJ, Caprioli J	Circumferential profiles of peripapillary surface height with confocal scanning laser ophthalmoscopy.	Korean J. Ophthalmol. 1997	7	8	8
Vihanninjoki K, Teesalu P, Burk RO, Laara E, Tuulonen A, Airaksinen PJ	Search for an optimal combination of structural and functional parameters for the diagnosis of glaucoma. Multivariate analysis of confocal scanning laser tomograph, blue-on-yellow visual field and retinal nerve fiber layer data.	Graefes Arch Clin Exp Ophthalmol. 2000	5	6	12

  n = 29. Listed in order from the highest number scored for “Yes.” AJO, Am J Ophthalmol.; BJO, Br J Ophthalmol.

The authors thank M. A. Rehman Siddiqui for comments and suggestions.