Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma: secondary analyses of the GATE study

Gianni Virgili, Manuele Michelessi, Jonathan Cook, Charles Boachie, Jennifer Burr, Kathryn Ruth Banister, David Garway-Heath, Rupert R. A. Bourne, Almudena Asorey, Craig Robert Ramsay, Augusto Azuara-Blanco (Corresponding Author)

Research output: Contribution to journalArticle

2 Citations (Scopus)
8 Downloads (Pure)

Abstract

Background/Aims: To assess the diagnostic performance of retinal nerve fiber layer (RNFL) data of optical coherence tomography (OCT) for detecting glaucoma.

Methods: Secondary analyses of a prospective, multicenter diagnostic study (GATE) referred to hospital eye services in the United Kingdom. We included data from 899 of 966 participants, referred to hospital eye services with suspected glaucoma or ocular hypertension. We used both eyes’ data and logistic
regression-based Receiver Operator Characteristics (ROC) analysis to build a set of models to measure the sensitivity and specificity of the average and inferior quadrant RNFL thickness data of OCT. The reference standard was expert clinician examination including automated perimetry. The main outcome measures were sensitivity at 0.95 specificity and specificity at 0.95
sensitivity and the corresponding RNFL thickness thresholds. We explored the possibility of accuracy improvement by adding measures of within-eye and between-eye variation, scan quality, intraocular pressure, and age.


Results: Glaucoma was diagnosed in at least one eye in 17% of participants. Areas under the curve were between 0.83 and 0.88. When specificity was fixed at 0.95, the sensitivity was between 0.38 and 0.55, the highest values were reached with models including the inferior quadrant rather than the average RNFL thickness. Fixing specificity at 0.95 the sensitivity was between 0.36 and 0.58. The addition of age, refractive error, IOP, or within-subject variation did not improve the accuracy.

Conclusion: RNFL thickness data of OCT can be used as a diagnostic test but accuracy estimates remain moderate even in exploratory multivariable modelling of aiming to improve accuracy.
Original languageEnglish
Pages (from-to)604-610
Number of pages7
JournalBritish Journal of Ophthalmology
Volume102
Early online date30 Aug 2017
DOIs
Publication statusPublished - 2018

Fingerprint

Optical Coherence Tomography
Glaucoma
Nerve Fibers
Ocular Hypertension
Visual Field Tests
Refractive Errors
Intraocular Pressure
Routine Diagnostic Tests
Multicenter Studies
Area Under Curve
Outcome Assessment (Health Care)
Sensitivity and Specificity

Keywords

  • Glaucoma
  • Imaging
  • Diagnostic tests/Investigation

Cite this

Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma : secondary analyses of the GATE study. / Virgili, Gianni ; Michelessi, Manuele ; Cook, Jonathan; Boachie, Charles; Burr, Jennifer; Banister, Kathryn Ruth; Garway-Heath, David; Bourne, Rupert R. A.; Asorey, Almudena; Ramsay, Craig Robert; Azuara-Blanco, Augusto (Corresponding Author).

In: British Journal of Ophthalmology, Vol. 102, 2018, p. 604-610.

Research output: Contribution to journalArticle

Virgili, G, Michelessi, M, Cook, J, Boachie, C, Burr, J, Banister, KR, Garway-Heath, D, Bourne, RRA, Asorey, A, Ramsay, CR & Azuara-Blanco, A 2018, 'Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma: secondary analyses of the GATE study', British Journal of Ophthalmology, vol. 102, pp. 604-610. https://doi.org/10.1136/bjophthalmol-2017-310642
Virgili, Gianni ; Michelessi, Manuele ; Cook, Jonathan ; Boachie, Charles ; Burr, Jennifer ; Banister, Kathryn Ruth ; Garway-Heath, David ; Bourne, Rupert R. A. ; Asorey, Almudena ; Ramsay, Craig Robert ; Azuara-Blanco, Augusto. / Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma : secondary analyses of the GATE study. In: British Journal of Ophthalmology. 2018 ; Vol. 102. pp. 604-610.
@article{e40067fb85ac41759609921d2f9aa7d2,
title = "Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma: secondary analyses of the GATE study",
abstract = "Background/Aims: To assess the diagnostic performance of retinal nerve fiber layer (RNFL) data of optical coherence tomography (OCT) for detecting glaucoma.Methods: Secondary analyses of a prospective, multicenter diagnostic study (GATE) referred to hospital eye services in the United Kingdom. We included data from 899 of 966 participants, referred to hospital eye services with suspected glaucoma or ocular hypertension. We used both eyes’ data and logisticregression-based Receiver Operator Characteristics (ROC) analysis to build a set of models to measure the sensitivity and specificity of the average and inferior quadrant RNFL thickness data of OCT. The reference standard was expert clinician examination including automated perimetry. The main outcome measures were sensitivity at 0.95 specificity and specificity at 0.95sensitivity and the corresponding RNFL thickness thresholds. We explored the possibility of accuracy improvement by adding measures of within-eye and between-eye variation, scan quality, intraocular pressure, and age.Results: Glaucoma was diagnosed in at least one eye in 17{\%} of participants. Areas under the curve were between 0.83 and 0.88. When specificity was fixed at 0.95, the sensitivity was between 0.38 and 0.55, the highest values were reached with models including the inferior quadrant rather than the average RNFL thickness. Fixing specificity at 0.95 the sensitivity was between 0.36 and 0.58. The addition of age, refractive error, IOP, or within-subject variation did not improve the accuracy.Conclusion: RNFL thickness data of OCT can be used as a diagnostic test but accuracy estimates remain moderate even in exploratory multivariable modelling of aiming to improve accuracy.",
keywords = "Glaucoma, Imaging, Diagnostic tests/Investigation",
author = "Gianni Virgili and Manuele Michelessi and Jonathan Cook and Charles Boachie and Jennifer Burr and Banister, {Kathryn Ruth} and David Garway-Heath and Bourne, {Rupert R. A.} and Almudena Asorey and Ramsay, {Craig Robert} and Augusto Azuara-Blanco",
note = "Financial Support: - The GATE study was funded by the National Institute of Health Research HTA programme (09/22/111). The views and opinions expressed are those of the authors and do not necessarily reflect those of the HTA programme, NIHR, NHS or the Department of Health. - The contribution of Manuele Michelessi of the IRCCS Fondazione Bietti in this paper was supported by the Italian Ministry of Health and by Fondazione Roma",
year = "2018",
doi = "10.1136/bjophthalmol-2017-310642",
language = "English",
volume = "102",
pages = "604--610",
journal = "British Journal of Ophthalmology",
issn = "0007-1161",
publisher = "BMJ Publishing Group",

