Purchase this article with an account.
Michael West, Glen Sharpe, Donna Hutchison, Lesya M Shuba, Paul E Rafuse, Marcelo T Nicolela, Jayme R Vianna, Balwantray C Chauhan; Value of central 10-2 perimetry in patients with existing 24-2 visual field defects. Invest. Ophthalmol. Vis. Sci. 2019;60(9):2457. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent studies have suggested the 10-2 Humphrey visual field pattern identifies central visual field defects missed by the 24-2 pattern. We compared the diagnostic accuracy of these two test patterns in a cross-sectional analysis of early glaucoma patients and healthy controls followed in a longitudinal study on the value of macular imaging and perimetry.
We performed 24-2 and 10-2 perimetry (SITA standard, Humphrey Field Analyzer) in 62 glaucoma patients (62 eyes) and 36 healthy controls (68 eyes). Glaucoma patients had a positive Glaucoma Hemifield Test (GHT) and mean deviation ≥ -6 dB in the 24-2 test, whereas controls were required to have GHT within normal limits. Analysis included only the test points falling within 9° of fixation, encompassing the full 10-2 visual field and the central 12 points of the 24-2 visual field. Individual points with P < 5%, 2%, or 1% in the total deviation (TD) and pattern deviation (PD) values were analyzed. Receiver operating characteristic (ROC) curves for each test were compared with DeLong’s test, and the overlap of test positives identified by each test were compared at matched specificities.
The area under the ROC curve (AUC) at P < 5% was greater for the 10-2 pattern compared to the 24-2 pattern with both TD and PD values (Figure), however the differences were small and not statistically significant [difference (95% CI): TD, 0.05 (-0.01, 0.11) and PD, 0.05 (-0.01, 0.11)]. Similar results were obtained at P < 2% and 1% levels. Points along the 2 ROC curves matched for specificity showed high agreement in individuals identified by both test patterns, with each test pattern uniquely identifying a small subset of patients at all specificity values and probability levels.
This study indicates that in patients with existing 24-2 visual field defects and matched for specificity, the 10-2 test does not yield a meaningfully higher sensitivity or uniquely detect more cases than the 24-2 test.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
Figure. Receiver operating characteristics curves for 24-2 (blue) and 10-2 (red) patterns for total (A) and pattern (B) deviation analysis at P < 5%. Labelled points on the curves indicates the number of test points required for abnormality. Venn diagrams for 3 criteria that yielded the same specificity (number of points for each test pattern and corresponding sensitivity shown) indicating the overlap in the patients identified by each test pattern.
This PDF is available to Subscribers Only