Confocal Scan Laser Ophthalmoscope for Diagnosing Glaucoma: A Systematic Review and Meta-analysis

REVIEW ARTICLE

Confocal Scan Laser Ophthalmoscope for Diagnosing Glaucoma: A Systematic Review and Meta-analysis Mohsen Yaghoubi, MS,* Maziar Moradi-Lakeh, MD, MPH,* Mehdi Mokhtari-Payam, MS,Þ Ghasem Fakhraie, MD,þ and Farhad Shokraneh, MS§

Abstract: This systematic review was performed to estimate the diagnostic accuracy of the confocal scanning laser ophthalmoscopy in diagnosing glaucoma. We did a sensitive electronic search to find relevant studies. Two reviewers independently screened relevant articles and extracted required data about study methods and reported results of sensitivity and specificity. A meta-analysis was conducted for pooling data to compare different editions of the Heidelberg Retina Tomograph (HRT) with one of its alternatives, scanning laser polarimetry (GDx) with the criteria of ‘‘visual field defect’’ and ‘‘changes of nerve fiber layer’’ as the reference standard. We identified 37 evaluations from 28 relevant primary studies. In these studies, 9573 eyes (4883 glaucomatous and 4689 non-glaucomatous) were assessed with regards to the reference standard using one of the HRT editions with or without GDx. Diagnostic odds ratios were 9.35 [95% confidence interval (CI): 7.58Y11.53] for HRT, 11.84 (95% CI: 9.97Y14.06) for HRT II, 11.86 (95% CI: 9.16Y15.35) for HRT III, and 21.33 (95% CI, 13.56Y33.55) for GDx. Although GDx was more accurate than HRT, all editions of HRT had acceptable performance in diagnosing glaucomatous eyes with an ophthalmologist’s clinical examination as the reference standard. Key Words: glaucoma, confocal laser scanning, Heidelberg Retina Tomograph, scanning laser polarimeter (Asia Pac J Ophthalmol 2015;4: 32Y39)

T

he diagnosis of glaucoma currently is mainly based on an ophthalmologist’s examination. However, objective tests provided by modern technology have become increasingly necessary for screening, diagnosing, monitoring the progress of glaucoma, documenting patients’ status, and conducting research. Objective testing is usually based on detection of structural and functional changes concerning the optic nerve head (ONH) and the peripapillary retinal nerve fiber layer (RNFL).1 Visual field examination is considered a subjective measure of central and peripheral vision. The assessment of RNFL thickness has traditionally been based on a clinical examination, with primarily qualitaive documentation of changes. The latest progress in diagnostic

From the *Clinical Knowledge Management Unit, Department of Community Medicine, School of Medicine, Iran University of Medical Sciences, Tehran; †Iran University of Medical Science, Tehran, Iran; ‡Glaucoma Service, Farabi Eye Hospital, Tehran University of Medical Sciences, Tehran; and §Cochrane Schizophrenia Group, the Institute of Mental Health, a partnership between the University of Nottingham and Nottinghamshire Healthcare NHS Trust, Nottingham, UK. Received for publication November 17, 2013; accepted July 24, 2014. This study was funded by the Iranian National Institute of Health Researches (NIHR) under contract 241/M/308. The authors have no conflicts of interest to declare. Reprints: Maziar Moradi-Lakeh MD, Department of Community Medicine, Iran University of Medical Sciences, Hemmat Highway, Tehran, Iran. E-mail: [email protected]; [email protected]. Copyright * 2015 by Asia Pacific Academy of Ophthalmology ISSN: 2162-0989 DOI: 10.1097/APO.0000000000000085

32

www.apjo.org

technology makes objective and quantitative measurements of the RNFL possible now.2 Changes in the thickness of the peripapillary RNFL may occur before changes in visual field. Therefore, objective testing for structural changes may be more important than subjective measurements of functional changes for early detection of glaucoma. Structural evaluation of the ONH and the peripapillary RNFL by imaging devices such as the optical coherence tomograph, scanning laser polarimeter (GDx), nerve fiber analyzer, confocal scanning laser ophthalmoscope (cSLO), and the Heidelberg Retina Tomograph (HRT) has gained much ground in the early detection of glaucoma.3 There is a notable lack of studies that review the outcome of the HRT and GDx evaluations in groups of glaucoma patients and those with suspected glaucoma.3 The aim of this study was to assess the accuracy of structural evaluation tests for diagnosing glaucoma based on odds ratios (ORs). A systematic literature review and a meta-analysis were performed to evaluate the accuracy of the instruments used to detect the structural changes of glaucoma. The study focused on the HRT and GDx, which provide topographic examination of the ONH and measurement of retardation, respectively.3 The diagnostic accuracy of a prototype of a cSLO, known as the HRT test, was reviewed and compared with a possible rival technology (scanning laser polarimeter and its prototype, GDx). Each of these instruments provides examination by a different optical rule. The HRT measures the topography of the ONH and does not differentiate between different layers of the retina, whereas GDx measures retardation, which is a surrogate for RNFL thickness.4

MATERIALS AND METHODS Resources and Search Methods We performed PICO framework (determining patient problem or population, intervention, comparison, and outcome) in order to define research questions for a focused search.

