Friday, September 22, 2023
| Study | Description | Major exclusions | Participants | Sensitivity/Specificity |
|---|---|---|---|---|
| Bagchi et al. 2019[26] | Retrospective audit of patients who presented to a retinal clinic in the United Kingdom with high myopia (<−6D or AL > 26 mm) and new onset visual disturbance who received FA, OCTA and SD-OCT imaging. | Excluded patients who did not receive all three imaging modalities. Excluded patients with poor quality images. Excluded patients with other co- existing major ocular conditions. | 27 eyes of 26 patients (18 female, 6 male) Mean age 47.7 ± 19.7 years | OCTA vs FA: Sensitivity 19/23, specificity 3/4 SD-OCT vs FA: Sensitivity 23/23, specificity 0/4 |
| Milani et al.[24] | Retrospective audit of patients seen at a research hospital in Italy with recent vision deterioration, pathologic myopia (<−6D and staphyloma) and suspected mCNV who received near infrared, autofluorescence, FA and SD-OCT imaging at first presentation. | Excluded patients who did not receive all four imaging modalities. Excluded patients with poor quality images. Excluded patients who had previous vitreoretinal surgery, diabetes, signs of age-related macular degeneration, or vitreoretinal interface-related pathologies. | 65 eyes of 62 patients (44 female, 21 male) Mean age 66.72 years, range 18–89 Mean refraction −9.72D, range −6 to −22 | SD-OCT vs FA: Sensitivity 48/49, specificity 16/16 |
| Miyata et al. 2016[25] | Prospective study of consecutive patients who presented to a university ophthalmology clinic in Japan with pathologic myopia (<−6D or AL > 26 mm, plus chorioretinal abnormalities) and treatment naïve exudative lesions. | Patients with OCTA images of insufficient quality were excluded from analysis. | 28 eyes of 26 patients (22 female, 4 male) Included in analysis: 21 eyes of 20 patients (17 female, 3 male) Mean age 63.0 ± 13.6 years | OCTA vs FA: Sensitivity 16/17, specificity 4/4 |
| Querques et al. 2017[28] | Retrospective audit of patients who presented to a university hospital’s retinal clinic in Italy with pathologic myopia (<−8D or AL > 26.5 mm, plus characteristic degenerative changes of the sclera/ choroid/retina) who were diagnosed with mCNV using FA.a An additional cohort of patients with pathologic myopia and no evidence of mCNV were enrolled as a negative control group. | Excluded patients with co-existing retinal conditions, history of ocular inflammation in the study eye, significant media opacities, or large haemorrhage. Patients with OCTA images of insufficient quality or who did not have FA performed on the same day as OCTA were excluded from analysis. Negative control group: Excluded patients with co- existing retinal conditions, or previous ocular treatments in the study eye. | 36 eyes of 28 patients (23 female, 5 male) Included in analysis: 21 eyes of 17 patients (14 female, 3 male) Mean age 57.8 ± 14.5 years= Negative control group: 32 eyes of 32 patients (27 female, 5 male) Mean age 56.2 ± 14.4 years, range 26–84 | OCTA vs FAa: Sensitivity 19/21, specificity 30/32 |
| Su et al. 2014[27] | Prospective study of patients who presented to a macular service centre in China with high myopia (< −6D and AL > 26.5 mm) and myopic maculopathy. | Excluded patients with other retinal or choroidal diseases, or dense cataracts. | 69 eyes of 42 patients (23 female, 19 male) Mean age 47.3 ± 17.3 years, range 20–79 | SD-OCT vs FA: Sensitivity 16/16, specificity 53/53 |
Table 1. Key details of included studies.
D Dioptres, AL Axial length, FA Fluorescein angiography, OCTA Optical coherence tomography angiography, SD-OCT Spectral domain optical coherence tomography, mCNV Myopic choroidal neovascularisation.
aUse of FA as reference standard was clarified by direct communication with the corresponding author.
