A comparison of relative diagnostic precision between the Compass fundus perimeter and the Humphrey Field Analyzer

Abstract

Purpose : To evaluate relative diagnostic precision of two devices, the Compass (CMP, CenterVue, Italy) fundus perimeter and the Humphrey Field Analyzer (HFA, Zeiss, Dublin), in detecting glaucomatous optic neuropathy (GON) in a large multicentre cross-sectional study, .

Methods : One eye of 498 glaucoma patients and 436 age similar normals was tested with the index tests: HFA (SITA Standard strategy) and CMP (ZEST strategy) with a 24-2 grid. The reference test for GON was specialist evaluation of fundus photographs or OCT, independent of the visual field. For both devices, linear regression was used to calculate the sensitivity decrease with age in the normal group to compute pointwise Total Deviation (TD) values and Mean Deviation (MD). We derived 5% and 1% pointwise age corrected normative limits.
MD and the total number of TD values below the 5% (TD 5%) or the 1% (TD 1%) limit per field were used as classifiers. We used partial ROC curves and partial Area Under the Curve (pAUC) to compare the diagnostic precision of the devices.
Additionally, 44 glaucoma and 54 normal subjects were tested twice on both instruments. Pointwise Mean Absolute Deviation (MAD) and Bland Altman plots for the mean sensitivity (MS) were computed.
Results are reported as mean difference ± standard error.

Results : Both devices showed similar discriminative power (Figure 1). Differences between pAUC were negligible (for MD 0.019, p = 0.03; for TD 5% 0.012, p = 0.023; for TD 1% (0.003, p-value = 0.54).
The 95% limits of agreement for the MS were reduced by 14% in CMP compared to HFA in glaucoma subjects, and by 49% in normal subjects (Figure 2). MAD was very similar, being slightly smaller (not signifincant) in CMP compared to HFA for glaucoma (0.03 ± 0.2 dB) and for normals (0.08 ± 0.16 dB).
Average MS was lower with CMP than with HFA, both for glaucoma (-1.45 ± 0.01 dB) and normal subjects (-1.82 ± 0.11 dB).

Conclusions : Relative diagnostic precision of the two devices are equivalent. Test-retest variability of mean sensitivity for CMP was better than HFA. Fundus perimetry might be a promising tool to monitor progression glaucoma patients.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Figure 1. pROC curves for MD, TD 5% and TD 1% for HFA (red) and CMP (blue). The shaded regions represent the 95% CIs.

 

Figure 2. Bland – Altman plots for HFA (in red) and CMP (in blue) with 95% limits of agreement (shaded region) and mean difference (solid line).