OBJECTIVES: This study compares the diagnostic performance of two rapid visual field testing strategies, SITA Faster (SF) and SPARK Quick (SQ), against the gold standard SITA Standard (SS). The measurement accuracy, agreement, and repeatability were evaluated. DESIGN: Prospective, cross-sectional study. PARTICIPANTS: A total of 118 eyes of 118 participants were enrolled. METHODS: Participants from a glaucoma clinic underwent multiple visual field tests. All subjects were naïve to SF and SQ, with SS 24-2 serving as the initial test. SF and SQ were then administered in random order. Among the 74 participants with visual field defects detected by SS, 49 underwent testing twice with both rapid strategies. Parameters were compared using one-way ANOVA with Bonferroni post hoc tests, and repeatability was assessed using intraclass correlation coefficients (ICC). A bootstrap method (5000 samples) evaluated the differences in repeatability. Agreement was analyzed using Bland-Altman quantile regression plots, and diagnostic performance was assessed via receiver operating characteristic (ROC) curves. MAIN OUTCOME MEASURES: Measurement accuracy, agreement, and repeatability. RESULTS: No significant differences in mean deviation (MD) and pattern standard deviation (PSD) were found between SS and SF; however, SQ demonstrated significantly better MD and PSD (p<0.001). Test duration was reduced by 57.8% for SF and 73.8% for SQ compared to SS (p<0.001). Bland-Altman analysis revealed that the mean differences in MD between SS and SF and between SS and SQ were -0.35 and -3.1, respectively. ICC values indicated excellent repeatability for SF, whereas SQ exhibited significantly lower ICCs (p<0.001). Among the 118 participants, 74 (62.7%) had visual field defects identified by SS, 72 (61.0%) by SF, and 31 (26.2%) by SQ. Sensitivity and specificity for SF were 83.8% and 77.3%, respectively, compared to 40.5% and 97.7% for SQ. ROC analysis yielded AUCs for MD of 0.89 for SF and 0.86 for SQ (p=0.306); for PSD, AUCs were 0.91 and 0.77 (p=0.005). The Conger's Kappa coefficient indicated fair agreement among the strategies. CONCLUSIONS: Both SF and SQ significantly shorten testing duration. SF yields results comparable to SS, while SQ may underestimate defects.
Wang et al. (Mon,) studied this question.