Purchase this article with an account.
Elham Ghahari, Christopher Bowd, Linda M Zangwill, James Proudfoot, Rafaella Penteado, Huiyuan Hou, Robert N Weinreb; Agreement between Compass Fundus Automated Perimetry and Humphrey Field Analyzer measurements.. Invest. Ophthalmol. Vis. Sci. 2019;60(9):2478.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate the agreement between threshold sensitivities and global indices from the Compass Fundus Automated Perimeter (CMP, CenterVue, Padova, Italy) and the Humphrey Field Analyzer (HFA, Zeiss, Dublin, CA).
Reliable CMP and HFA measurements from one or both eyes of 33 glaucoma suspects (by stereophotograph assessment or ocular hypertensive) or glaucoma patients (with consecutive repeatable VF defects) from the Diagnostic Innovations in Glaucoma Study were included [mean HVF MD (95% CI): -3.16 (-5.20, -1.12)]. CMP (ZEST algorithm 10-2, 24-2, and New Grid test patterns) and HFA (SITA Standard algorithm 10-2 and 24-2 test patterns) tests were performed in a random order on the same or consecutive days. Variance components of linear mixed models were used to compute the intra-class correlations (ICC) of corresponding visual field points’ raw threshold sensitivity and average MD and PSD between the two devices.
Mean (95% CI) visual field sensitivities were lower for CMP compared to HVF for 10-2 [26.67 (25.26, 28.07) dB and 30.33 (28.67, 31.99) dB, respectively], 24-2 [ 22.42 (20.48, 24.36) dB and 25.91 (23.71, 28.11) dB, respectively] and for New Grid [23.04 (21.40, 24.68) dB] and HVF 24-2 [26.95 (24.90, 29.00) dB] test patterns. ICCs comparing point-wise threshold sensitivities ranged from poor to excellent with 70.6% of tested points ≥ 0.60 for 10-2, 61.5% ≥ 0.60 for 24-2, and 50.0% ≥ 0.60 for comparing CMP New Grid and HVF 24-2.ICCs comparing MD and PSD for CMP and HVF for 10-2 were 0.93 (0.87, 0.97) and 0.95 (0.90, 0.97), respectively. For 24-2 were 0.87 (0.80, 0.92) and 0.84 (0.76, 0.90), respectively and for New Grid and 24-2 were 0.86 (0.75, 0.93) and 0.84 (0.75, 0.93), respectively.CPM MD was lower than HVF MD (from D MD = -0.78 dB for 10-2 to D MD = -1.58 dB between CMP New Grid and HFA 24-2, all comparisons p ≤ 0.002) and CPM PSD was higher than HVF PSD (from D PSD = 0.58 dB for 10-2 to D PSD = 0.97 dB between CMP New Grid and HFA 24-2, all comparisons p ≤ 0.004).
Threshold sensitivities are lower in CMP compared to HVF for the 3 test patterns investigated. Normative database-based MD and PSD between the Compass Fundus Perimeter and Humphrey Field Analyzer are in moderate to good agreement suggesting that these instruments, although not interchangeable, provide comparable assessment of disease severity.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only