Problem/Population Eyes that are suggestive of having glaucomatous changes were included. Glaucoma is a group of eye conditions that result in progressive optic neuropathy, with characteristic morphological changes at the ONH and associated visual field defects.1

Intervention/Index Test The HRT (Heidelberg Engineering, Germany) is a cSLO. The instrument utilizes a 670-nm diode laser with a maximum output of 180 W to acquire images of the posterior segment of the eye. For the purposes of diagnosing and monitoring of glaucoma, the HRT is used to generate 3-dimensional topographic images of the ONH. Three models of the HRT are available: HRT I, HRT II, and HRT III.3

Asia-Pacific Journal of Ophthalmology

&

Volume 4, Number 1, January/February 2015

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

Asia-Pacific Journal of Ophthalmology

&

Volume 4, Number 1, January/February 2015

Comparison The GDx is seen as a possible alternative to cSLO. Optical coherence tomography was not included in the comparison, even though it is gaining importance in the diagnosis and monitoring of glaucoma, because it was outside the approved study protocol.

Outcome Diagnostic accuracy, sensitivity, and specificity were compared with a reference standard test for the diagnosis of glaucoma. We used the clinical diagnosis of an ophthalmologist, based on at least 2 criteria (visual field defect and changes in nerve fiber layer) as the acceptable reference standard test for the final diagnosis of glaucoma.5 Studies that did not use both criteria were excluded although some of the included studies had additional criteria. To retrieve all the relevant studies, we used a sensitive librarian-assisted search strategy in relevant databases including MEDLINE, HTA, NEED, CDSR, DARE, CENTRAL (via Ovid SP), EMBASE (via EMBASE.com), Elsevier’s Science Direct, Springer Link, Wiley Online Library, and Proquest Theses Database. Search in databases was performed on October 10, 2012. We used search terms based on PICO. We only included problem and intervention to achieve higher sensitivity. The

Ophthalmoscope for Diagnosing Glaucoma

following search strategy was developed for MEDLINE via Ovid SP and then modified for the other databases: 1-expglaucoma/2-hydrophthalmos/3-glaucoma$.ti,ab/4hydrophthalmos.ti,ab/5-or/1-4/6-microscopy, confocal/ 7-(confocal adj6 microscop$).ti,ab/8-((confocal adj6 scan$) or cslo).ti,ab/ 9-(confocal adj6 laser).ti, ab/10-(heidelberg adj6 retina? adj6 tomograph$).ti, ab/11-hrt.ti, ab/ 12-or/6-11/13-5 and 12/14-limit 13 to animals/15-limit 13 to humans/16-14 not 15/17-13 not 16 We did not limit our search to a certain language or time period to avoid missing possible related works. Besides, we factored in manual reference checking and citation tracking of related papers through Thomson Reuters’ Web of Science, Elsevier’s Scopus, and Google Scholar. Also, the search focused on secondary studies for rapid health technology assessment.6 The search method of high sensitivity proposed by the Centre for Reviews and Dissemination was used in order to get all related secondary studies such as meta-analyses, systematic reviews, health technology assessments, and economic evaluations, as well as review articles.

Inclusion Criteria The following types of studies were included: those compatible with our PICO, those conducted on human subjects with

FIGURE 1. Flow diagram of articles included in the systematic review of diagnostic accuracy of HRT editions and GDx. * 2015 Asia Pacific Academy of Ophthalmology

www.apjo.org

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

33

34

www.apjo.org

2007

Andreou et al18

Badala et al20

2006

2007

Lester et al21

2005

2005

2005

15

2007

2007

Zangwill et al17

18

2007

23

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

2007

2007

Badala et al20

de Leon-Ortega et al

2010

2010

Rao et al34

Healey et al

2008

Toth et al26

33

2008

Yip et al27

19

2008

Borque et al24

Ferreras et al

2007

de Leon-Ortega et al19

89

46

98

488

118

58

93

348

89

102

62

97

242

78

46

79

1464

13

52

90

103

78

31

99

94

19

52

49

66

80

283

342

296

68

20

148

51

50

53

13

46

31

148

66

Gl+

167

92

177

1952

131

110

183

450

167

133

161

191

261

265

264

141

190

438

818

445

131

40

241

123

99

98

131

92

133

241

141

Total

Pr

0.467

0.5

0.4463

0.75

0.0992

0.4727

0.4918

0.2289

0.467

0.2331

0.6149

0.4921

0.0728

0.1962

0.1856

0.4681

0.4211

0.6461

0.4181

0.6652

0.519

0.5

0.6141

0.4146

0.5051

0.5408

0.0992

0.5

0.2331

0.6141

0.4681

Test

HRT III

HRT III

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT II

HRT

HRT

HRT

HRT

HRT

HRT

HRT

HRT

HRT

GDx

GDx

GDx

GDx

GDx

Sn (95% CI)

