Comparison of 30-2 Visual Field Using Melbourne Rapid Fields Online Perimetry and Humphrey Field Analyzer.

Abstract

Prcis: Protocol 30-2 of Melbourne Rapid Fields, online computer perimetry, provides a portable, reliable, and patient-friendly alternative to Humphrey Field Analyzer 30-2 SITA fast protocol for Japanese all severity stages of glaucoma patients.

Purpose: Melbourne Rapid Fields (MRF) online computer perimetry is a web-browser-based software that offers white-on-white threshold perimetry using any computer. This study evaluates the perimetric results of 30-2 protocol from MRF performed using a laptop computer in comparison to Humphrey Field Analyzer (HFA).

Methods: A prospective and cross-sectional study of 87 eyes from 87 Japanese glaucoma patients. The MRF software includes features such as computer vision gaze monitoring and thresholding using Bayes logic. MRF's 30-2 VF results were compared to HFA 30-2 SITA-Fast, including Mean Deviation (MD), Pattern Deviation (PD), and reliability indices. Patients underwent 2 assessments on the MRF to establish test-retest reliability.

Results: Of the 87 eyes, 43 eyes had mild field defect (MD>-6 dB), 26 had moderate field defect (-12 dB≤MD≤-6 dB), and 18 had advanced field defects (MD<-12 dB). MRF demonstrated a high level of agreement with HFA in evaluating MD (Intraclass Correlation Coefficient ICC = 0.97 {95% CI 0.95 to 0.98}) and PSD (ICC=0.91 {95% CI 0.86 to 0.94}). Bland-Altman analysis revealed a mean bias of -0.76 decibels (dB) (95% Limits of Agreement LoA -5.82 dB, +4.30 dB) for MD and 0.79 dB (LoA -4.24 dB, +5.82 dB) for PSD. Regarding MRF test-retest, Bland-Altman analysis demonstrated a mean bias of 0.25 dB (LoA - 2.48 dB, +2.99 dB) for MD and -0.21 dB (LoA -3.22 dB, +2.79 dB) for PSD. Although false positives and fixation losses were comparable between MRF and HFA, the MRF showed slightly higher false negatives and longer test times than HFA, though these differences did not reach statistical significance. In the mild group, MRF has a sensitivity of detecting field defect of 80% and a specificity of 72%.

Conclusion: MRF provides a portable and accessible alternative to HFA for 30-2 visual field testing, with good agreement in moderate to advanced glaucoma. However, its slightly higher false negatives, longer test duration, and systemic difference in output to HFA should be considered when interpreting results. Further improvements may enhance its clinical utility.