Automated imaging technologies for the diagnosis of glaucoma: a comparative diagnostic study for the evaluation of the diagnostic accuracy, performance as triage tests and cost-effectiveness (GATE study)

Abstract

BackgroundMany glaucoma referrals from the community to hospital eye services are unnecessary. Imaging technologies can potentially be useful to triage this population.ObjectivesTo assess the diagnostic performance and cost-effectiveness of imaging technologies as triage tests for identifying people with glaucoma.DesignWithin-patient comparative diagnostic accuracy study. Markov economic model comparing the cost-effectiveness of a triage test with usual care.SettingSecondary care.ParticipantsAdults referred from the community to hospital eye services for possible glaucoma.InterventionsHeidelberg Retinal Tomography (HRT), including two diagnostic algorithms, glaucoma probability score (HRT-GPS) and Moorfields regression analysis (HRT-MRA); scanning laser polarimetry [glaucoma diagnostics (GDx)]; and optical coherence tomography (OCT). The reference standard was clinical examination by a consultant ophthalmologist with glaucoma expertise including visual field testing and intraocular pressure (IOP) measurement.Main outcome measures(1) Diagnostic performance of imaging, using data from the eye with most severe disease. (2) Composite triage test performance (imaging test, IOP measurement and visual acuity measurement), using data from both eyes, in correctly identifying clinical management decisions, that is ‘discharge’ or ‘do not discharge’. Outcome measures were sensitivity, specificity and incremental cost per quality-adjusted life-year (QALY).ResultsData from 943 of 955 participants were included in the analysis. The average age was 60.5 years (standard deviation 13.8 years) and 51.1% were females. Glaucoma was diagnosed by the clinician in at least one eye in 16.8% of participants; 37.9% of participants were discharged after the first visit. Regarding diagnosing glaucoma, HRT-MRA had the highest sensitivity [87.0%, 95% confidence interval (CI) 80.2% to 92.1%] but the lowest specificity (63.9%, 95% CI 60.2% to 67.4%) and GDx had the lowest sensitivity (35.1%, 95% CI 27.0% to 43.8%) but the highest specificity (97.2%, 95% CI 95.6% to 98.3%). HRT-GPS had sensitivity of 81.5% (95% CI 73.9% to 87.6%) and specificity of 67.7% (95% CI 64.2% to 71.2%) and OCT had sensitivity of 76.9% (95% CI 69.2% to 83.4%) and specificity of 78.5% (95% CI 75.4% to 81.4%). Regarding triage accuracy, triage using HRT-GPS had the highest sensitivity (86.0%, 95% CI 82.8% to 88.7%) but the lowest specificity (39.1%, 95% CI 34.0% to 44.5%), GDx had the lowest sensitivity (64.7%, 95% CI 60.7% to 68.7%) but the highest specificity (53.6%, 95% CI 48.2% to 58.9%). Introducing a composite triage station into the referral pathway to identify appropriate referrals was cost-effective. All triage strategies resulted in a cost reduction compared with standard care (consultant-led diagnosis) but with an associated reduction in effectiveness. GDx was the least costly and least effective strategy. OCT and HRT-GPS were not cost-effective. Compared with GDx, the cost per QALY gained for HRT-MRA is £22,904. The cost per QALY gained with current practice is £156,985 compared with HRT-MRA. Large savings could be made by implementing HRT-MRA but some benefit to patients will be forgone. The results were sensitive to the triage costs.ConclusionsAutomated imaging can be effective to aid glaucoma diagnosis among individuals referred from the community to hospital eye services. A model of care using a triage composite test appears to be cost-effective.Future workThere are uncertainties about glaucoma progression under routine care and the cost of providing health care. The acceptability of implementing a triage test needs to be explored.FundingThe National Institute for Health Research Health Technology Assessment programme.

Review of economic evaluations of teleophthalmology as a screening strategy for chronic eye disease in adults

“…As technologies improve, it will be important to study methods of providing cost-effective solutions for the teleophthalmic management of visually impairing eye conditions including using machine learning approaches for image processing. A recent study evaluating the cost-effectiveness of automated imaging in glaucoma diagnosis found a composite triage test involving imaging test, IOP and visual acuity measurements to be cost-effective compared with clinical examination 43. Furthermore, all studies in our review (except for one) were conducted in developed countries in North America and Europe.…”

Diagnostic accuracy of optical coherence tomography for diagnosing glaucoma: secondary analyses of the GATE study

“…A complete description of the patients’ enrolment and flow in the GATE study has been previously published 11. A flow diagram of participants included in this study is shown in figure 1.…”

“…The GATE study11 found OCT sensitivity was 0.77 and specificity was 0.79 for diagnosing manifest glaucoma using the Spectralis RNFL classification of outside normal limits. Combination of two different imaging tests did not improve substantially the diagnostic accuracy.…”

“…Study design and participants have been described before 11. In brief, in the GATE study we conducted a pragmatic, multicentre, within-patient comparative evaluation of the diagnostic accuracy of automated imaging techniques for diagnosis of glaucoma 11…”

Effect of quantitative intraocular pressure reduction on visual field defect progression in normal tension glaucoma under medical therapy applying Markov model

“…The available methods for evaluating malignant disease transitions, such as courses and outcomes, include the Markov model 5. The Markov model is employed in health economics6 and disease treatment evaluations,7 as well as in the glaucoma field to determine gain and loss of treatment costs,8 perform automated diagnoses using imaging technologies,9 and assess imaging-based progression 10. However, few studies have reported on this model for clinical course evaluation in VFD 11.…”