Abstract Background While endocytoscopy allows for in vivo cellular-level observation, achieving high-quality staining remains a technical challenge. Aims To develop and validate a simple, reliable scale for assessing and improving endocytoscopic staining quality. Methods We developed a four-dimension staining quality score and validated it on a retrospective image cohort (n = 2692). The scale’s reliability, validity, and internal consistency were assessed, and associated factors were identified using multinomial regression. The scale was then externally validated in a separate retrospective image cohort (n = 2413) to confirm its clinical applicability. Results The scale showed inter-rater agreement (ICC = 0.992) and reliability (Cronbach’s alpha = 0.81), with a single principal factor explaining 62.3% of variance.The Spearman correlation between the gold standard score and AOD value was 0.81. Staining quality was influenced by lesion characteristics and physician experience (P < 0.05). After one month of training, junior physicians showed improved agreement with gold standard scores, with Kappa values rising from 0.61, 0.66, and 0.69 to 0.81, 0.83, and 0.82. Higher staining quality was also significantly associated with better concordance between Endocytoscopy imaging and histopathology (P < 0.001). Conclusion This endocytoscopic staining quality score is a reliable and valid tool to guide technique refinement and improve diagnostic accuracy.
Fan et al. (Wed,) studied this question.