0.77 (0.67Y0.86)

0.80 (0.77Y0.83)

0.56 (0.47Y0.61)

0.64 (0.62Y0.67)

0.71 (0.46Y0.95)

0.65 (0.52Y0.78)

0.74 (0.80Y0.97)

0.65 (0.55Y0.74)

0.62 (0.51Y0.72)

0.79 (0.65Y0.93)

0.71 (0.62Y0.89)

0.80 (0.72Y0.88)

0.95 (0.85Y1.00)

0.44 (0.31Y0.59)

0.51 (0.37Y0.65)

0.73 (0.62Y0.84)

0.90 (0.83Y0.96)

0.66 (0.61Y0.72)

0.65 (0.60Y0.70)

0.75 (0.69Y0.81)

0.76 (0.65Y0.86)

0.80 (0.62Y0.97)

0.84 (0.78Y0.90)

0.84 (0.74Y0.94)

0.60 (0.47Y0.74)

0.83 (0.73Y0.93)

0.23 (0.00Y0.45)

0.80 (0.68Y0.92)

0.80 (0.66Y0.94)

0.76 (0.69Y0.83)

0.87 (0.79Y0.95)

Sp (95% CI)

0.83 (0.75Y0.90)

0.87 (0.02Y0.09)

0.96 (0.92Y0.99)

0.86 (0.83Y0.89)

0.79 (0.71Y0.86)

0.79 (0.68Y0.89)

0.95 (0.91Y0.99)

0.77 (0.72Y0.81)

0.92 (0.86Y0.97)

0.70 (0.61Y0.79)

0.82 (0.72Y0.91)

0.86 (0.79Y0.93)

0.81 (0.76Y0.86)

0.93 (0.88Y0.96)

0.90 (0.85Y0.93)

0.80 (0.71Y0.89)

0.89 (0.83Y0.95)

0.92 (0.88Y0.96)

0.69 (0.65Y0.73)

0.91 (0.86Y0.94)

0.79 (0.68Y0.89)

0.75 (0.56Y0.93)

0.72 (0.63Y0.81)

0.96 (0.91Y1.00)

0.94 (0.87Y1.00)

0.86 (0.76Y0.96)

0.96 (0.92Y0.99)

0.89 (0.80Y0.98)

0.72 (0.63Y0.81)

0.91 (0.85Y0.97)

0.80 (0.71Y0.89)

4.5 (2.8Y7.3)

6.2 (2.9Y13.2)

14 (5.2Y37.7)

4.5 (3.6Y5.6)

3.4 (2.1Y5.5)

3.1 (1.8Y5.3)

14.9 (6.1Y36.4)

2.8 (2.2Y3.6)

7.8 (3.8Y16.0)

2.6 (1.9Y3.7)

3.9 (2.3Y6.8)

5.6 (3.4Y9.1)

4.9 (3.7Y6.4)

5.9 (3.4Y10.3)

5.0 (3.1Y8.1)

3.7 (2.3Y5.9)

8.2 (4.8Y14.1)

8.3 (4.8Y14.2)

2.1 (1.8Y2.4)

8.4 (5.0Y14.2)

3.6 (2.2Y6.0)

3.2 (1.5Y7.1)

3.0 (2.1Y4.2)

20.1 (6.6Y60.9)

9.9 (3.2Y30.3)

5.9 (2.8Y12.6)

7.4 (2.3Y23.6)

7.3 (3.2Y16.8)

2.9 (2.0Y4.1)

8.4 (4.4Y16.2)

4.3 (2.7Y6.9)

PLR (95% CI)

NLR (95% CI)

0.28 (0.18Y0.42)

0.23 (0.13Y0.41)

0.46 (0.36Y0.59)

0.42 (0.39Y0.45)

0.37 (0.16Y0.86)

0.44 (0.30Y0.66)

0.27 (0.19Y0.38)

0.45 (0.35Y0.60)

0.41 (0.31Y0.55)

0.30 (0.15Y0.60)

0.35 (0.25Y0.49)

0.23 (0.16Y0.35)

0.07 (0.01Y0.44)

0.60 (0.47Y0.77)

0.55 (0.41Y0.73)

0.34 (0.22Y0.51)

0.12 (0.06Y0.22)

0.37 (0.31Y0.44)

0.51 (0.43Y0.59)

0.27 (0.22Y0.33)

0.30 (0.20Y0.47)

0.27 (0.11Y0.66)

0.22 (0.15Y0.33)

0.16 (0.09Y0.31)

0.42 (0.30Y0.60)

0.20 (0.11Y0.36)

0.78 (0.58Y1.06)

0.22 (0.13Y0.40)

0.28 (0.14Y0.57)

