Comparison of 10-2 and 24-2 visual field event analyses

Abstract

Purpose : The 10-2 visual field (VF) provides more information in the center of the VF compared to 24-2. An extension of the HFA guided progression analysis (GPA) adding event (i.e. change from baseline) analysis for 10-2 was recently proposed¹. In this preliminary, retrospective study, we compare the results of the GPA event analyses in 24-2 and 10-2 VFs.

Methods : A retrospective series of VFs were analyzed from 22 eyes of 11 clinically managed glaucoma subjects with at least 5 visits each with 10-2 and 24-2 (or 30-2) SITA Standard VFs using HFA™ II-i or HFA3 (ZEISS, Dublin, CA). 10-2 and 24-2 VFs were typically alternated between visits. 24-2 VFs were extracted if the 30-2 VF was available. VFs were selected to cover the same follow-up period for the shorter of the 10-2 or 24-2 series. A GPA Alert result of “Possible” or “Likely” Progression was considered a progression event. Cohen’s kappa was used to assess agreement between the proposed 10-2 and reference 24-2 GPA event analyses.

Results : Age at first exam ranged from 32.2 to 74.1 years (see Table 1). Mean 24-2 and 10-2 MD at first exam were -1.16 (standard deviation, SD: 3.00, range: -12.47 to 1.43) dB and -0.88 (SD: 2.87, range: -12.33 to 1.08) dB, respectively. Mean follow-up time for the 24-2 series was 5.3 (SD: 2.8, range: 1.2 to 10.0) years, with a mean inter-visit interval of 0.4 (0.1, range: 0.3 to 0.6) years.Mean follow-up time for the 10-2 series was 5.2 (SD: 2.8, range: 1.4 to 9.9) years, with a mean inter-visit interval of 0.6 (0.3, range: 0.3 to 1.2) years. 21/22 cases (95.5%) were observed to agree, resulting in a kappa of 0.83 (95% CI: 0.51 to 1.00).

Conclusions : In this preliminary study, strong agreement was observed overall between the event analysis results for a proposed 10-2 GPA and a reference 24-2 GPA. This may suggest the potential clinical utility of using change from baseline analyses for 10-2 VFs to highlight progression in the central VF that may or may not be detected in the 24-2 VF. Further work may include more extensive comparisons between 10-2 and 24-2 GPA event analyses in cohorts with clinically confirmed diagnoses of progression or non-progression.

References
[1] Durbin et al. IOVS 2021; 62(8): Abstract 3482.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Table 1. Summary of demographics and GPA event analysis results.

 

Figure 1. Examples of 10-2 GPA event analysis alerts indicating progression (A) or no progression (B) at Nth Follow-up exam.