Purchase this article with an account.
K. Jahanbakhsh, R.R. A. Bourne, C. Boden, L.M. Zangwill, E.M. Hoffmann, F.A. Medeiros, R.N. Weinreb, P.A. Sample; Reproducibility of Selected Visual Field Endpoint Criteria Between SAP–FT and SAP–SITA Testing Strategies in the Diagnostic Innovations in Glaucoma Study (DIGS) . Invest. Ophthalmol. Vis. Sci. 2005;46(13):3724.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine agreement of commonly used visual field (VF) endpoint criteria in subjects who performed visual field tests using both Full Threshold (FT) and Swedish Interactive Testing Algorithm (SITA) testing strategies (inter–strategy agreement), and to compare with agreement between pairs of similar testing strategies (intra–strategy agreement).
Inter–strategy agreement was tested with a randomly selected eye of 173 subjects who had performed this pair of tests within a period of 3 months (Sequence FT&SITA). 37 subjects had stereophotographic evidence of glaucomatous optic neuropathy (GON) and an abnormal VF (abnormal Glaucoma Hemifield Test (GHT) &/or Pattern Standard Deviation (PSD)<5%)), 56 had GON only, 4 had only abnormal VF, and 76 had neither. Criteria tested: PSD<1%; PSD<5%; GHT; 4 pattern deviation locations <5% probability level (PDP). Intra–strategy agreement was tested for 44 (25.4%) subjects who had performed a FT within a year of the FT used in the inter–strategy pairing (FT&FT), and for 89 patients (51.4%) who had performed a SITA within a year before (SITA&SITA).
The levels of agreement (% and kappa) for endpoint criteria are given below. Agreement with GHT was significantly higher (P<0.01) for FT&FT than FT&SITA, and approached significance (P=0.07) when comparing FT&FT with SITA&SITA. There were no significant agreement differences between the sequences for other endpoints.
Inter– and intra–strategy agreement were not significantly different for the PSD and PDP criteria. However, the agreement of the GHT result between successive FT tests was significantly higher than between SITA and FT. This finding has implications when requiring an abnormal GHT result to be repeatably abnormal across different test strategies as a means of confirming abnormality.
This PDF is available to Subscribers Only