0.26 (0.20Y0.35)

0.16 (0.09Y0.31)

DOR (95% CI)

16.3 (12.1Y22)

26.8 (14Y51.2)

32.4 (17.6Y59.6)

10.7 (10.3Y11.1)

9.2 (4.1Y20.6)

7.0 (4.8Y10.1)

55.2 (32.5Y93.7)

6.2 (5.5Y7.0)

18.8 (12.4Y28.3)

8.8 (5.6Y13.8)

11.2 (8.2Y15.2)

23.8 (17.8Y31.9)

74.2 (8.9Y618.6)

9.8 (7.4Y13.0)

9.2 (7.1Y11.9)

10.8 (7.9Y14.8)

71.2 (45.1Y112.3)

22.3 (18.1Y27.5)

4.1 (3.9Y4.3)

31.0 (25.4Y37.9)

11.9 (8.4Y16.9)

12.0 (3.9Y37.3)

13.5 (11.0Y16.5)

122.5 (46.3Y323.9)

23.3 (9.9Y55.0)

30.0 (15.8Y57.0)

9.4 (1.8Y48.2)

32.4 (15.9Y65.9)

10.3 (6.3Y17.0)

32.0 (22.8Y45.0)

26.8 (17.7Y40.6)

&

Andreou et al

2006

213

215

75

110

155

476

149

63

20

93

72

49

43

118

46

102

93

75

Glj

Asia-Pacific Journal of Ophthalmology

Larrosa et al16

Robin et al

Harasymowycz et al*

Harasymowycz et al†14

2004

Medeiros et al12

14

2000

Ahn et al11

Weinreb et al

2010

2010

Saarela et al31

32

2009

Saito et al30

Saarela et al

22

2008

2000

Wollstein et al

Kanamori et al13

1998

Bathija et al9

10

1996

Uchida et al8

2008

2007

Kanamori et al13

Toth et al

2006

Medeiros et al12

26

Year

2004

Author

TABLE 1. Specifications of 28 Studies That Estimated Diagnostic Accuracy of HRT, HRT II, or HRT III (With or Without Diagnostic Accuracy of GDx) in Comparison With Standard Reference Test in Diagnosis of Glaucoma

Yaghoubi et al Volume 4, Number 1, January/February 2015

* 2015 Asia Pacific Academy of Ophthalmology

13.1 (9.3-18.4)

13.6 (11.8Y15.7)

0.42 (0.32Y0.55)

0.48 (0.37Y0.63)

0.34 (0.29Y0.41) 4.6 (3.1Y7.1) 0.85 (0.83Y0.90) 0.71 (0.62Y0.79) HRT III 0.7597 516 392 124 2010 Zheng et al

35

Moreno-Montanes et al

Volume 4, Number 1, January/February 2015

Ophthalmoscope for Diagnosing Glaucoma

glaucoma that used cSLO with or without GDx, and those that evaluated at least 1 of the outcomes of our research. No age, sex, nationality, or race limitations were imposed.

Exclusion Criteria Laboratory studies conducted on animals were excluded. Studies that did not use our stated criteria as the reference standard test were excluded.

Study Selection Full texts of the relevant articles were critically appraised for eligibility criteria by 2 researchers independently. The third colleague were asked to assess the articles in case of disagreement.

Gl+ indicates glaucomatous eyes; Glj, nonglaucomatous eyes; Sn, sensitivity; Sp, specificity; LL and UL, lower and upper limits of 95% CI. *Right-sided eyes. †Left-sided eyes.

HRT III 177 79 2010 Rao et al34

98

0.4463

0.56 (0.47Y0.67)

0.91 (0.85Y0.96)

6.2 (3.2Y12.0)

8.5 (5.7Y12.7)

12.5 (10.0-15.6)

0.66 (0.55Y0.79) 5.6 (2.6Y12.0)

5.3 (3.3Y8.5) 0.88 (0.82Y0.93)

0.93 (0.87Y0.98) 0.39 (0.28Y0.49)

0.63 (0.53Y0.73)

HRT III

HRT III 96

0.456 182 83 99

2009

2008

137

0.412

3.9 (3.2Y4.9)

25

Oddone et al29

233

24.4 (18.0Y33.1)

0.46 (0.30Y0.68) 1.7 (1.3Y2.3) 0.57 (0.47Y0.68)

0.83 (0.75Y0.91) 0.83 (0.76Y0.91) HRT III

HRT III 0.512 166 85

183 90

2008 Ferreras et al28

81

2007 Ferreras et al23

&

Data Extraction

93

0.4918

0.74 (0.64Y0.83)

4.9 (3.1Y7.8)

0.20 (0.13Y0.32)

