Impact of scan size on foveal avascular zone metrics measured by optical coherence tomography angiography

Abstract Background Although scan size affects optical coherence tomography angiography (OCTA) measurements, it remains unclear whether it similarly affects multiple foveal avascular zone (FAZ)-related parameters across retinal vascular diseases. We therefore aimed to assess the effect of scan size on FAZ and associated metrics. Methods This retrospective observational study included 394 eyes, comprising 135 healthy eyes, 81 with diabetic retinopathy (DR), and 178 with branch retinal vein occlusion (BRVO). FAZ area, FAZ perimeter (PER), acircularity index (AI), and foveal density-300 (FD-300) were measured and compared between 3 × 3-mm and 6 × 6-mm OCTA scans acquired using the RTVue XR Avanti™. Differences between scan sizes were evaluated using the Wilcoxon signed-rank test. Correlation and agreement between the two scan protocols were assessed using Spearman’s rank correlation coefficients, intraclass correlation coefficients (ICCs), and Bland–Altman analysis. Scan size-related differences among disease groups were assessed using the Kruskal–Wallis test with Steel–Dwass post-hoc comparisons. Results FAZ area, FAZ PER, AI, and FD-300 differed significantly between the 3 × 3-mm and 6 × 6-mm scans in all disease groups ( p < 0.0001). ICC analysis showed good agreement for FAZ area (0.92–0.98) and FAZ PER (0.87–0.92), poor agreement for AI (0.23–0.49), and moderate-to-good agreement for FD-300 (0.65–0.78). Bland–Altman analysis identified systematic bias across all FAZ-related metrics. Scan size-related measurement differences also differed significantly among disease groups. Significant differences in FAZ area, FAZ PER, and AI were observed between the healthy eye group and both the DR and BRVO groups, whereas no significant difference in FD-300 was observed among the disease groups. Conclusion The 3 × 3-mm and 6 × 6-mm OCTA scan protocols showed differences in FAZ-related metrics, despite showing a degree of correlation between the two protocols. These findings suggest that scan size should be considered when interpreting FAZ-related metrics, particularly when different scan protocols are used in clinical practice and research. Trial registration Not applicable.