Purchase this article with an account.
William G. Hodge, Zainab Khan, Kevin Warrian, Francie Si, Cindy Hutnik, Alex Mao, Irene Pan, David Moher, Ron Goeree, Fatemeh Yazdi; Summary of Glaucoma Testing Accuaracy: A Comprehensive Systematic Review of Clinical Studies. Invest. Ophthalmol. Vis. Sci. 2011;52(14):246. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
A plethora of studies now exist that evaluate imaging technologies for glaucoma. However there are many new studies weekly and assimilating this information into a useful synthesis is now overdue for both clinicians and policy makers. We undertook this task to assimilate the studies from 3 new technologies OCT, HRT and GDx compared to the gold standard of optic disc photography and white on white visual field.
Standard Information retrieval methods, inclusion and exclusion criteria, study selection methods, screening and data abstraction and analysis were performed as per the Agency for Health Care Research and Quality Systematic Review Protocols. 1035 articles were screened and after exclusions based on our a priori protocol, 303 atricles were used for data extraction. The QUORUM Statement was used to assess quality and a PRISMA flow diagram was generated. Inter-rater reliability was rated with the kappa statistic. Synthesized sensitivity, specificity and diagnostic odds ratios (DOR) were calculated vs the gold standard methods. Both point estimates and 95% confidence intervals were created. ROC curves were also aggregated whenever possible. A test for heterogeneity was performed-I2.
Both OCT (DOR 29.8 95% CI: 21.7,40.8) and GDx (DOR: 28.3, 95% CI: 21.6, 36.8) had similar accuracy compared to white on white perimetry but HRT lagged behind (DOR:20.9, 95% CI: 16.2, 27.2). Stratified data by new technology (Stratus OCT, OCT 2000, OCT1, GDx VCC, GDx FCC, HRT 1, 2 and 3) showed similar results to the overall results for each technology. Of importance, the I2 was well over 50% for all technologies and all categories.
The synthesized data favors GDx and OCT over HRT for diagnostic accuracy in glaucoma compared to a gold standard of white on white visual field. However all I2values are high indicating great heterogeneity in terms of chosen cutoffs and their values. Hence a more agreed upon uniform methodology to assess each technology consistently is needed and content experts in this field should work toward this goal.
This PDF is available to Subscribers Only