Assessment of observer agreement in conventional and digital analysis of salivary gland biopsies in Sjögren’s disease

Objective To perform a comparison between conventional and digital methods of histopathologic analysis followed by observer reliability of positive biopsies for Sjögren’s disease (SjD). Materials and methods Histological characterization of 19 minor salivary glands (MSG) from lip biopsies of patients with a positive biopsy/histology for Sjögren's disease (meaning a Focus Score (FS) ≥ 1) was performed using the criteria of Fisher et al. [6]. The FS of all MSG biopsies were calculated and a conventional and a digital method were used to measure gland area. The Bland–Altman method was used to compare the conventional and digital strategies. Intra- and inter-observer reliability of both analyses was performed using the intraclass reliability coefficient (ICC) using the continuous variables number of foci and area of glandular parenchyma. Results The Bland–Altman method showed that there was no agreement between the conventional and digital methods. However, histologic analysis showed diagnostic reliability in terms of intra- and inter-observer levels for SjD-positive biopsy/histology in 19 lip MSG. The mean FS across observers for conventional analysis was 1.73 versus 2.67 for digital analysis, with an ICC of 0.775 and 0.866 for both analyses, respectively. The mean age of the patients was 39.6 years and all were female. Conclusion It is recommended to be aware that conventional and digital methods differ in terms of the final value of FS in patients with positive biopsy for SjD. Regarding mathematical calculation, FS tends to be higher with digital analysis than conventional analysis because the latter yields a greater value in measuring glandular area. However, both methods appear to be reliable when carefully standardized histologically. The prognostic implications of FS values obtained by the different methods should be carefully evaluated.