Feasibility study in teledermatopathology: An examination of the histopathologic features of mycosis fungoides and spongiotic dermatitis

• Published in: Journal of cutaneous pathology 2017-11, Vol.44 (11), p.919-924
• Main Authors: Fertig, Raymond M, Gaudi, Sudeep, Cervantes, Jessica, Maddy, Austin, Sangueza, Omar, Vu, John, Ho, Jonhan, Jukic, Drazen M
• Format: Article
• Language: English
• Subjects: Dermatitis; Dermatitis - diagnosis; Dermatology - methods; Diagnosis, Differential; digital; Discordance; Feasibility Studies; Fungal infections; histopathology; Humans; Mycosis; Mycosis fungoides; Mycosis Fungoides - diagnosis; Observer Variation; Pathology, Clinical - methods; Skin Neoplasms - diagnosis; spongiotic dermatitis; Statistical analysis; teledermatopathology; Telemedicine - methods
• ISSN: 0303-6987; 1600-0560
• DOI: 10.1111/cup.13018

Background Digital pathology offers numerous advantages, allowing remote information sharing using whole slide imaging (WSI) to digitize an entire glass slide (GS) at high resolution, creating a digital slide (DS). Methods In this study, we examine the concordance in diagnoses made on 40 digital slides (DSs) vs traditional GSs in differentiating between spongiotic dermatitis (SD) and patch/plaque‐stage mycosis fungoides (MF). Results Greater interobserver concordance rate in final diagnosis of SD vs MF was observed with the utilization of DS (86.7%) compared with the utilization of GS (80%). Intraobserver concordance rate between the diagnoses rendered by a particular dermatopathologist on GS and DS was 86.7%. For all histopathological criteria, a correlation in the magnitudes of interobserver vs intraobserver discordances suggests that discordance between glass vs digital evaluation of these criteria may be largely expected subjective read variation independent of the media. Discordance in identification of histopathological features did not have a statistically significant link to discordance in diagnosis for 7 out of the 8 features. Conclusions The similarity between interobserver and intraobserver discordances suggests that WSI does not introduce additional barriers or variability to accurately identify histopathologic feature and to discriminate between MF and SD beyond interobserver variability.