Evaluating equivalence between the new Humphrey Field Analyzer (HFA3) and HFA II-i

Abstract

Purpose : A new generation of the Humphrey Field Analyzer (HFA) from ZEISS (Dublin, CA) has been introduced. Known as the HFA3, this device modernizes many of the original HFA II-i series components but preserves the bowl and projector relationship, as well as the normative reference databases.
This study investigated the agreement between the two devices for Mean Deviation (MD) and Pattern Standard Deviation (PSD) using the SITA Standard 24-2 threshold test. In addition, the inter-device, inter-operator variability and reproducibility of MD and PSD were determined.

Methods : A group of 33 normal subjects and 33 glaucoma subjects were tested once on three HFA3 perimeters and three HFA II-i perimeters over a period of two visits (six tests total). The 24-2 SITA Standard threshold tests were administered by three operators. Each operator was assigned to one HFA3 and one HFA 2i device.

Results : Results of the agreement analyses for MD and PSD are displayed in Table 1. All differences between the two devices were less than 0.4 dB. For determining the correlation between devices, both the normal and glaucoma groups were combined. HFA3 within-subject average versus the HFA II-i within-subject average produced R² values of 0.997 (MD) and 0.996 (PSD). The HFA3 individual versus the HFA II-i within subject average achieved R² values of 0.990 (MD) and 0.989 (PSD). For both conditions, the R² results exceeded the desired value of 0.95 for both MD and PSD. These R² findings confirmed the ability to utilize the same normative reference databases from the HFA II-i for the HFA3. Figure 1 illustrates the regression plot for MD.

Conclusions : The MD and PSD measurements made with the HFA3 were similar to measurements taken with the HFA II-i. With the same normative reference databases on the HFA3 and HFA II-i, the devices can share data and give equivalent results. Based on the findings of this study, the HFA3 is considered to be equivalent to the HFA II-i.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.

 

Table 1: Agreement of MD and PSD between HFA II-i and HFA3 (Glaucoma and Normal Eyes)

 

Figure 1: OLS Regression Plot for MD (glaucoma and normal eyes combined)