Asia-Pacific Journal of Ophthalmology

* 2015 Asia Pacific Academy of Ophthalmology

Data extraction was carried out by 2 researchers. Data pertaining to definition of glaucoma and specificities of the reference standard test, applied diagnostic test, number of studied and glaucomatous eyes, and specificities of diagnostic tests including sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), together with their confidence intervals (CIs), were extracted from the articles. In cases where the mentioned indices were not reported directly, they were calculated and recorded using other indices. Diagnostic ORs (DORs) and its CIs have also been calculated by the following formulas: PLR = sensitivity / (1 j specificity) NLR = (1 j sensitivity) / specificity DOR = PLR / NLR LnDOR = natural logarithm of DOR SE (LnDOR) = (1 / A + 1 / B + 1 / C + 1 / D)^0.5 SE (LnPLR) = (1 / A + 1 / B - 1 / Dis j 1 / NDis)^0.5 SE (LnNLR) = (1 / C + 1 / D - 1 / Dis Y 1 / NDis)^0.5 In a 2  2 table between the results of a reference standard and another test, A, B, C, and D cells contain true-positive, false-positive, false-negative, and true-positive cases, respectively. Dis and NDis remark the number of glaucomatous and nonglaucomatous eyes, respectively. 95% CI for DOR: exp(LnDOR T 1.96 SE for LnDOR) 95% CI for PLR: exp(LnPLR T 1.96 SE for LnPLR) 95% CI for NLR: exp(LnNLR T 1.96 SE for LnNLR)

Quantitative Data Synthesis Meta-analysis was performed using the STATA statistical package version 13. First, we assessed heterogeneity by Q test (P G 0.1 as meaningful) in order to estimate DOR. We used random or fixed model for pooling measures of DOR (and 95% CIs) in case of absence or presence of considerable heterogeneity, respectively. Presence of publication bias was assessed by Egger regression test (Fig. 1).

RESULTS Summary of specifications of 37 examinations was obtained from 28 studies and taken into the account in the final analysis as shown in Table 1. In these studies, diagnostic accuracy of 1 of the systems of HRT, HRT II, or HRT III (together with diagnostic accuracy of GDx in 5 studies) had been tested. In these studies, 9573 eyes (4883 glaucomatous and 4689 nonglaucomatous) were assessed by reference standard test and by at least 1 of the HRT editions with or without GDx (Table 2). Figures 2 to 5 are forest plots showing PLR and NLR and their pooled estimates based on meta-analysis related to each study conducted on HRT, HRT II, HRT III, and GDx, respectively. The PLR (right-side bars) and NLR (left-side bars) for each study can be seen in front of the name of first author; the www.apjo.org

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

35

Asia-Pacific Journal of Ophthalmology

Yaghoubi et al

&

Volume 4, Number 1, January/February 2015

TABLE 2. Results of Pooled Estimated in PLR, NLR, and DOR in HRT, HRT II, HRT III, and GDx Pooled Estimates Device

Total

GL+

GLj

PLR

NLR

DOR

HRT

2433

1311

1120

2.86 (2.53Y3.23)

0.36 (0.33Y0.39)

9.35 (7.58Y11.53)

HRT II

4776

2369

2408

4.15 (3.74Y4.60)

0.41 (0.39Y0.44)

11.84 (9.97Y14.06)

HRT III

1716

949

767

3.51 (2.98Y4.13)

0.42 (0.38Y0.47)

11.86 (9.16Y15.35)

648

254

394

4.14 (3.24Y5.28)

0.37 (0.31Y0.45)

21.33 (13.56Y33.55)

GDx

diamonds at the bottom line of each figure show the pooled estimates (vertical diagonals) of PLR and NLR and their 95% CIs (horizontal diagonals). Overall, we found 9 qualified studies in which 2433 eyes (including 1120 nonglaucomatous and 1311 glaucomatous) were examined using HRT. The sensitivity was 0.76 ( ranged 0.60Y0.84) and the specificity was 0.86 (ranged 0.69Y0.96). There was statistically significant heterogeneity between results of the studies (Q = 71.7, P G 0.001, I2 = 88%). The pooled estimate of DOR (95% CIs) was 9.35 (7.58Y11.53) for HRT (Fig. 2). Fifteen studies on 4776 eyes (including 2408 nonglaucomatous and 2369 glaucomatous) had assessed diagnostic accuracy of HRT II. The sensitivity was 0.71 ( ranged 0.44Y0.90) and the specificity was 0.85 ( ranged 0.70Y0.96). There was considerable heterogeneity between results of the studies (Q = 43.29, P G 0.001, I2 = 67%), and pooled estimate of DOR (95% CIs) was 11.84 (9.97Y14.06) for HRT II (Fig. 3). Diagnostic accuracy of HRT III had been assessed in 8 eligible studies on the 1716 eyes (949 glaucomatous and 767 nonglaucomatous). The sensitivity was 0.72 ( ranged 0.39Y0.83) and the specificity as 0.85 ( ranged 0.57Y0.93). There was heterogeneity between results of the studies (Q = 18.61, P = 0.001, I 2 = 62.4%), and DOR (95% CIs) was 11.86 (9.16Y15.35) by random model (Fig. 4). We found 3 studies 19 that had directly compared HRT II and HRT III. Pooled DORs (95% CI) for these studies were 3.46 (2.79Y4.12) for HRT II and 2.86 (2.51Y3.21) for HRT III.