}

TY - JOUR

T1 - Diagnostic accuracy of Optical Coherence Tomography for diagnosing glaucoma

T2 - secondary analyses of the GATE study

AU - Virgili, Gianni

AU - Michelessi, Manuele

AU - Cook, Jonathan

AU - Boachie, Charles

AU - Burr, Jennifer

AU - Banister, Kathryn Ruth

AU - Garway-Heath, David

AU - Bourne, Rupert R. A.

AU - Asorey, Almudena

AU - Ramsay, Craig Robert

AU - Azuara-Blanco, Augusto

N1 - Financial Support: - The GATE study was funded by the National Institute of Health Research HTA programme (09/22/111). The views and opinions expressed are those of the authors and do not necessarily reflect those of the HTA programme, NIHR, NHS or the Department of Health. - The contribution of Manuele Michelessi of the IRCCS Fondazione Bietti in this paper was supported by the Italian Ministry of Health and by Fondazione Roma

PY - 2018

Y1 - 2018

N2 - Background/Aims: To assess the diagnostic performance of retinal nerve fiber layer (RNFL) data of optical coherence tomography (OCT) for detecting glaucoma.Methods: Secondary analyses of a prospective, multicenter diagnostic study (GATE) referred to hospital eye services in the United Kingdom. We included data from 899 of 966 participants, referred to hospital eye services with suspected glaucoma or ocular hypertension. We used both eyes’ data and logisticregression-based Receiver Operator Characteristics (ROC) analysis to build a set of models to measure the sensitivity and specificity of the average and inferior quadrant RNFL thickness data of OCT. The reference standard was expert clinician examination including automated perimetry. The main outcome measures were sensitivity at 0.95 specificity and specificity at 0.95sensitivity and the corresponding RNFL thickness thresholds. We explored the possibility of accuracy improvement by adding measures of within-eye and between-eye variation, scan quality, intraocular pressure, and age.Results: Glaucoma was diagnosed in at least one eye in 17% of participants. Areas under the curve were between 0.83 and 0.88. When specificity was fixed at 0.95, the sensitivity was between 0.38 and 0.55, the highest values were reached with models including the inferior quadrant rather than the average RNFL thickness. Fixing specificity at 0.95 the sensitivity was between 0.36 and 0.58. The addition of age, refractive error, IOP, or within-subject variation did not improve the accuracy.Conclusion: RNFL thickness data of OCT can be used as a diagnostic test but accuracy estimates remain moderate even in exploratory multivariable modelling of aiming to improve accuracy.

AB - Background/Aims: To assess the diagnostic performance of retinal nerve fiber layer (RNFL) data of optical coherence tomography (OCT) for detecting glaucoma.Methods: Secondary analyses of a prospective, multicenter diagnostic study (GATE) referred to hospital eye services in the United Kingdom. We included data from 899 of 966 participants, referred to hospital eye services with suspected glaucoma or ocular hypertension. We used both eyes’ data and logisticregression-based Receiver Operator Characteristics (ROC) analysis to build a set of models to measure the sensitivity and specificity of the average and inferior quadrant RNFL thickness data of OCT. The reference standard was expert clinician examination including automated perimetry. The main outcome measures were sensitivity at 0.95 specificity and specificity at 0.95sensitivity and the corresponding RNFL thickness thresholds. We explored the possibility of accuracy improvement by adding measures of within-eye and between-eye variation, scan quality, intraocular pressure, and age.Results: Glaucoma was diagnosed in at least one eye in 17% of participants. Areas under the curve were between 0.83 and 0.88. When specificity was fixed at 0.95, the sensitivity was between 0.38 and 0.55, the highest values were reached with models including the inferior quadrant rather than the average RNFL thickness. Fixing specificity at 0.95 the sensitivity was between 0.36 and 0.58. The addition of age, refractive error, IOP, or within-subject variation did not improve the accuracy.Conclusion: RNFL thickness data of OCT can be used as a diagnostic test but accuracy estimates remain moderate even in exploratory multivariable modelling of aiming to improve accuracy.

KW - Glaucoma

KW - Imaging

KW - Diagnostic tests/Investigation

U2 - 10.1136/bjophthalmol-2017-310642

DO - 10.1136/bjophthalmol-2017-310642

M3 - Article

VL - 102

SP - 604

EP - 610

JO - British Journal of Ophthalmology

JF - British Journal of Ophthalmology

SN - 0007-1161

ER -