| Outcome (95% CI) | Outcome (95% CI) | |||||
|---|---|---|---|---|---|---|
| OCTA compared to FA | ||||||
| Individual studiesa | TP | FP | FN | TN | Sensitivity | Specificity |
| Bagchi 2019 | 19 | 1 | 4 | 3 | 0.83 (0.61–0.95) | 0.75 (0.19–0.99) |
| Miyata 2016 | 16 | 0 | 1 | 4 | 0.94 (0.71–1.00) | 1.00 (0.40–1.00) |
| Querques 2017 | 19 | 2 | 2 | 30 | 0.90 (0.70–0.99) | 0.94 (0.79–0.99) |
| Pooled results from meta-analysisb | Sensitivity | Specificity | ||||
| 0.89 (0.78–0.94) | 0.93 (0.79–0.98) | |||||
| LR of a positive test | LR of a negative test | |||||
| 11.8 (3.96–35.25) | 0.12 (0.061–0.25) | |||||
| Positive PV | Negative PV | |||||
| 0.95 (0.79–0.99) | 0.85 (0.61–0.94) | |||||
| SD-OCT compared to FA | ||||||
| Individual studiesa | TP | FP | FN | TN | Sensitivity | Specificity |
| Bagchi 2019 | 23 | 4 | 0 | 0 | 1.00 (0.85–1.00) | 0.00 (0.00–0.60) |
| Milani 2016 | 48 | 0 | 1 | 16 | 0.98 (0.89–1.00) | 1.00 (0.79–1.00) |
| Su 2014 | 16 | 0 | 0 | 53 | 1.00 (0.79–1.00) | 1.00 (0.93–1.00) |
| Pooled results from meta-analysisb | Sensitivity | Specificity | ||||
| 0.99 (0.91–1.00) | unestimatable | |||||
| LR of a positive test | LR of a negative test | |||||
| unestimatable | 0.01 (0.001–0.095) | |||||
| Positive PV | Negative PV | |||||
| unestimatable | unestimatable | |||||
Table 2. Test accuracy of individual studies and pooled results from meta-analysis.
CI confidence interval, OCTA optical coherence tomography angiography, FA fluorescein angiography, TP true positive, FP false positive, FN false negative, TN true negative, LR likelihood ratio, PV predictive value, SD-OCT spectral domain optical coherence tomography.
aCalculated using RevMan Ver 5.4.1.
bCalculated using SAS macro MetaDAS v1.3.
|
Test |
Recommendation |
|---|---|
|
OCTA |
Conditionally recommend the use of OCTA to achieve a diagnosis when mCNV is clinically suspected. Statement was conditional because all studies excluded patients from analysis due to image quality issues.
|
|
SD-OCT |
Conditionally suggest clinicians may consider the use of SD-OCT to achieve a diagnosis when mCNV is clinically suspected. Statement was conditional because of the inability to estimate a pooled specificity for SD-OCT resulting in an unknown false positive rate.
|
|
OCTA + SD-OCT |
Clinicians may consider using SD-OCT if an OCTA image of sufficient quality cannot be acquired.
|
Table 3. Key recommendations.
OCTA Optical coherence tomography angiography, mCNV Myopic choroidal neovascularisation, FA Fluorescein angiography, SD-OCT Spectral domain optical coherence tomography.
Physicians - maximum of 1.00 AMA PRA Category 1 Credit(s)™
This activity is intended for ophthalmologists and other clinicians caring for patients with myopic choroidal neovascularization (mCNV).
The goal of this activity is for learners to be better able to describe the test accuracy of optical coherence tomography angiography (OCTA) and spectral domain (SD)-OCT in diagnosing mCNV compared with the reference standard, fluorescein angiography, according to a meta-analysis with primary outcome of pooled sensitivity and specificity and additional outcomes of pooled likelihood ratios and patient-oriented outcomes.
Upon completion of this activity, participants will:
Medscape, LLC requires every individual in a position to control educational content to disclose all financial relationships with ineligible companies that have occurred within the past 24 months. Ineligible companies are organizations whose primary business is producing, marketing, selling, re-selling, or distributing healthcare products used by or on patients.
All relevant financial relationships for anyone with the ability to control the content of this educational activity are listed below and have been mitigated. Others involved in the planning of this activity have no relevant financial relationships.
Medscape, LLC designates this Journal-based CME activity for a maximum of 1.0 AMA PRA Category 1 Credit(s)™ . Physicians should claim only the credit commensurate with the extent of their participation in the activity.
For questions regarding the content of this activity, contact the accredited provider for this CME/CE activity noted above. For technical assistance, contact [email protected]
There are no fees for participating in or receiving credit for this online educational activity. For information on applicability
and acceptance of continuing education credit for this activity, please consult your professional licensing board.
This activity is designed to be completed within the time designated on the title page; physicians should claim only those
credits that reflect the time actually spent in the activity. To successfully earn credit, participants must complete the
activity online during the valid credit period that is noted on the title page. To receive AMA PRA Category 1 Credit(s)™, you must receive a minimum score of 75% on the post-test.