There were 5 qualified studies on the 648 eyes (254 glaucomatous and 394 nonglaucomatous) using GDx with at least 1 of the editions of HRT. Range and median were 0.23Y0.87 and 0.80 for sensitivity and 0.72Y0.96 and 0.89 for specificity. There was heterogeneity between results of these studies (Q = 5.61, P = 0.23, I2 = 28.7%). Diagnostic OR (95% CIs) was 21.33 (13.56-33.55) for GDx by random model (Fig. 5). The HRT arms of these 5 studies12 contain 3 HRT II, 1 HRT I and 1 HRT III. We combined all these HRT editions and calculated DOR; there was not a significant heterogeneity between the HRT results (Q = 2.71, P = 0.607, I2 = 0.0%) and the pooled estimate for DOR (95% CI) was 12.44 (8.45Y18.30) Egger regression test showed significant bias in the results of HRT (P = 0.01); there were no significant publication bias in the reported results of HRT II, HRT III, and GDx

DISCUSSION The aim of this literature review was to evaluate the accuracy of a CSLO and GDx for diagnosing glaucoma. In this meta-analysis, we reviewed 28 studies comparing HRT and GDx in the diagnosis of glaucoma based on the DOR. We chose the DOR because it is a helpful measure for collating diagnostic tests via meta-analysis, and it assigns accuracy with a unique value index rather than sensitivity and specificity. The technology of a confocal scanning microscope, applied as an HRT system, is safe with no significant reports of adverse effects on users or patients. If the HRT results because

FIGURE 2. Positive likelihood ratio (right) and NLR (left) of HRT in 9 primary studies and the pooled estimates based on meta-analysis.

36

www.apjo.org

* 2015 Asia Pacific Academy of Ophthalmology

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

Asia-Pacific Journal of Ophthalmology

&

Volume 4, Number 1, January/February 2015

Ophthalmoscope for Diagnosing Glaucoma

FIGURE 3. Positive likelihood ratio (right) and NLR (left) of HRT II in 15 primary evaluations and the pooled estimates based on meta-analysis.

of presence of publication bias are disregarded, accuracy of both the HRT II and the HRT III in diagnosing glaucoma is acceptable (Table 2). Diagnostic ORs may be between 0 and +V and will increase as the sensitivity and specificity of the diagnostic tests are approximately closer to 100%. In this case, odds of having glaucoma in eyes diagnosed positive by HRT II and HRT III are approximately 11.84 and 11.86 times of odds of having glaucoma in eyes with negative test results, respectively. It is of utmost importance to note that the DOR is not affected by the prevalence of the disease in the examined society. Estimates of DOR for HRT III and HRT II were almost the same, with the DOR of the HRTIII higher than the DOR of the HRT II. A direct comparison of the DOR estimate of HRT II to HRT III shows that the diagnostic accuracy of HRT II is better

than that of HRT III in 3 studies. Interestingly, Zelefsky et al.37 and de Leon-Ortega et al24 reported a better sensitivity of HRT III compared with HRT II for glaucoma detection. However, HRT III had lower specificity, so those authors did not reach a clear conclusion that HRT III was more accurate than its older editions. The GDx system is an alternative to HRT systems for diagnosing glaucoma. Our review found that GDx had a DOR of 21.33 (95% CI, 13.56Y33.55). This suggests improved diagnostic accuracy with GDx over the HRT editions; however, this should be interpreted with caution because of the large CIs. Also, it should be noted that HRT and GDx use different mechanisms rule in order to diagnostic glaucoma; therefore, HRT provides mostly an analysis of the ONH parameters,

FIGURE 4. Positive likelihood ratio (right) and NLR (left) conducted on HRT III in 8 primary studies and the pooled estimates based on meta-analysis. * 2015 Asia Pacific Academy of Ophthalmology

www.apjo.org

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

37

Asia-Pacific Journal of Ophthalmology

Yaghoubi et al

&

Volume 4, Number 1, January/February 2015

FIGURE 5. Positive likelihood ratio (right) and NLR (left) conducted on GDx in 5 primary studies and the pooled estimated based on meta-analysis.

whereas GDx is used for comparison as it focuses on the RNFL instead. The demographic characteristics of the populations (Asian, African, and white eyes) in the included studies, such as age, sex, and ethnicity, were not taken into account. This is another limitation of our study, as these characteristics could have affected the diagnostic accuracy of each device. In addition, severity of glaucoma was not detectable in included studies. Mowatt et al38 conducted a meta-analysis to evaluate several screening tests for glaucoma and concluded that no diagnostic test was more accurate than another, thus future search was needed. Kwartz et al3 performed an investigation to estimate the efficacy of the HRT and GDx and included that neither device provided adequate information for the diagnosis of glaucoma. The authors also suggested that the HRT and GDx could be used for ‘‘remote’’ or ‘‘telemedicine’’ clinics, where review of data and patient examinations were separated by time and/or location.

2. Zangwill LM1, Bowd C. Retinal nerve fiber layer analysis in the diagnosis of glaucoma. Curr Opin Ophthalmol. 2006;17:120Y131.

CONCLUSIONS

7. Centre for Reviews and Dissemination (CRD). Systematic Reviews: CRD’s Guidance for Undertaking Reviews in Health Care. York: University of York, 2008.

Comprehensive ophthalmic examination remains the reference standard in the diagnosis of glaucoma, and no other tests surpass its accuracy. However, the editions of the HRTs and GDx have acceptable performances in diagnosing glaucomatous eyes compared with the ophthalmologist’s clinical examination as the reference standard. Scanning laser polarimeter appeared to be more accurate than HRT. These diagnostics may be helpful in specific situations that require a more objective assessment of glaucoma such as research or monitoring its clinical progression. In addition, economic impact should be taken into consideration when evaluating the tools used in an ophthalmological care system. ACKNOWLEDGEMENTS The authors thank the NIHR staff, especially Ms Madadi, for the support. REFERENCES 1. Shah NN, Bowd C, et al. Combining structural and functional testing for detection of glaucoma. Ophthalmology. 2006;113:1593Y1602.

38

www.apjo.org

3. Kwartz AJ, Henson DB, Harper RA, et al. The effectiveness of the Heidelberg Retina Tomograph and laser diagnostic glaucoma scanning system (GDx) in detecting and monitoring glaucoma. Health Technol Assessment. 2006;9:1Y132, iii. 4. Kass MA, Heuer DK, Higginbotham EJ, et al. The Ocular Hypertension Treatment Study: a randomized trial determines that topical ocular hypotensive medication delays or prevents the onset of primary open-angle glaucoma. Arch Ophthalmol. 2002;120:701Y713, discussion 829Y30. 5. Alberta Heritage Foundation For Medical Research (AHFMR): confocal scanning laser ophthalmoscopy and scanning laser polarimetry for early diagnosis of glaucoma. TechNote 55, 2006, 39p. 6. Health Technology Assessment Handbook, Danish Centre for Health Technology Assessment, National Board of Health, Denmark. 2nd ed. 2008;45Y46. Available at: http://www.dacehta.dk.

8. Uchida H, Brigatti L, Caprioli J. Detection of structural damage from glaucoma with confocal laser image analysis. Invest Ophthalmol Vis Sci. 1996;37:2393Y2401. 9. Bathija R, Zangwill L, Berry CC, et al. Detection of Early Glaucomatous Structural Damage With Confocal Scanning Laser Tomography. San Diego, CA: Department of Ophthalmology, University of California, 1998. 10. Wollstein G, Garway-Heath DF, Fontana L, et al. Identifying early glaucomatous changes. Ophthalmology. 2000;107:2272Y2277. 11. Ahn BS, Kee C. Ability of a confocal scanning laser ophthalmoscope (TopSS) to detect early glaucomatous visual field defect. Br J Ophthalmol. 2000;84:852Y855. 12. Medeiros FA, Zangwill LM, Bowd C, et al. 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;122:827Y837. 13. Kanamori A, Nagai-Kusuhara A, Escan˜o MFT, et al. Comparison of confocal scanning laser ophthalmoscopy, scanning laser polarimetry