Follow these steps to earn CME/CE credit*:
You may now view or print the certificate from your CME/CE Tracker. You may print the certificate, but you cannot alter it.
Credits will be tallied in your CME/CE Tracker and archived for 6 years; at any point within this time period, you can print
out the tally as well as the certificates from the CME/CE Tracker.
*The credit that you receive is based on your user profile.
processing....
This systematic review and meta-analysis was conducted and reported in compliance with the Preferred Reporting Items for Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies (PRISMA-DTA)[10] and the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) guidelines[11, 12]. The study protocol was published online prior to commencing the literature search[13]. Study selection, data extraction and risk of bias appraisal were undertaken by at least two authors (SH, GSD). Disagreements were resolved by obtaining the opinion of a third independent author (AL) and majority decisions prevailed.
Data sources
MEDLINE (www.PubMed.org) and EMBASE (www.Embase.com) were searched from inception until 30 March 2021, without language restrictions. The query combined appropriate database specific statements using Medical Subject Heading (MeSH) or Emtree terms, and filters for studies of diagnostic accuracy were applied[14–16]. Reference lists of retrieved papers were also hand-searched to identify additional studies. Complete search strategy details are reported in (Supplement eTable 1).
Study selection
We included all studies evaluating the test accuracy of OCTA and/or OCT against the reference standard FA in diagnosing mCNV. Any combination of OCTA and/or OCT test device assessed against FA was included. Exclusion criteria were case reports, review articles, non-human studies and any article type where primary data to calculate sensitivity and specificity against the reference standard was not completely reported. Endnote software (Clarivate Analytics) was used to manage references. After removal of duplicate studies, articles were screened by title and abstract to identify studies that needed to be retrieved in full text for detailed assessment of eligibility.
Data extraction and risk of bias assessment
Data regarding study design, population, index test and reference test procedures, and outcomes were extracted from eligible studies. All included studies were appraised for risk of bias using the following domains from the revised Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool:[17] (1) was consecutive or random sampling used to obtain the patient sample; (2) was a case-control design avoided; (3) were there inappropriate exclusions; (4) were index test results interpreted without knowledge of the reference standard results; (5) were reference standard results interpreted without knowledge of the index test results; (6) was there an appropriate interval between the index test and reference standard; and (7) were all patients included in the analysis. Studies were rated ‘high’, ‘low’ or ‘unclear’ for risk of bias, and results graphed using a ‘traffic light system’ as proposed in the GRADE guidelines[11]. The ‘unclear’ category was used only when insufficient data was reported to permit a judgement.
Data synthesis and analysis
Clinical recommendations. The GRADE guidelines for assessing certainty of the evidence with respect to study design, risk of bias, indirectness, inconsistency, imprecision and publication bias were used to frame clinical recommendations[11, 12]. The GRADE assessment framework is supported by published reporting guidelines including PRISMA-DTA and Cochrane[10, 18–20].
Outcomes
The primary outcome was test accuracy, calculated as the sensitivity and specificity of OCTA or OCT against FA at the initial patient presentation. Downstream consequences of care (management decisions, health outcomes, resource utilisation)[21] delivered using OCTA or OCT alone vs FA were investigated as secondary outcomes.
Statistical analysis
Given the limitations of all available statistical models and methods to test for publication bias in test accuracy studies, and lack of a standardised method to register test accuracy studies[12], we intended to assess publication bias using Deeks’ test[22] only if 10 or more studies were identified for inclusion.
Pooled estimates with 95% confidence intervals (95% CI) of sensitivity and specificity for each index test (OCTA and OCT) were obtained using a bivariate model. Estimates of the positive likelihood ratio, negative likelihood ratio, and positive and negative predictive values were back-calculated from the pooled estimates of sensitivity and specificity. Study differences in patient populations, patient selection, risk of bias, clinical setting, disease severity, scan density, scan quality and retinal thickness were considered as sources of heterogeneity.
All statistical analyses were conducted using RevMan 5.4.1 (The Cochrane Collaboratio®, Oxford, England, 2020), and the SAS macro MetaDAS v1.3[23].
;)
EnlargeFigure 1. Flow diagram of the study selection process. N Number, mCNV Myopic choroidal neovascularisation, FA Fluorescein angiography, Sens Sensitivity, Spec Specificity.