* 2015 Asia Pacific Academy of Ophthalmology

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

Asia-Pacific Journal of Ophthalmology

&

Volume 4, Number 1, January/February 2015

and optical coherence tomography to discriminate ocular hypertension and glaucoma at an early stage. Graefe’s Arch Clin Exp Ophthalmol. 2006;244:58Y68. 14. Harasymowycz PJ, Papamatheakis DG, Fansi AK, et al. Validity of screening for glaucomatous optic nerve damage using confocal scanning laser ophthalmoscopy (Heidelberg Retina Tomograph II) in High-risk populations. Ophthalmology. 2005;112:2164Y2171. 15. Robin TA, Ller AM, Rait J, et al. Performance of community-based glaucoma screening using frequency doubling technology and Heidelberg Retinal Tomography. Ophthalmic Epidemiol. 2005;12:167Y178. 16. Larrosa JM, polo V, Ferreras A, et al. Multivariate analysis of structural parameters of the optic nerve head assessed by means of confocal scanning laser (Heidelberg Retina Tomograph II). Ann Ophthalmol (Skokie). 2006;38:329Y338. 17. Zangwill LM, Jain S, Racette L, et al. The effect of disc size and severity of disease on the diagnostic accuracy of the Heidelberg Retina Tomograph Glaucoma Probability Score. Invest Ophthalmol Vis Sci. 2007;48:2653Y2660. 18. Andreou PA, Wickremasinghe SS, Asaria RH, et al. A comparison of HRT II and GDx imaging for glaucoma detection in a primary care eye clinic setting. Eye. 2007;21:1050Y1055. 19. de Leon-Ortega JE, Sakata LM, Monheit BE, et al. Comparison of diagnostic accuracy of Heidelberg Retina Tomograph II and Heidelberg Retina Tomograph 3 to discriminate glaucomatous and nonglaucomatous eyes. J Ophthalmol. 2007;144:525Y532. 20. Badala F, Nouri-Mahdavi K, Raoof DA, et al. Optic disk and nerve fiber layer imaging to detect glaucoma. J Ophthalmol. 2007;144:724Y732. 21. Iester M, Zanini M, Vittone P, et al. Detection of glaucomatous optic nerve head by using Heidelberg topographic maps. Eye. 2007;21:609Y613. 22. Saarela V, Airaksinen PJ. Heidelberg retina tomography parameters of the optic disc in eyes with progressive retinal nerve fibre layer defects. Acta Ophthalmol. 2008;86:603Y608. 23. Ferreras A, Pablo L, Paharin AB, et al. Diagnostic ability of the Heidelberg Retina Tomograph 3 for glaucoma. Ophthalmology. 2007;144:1981Y1987. 24. Borque E, Ferreras A, Polo v, et al. Evaluation of valuation of four new discriminate functions for HRT II in glaucoma diagnosis. Arch Soc Esp Oftalmol. 2008;83:349Y356. 25. Moreno-Montanes J, Anton A, Garcia N, et al. Glaucoma probability score vs Moorfields classification in normal, ocular hypertensive, and glaucomatous eyes. Am J Ophthalmol. 2008;145:360Y368.

Ophthalmoscope for Diagnosing Glaucoma

26. Toth M, Kothy P, Hollo G. Accuracy of scanning laser polarimetry, scanning laser tomography, and their combination in a glaucoma screening trial. J Glaucoma. 2008;17:639Y646. 27. Yip LW, Mikelberg FS. A comparison of the Glaucoma Probability Score to earlier Heidelberg Retina Tomograph data analysis tools in classifying normal and glaucoma patients. J Glaucoma. 2008;17:513Y516. 28. Ferreras A, Pablo LE, Larrosa JM, et al. Discriminating between normal and glaucoma-damaged eyes with the Heidelberg Retina Tomograph 3. Ophthalmology. 2008;115:775Y781. 29. Oddone F, Centofanti M, Iester M, et al. Sector-based analysis with the Heidelberg Retinal Tomograph 3 across disc sizes and glaucoma stages Ophthalmology. 2009;116:1106Y1111. 30. Saito H, Tomidokoro A, Yanagisawa M, et al. Sensitivity and specificity with the Glaucoma Probability Score in Heidelberg Retina Tomograph II in Japanese eyes. J Glaucoma. 2009;18:227Y232. 31. Saarela V, Falck A, Airaksinen PJ, et al. The sensitivity and specificity of Heidelberg Retina Tomograph parameters to glaucomatous progression in disc photographs. J Ophthalmol. 2010;94:68Y73. 32. Weinreb RN, Zangwill LM, Jain S, et al. Predicting the onset of glaucoma: the Confocal Scanning Laser Ophthalmoscopy Ancillary Study to the Ocular Hypertension Treatment Study. Ophthalmology. 2010;117:1674Y1683. 33. Healey PR, Lee AJ, Aung T, et al. Comparison of the diagnostic capability of the Heidelberg Retina Tomographs 2 and 3 for glaucoma in the Indian population. Ophthalmology. 2010;117:275Y281. 34. Rao HL, Babu GJ, Sekhar GC, et al. Diagnostic accuracy of the Heidelberg Retina Tomograph for glaucoma a population-based assessment. Ophthalmology. 2010;117:1667Y1673. 35. Zheng Y, Wong TY, Lamoureux E, et al. Diagnostic ability of Heidelberg Retina Tomography in detecting glaucoma in a population setting, the Singapore Malay Eye Study. Ophthalmology. 2010;117:290Y297. 36. Zelefsky JR, Harizman N, Mora R, et al. Assessment of a race-specific normative HRT-III database to differentiate glaucomatous from normal eyes. J Glaucoma. 2006;15:548Y551. 37. Mowatt Graham, et al. Screening tests for detecting open-angle glaucoma: systematic review and meta-analysis. Invest Ophthalmol Vis Sci. 2008;49:5373Y5385. 38. Ekstrom C. Elevated intraocular pressure and pseudoexfoliation of the lens capsule as risk factors for chronic open angle glaucomaVa population based 5 year follow up study. Acta Ophthalmol (Copenh). 1993;71:189Y195.

Try to be a rainbow in someone’s cloud. V Maya Angelon

* 2015 Asia Pacific Academy of Ophthalmology

www.apjo.org

Copyright © 2015 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

39