A static-image telepathology system for dermatopathology consultation in East Africa: The Massachusetts General Hospital Experience

2012 
Background The histologic diagnosis of skin lesions in the developing world is complicated by the shortage of pathologists with subspecialty training in dermatopathology, limited access to ancillary diagnostic testing, and costly referrals for expert glass slide consultation in challenging cases. Objective In this study we evaluate the feasibility of a static-image telepathology platform in Africa for performing accurate dermatopathology consultations. Methods A static-image telepathology platform using the iPath server was utilized by referring pathologists in 4 African hospitals. Diagnostic interpretations were provided by Massachusetts General Hospital dermatopathologists at no cost. The diagnostic accuracy and interobserver correlation was evaluated. Results The static histopathologic images were diagnostic in 22 of 29 (76%) cases. Diagnostic accuracy between static image and glass slide diagnosis in 22 cases was 91%, ranging from 86% to 95% according to years of dermatopathology subspecialty expertise. Comparison with the glass slides showed that the telepathology diagnosis was limited by inappropriate field selection in only one case. Interobserver concordance between two pathologists was high (K = 0.86) suggesting that this platform is easy to use with minimal training of both referring and consulting pathologists. Limitations Concordance between conventional microscopy and static image telepathology was performed in 22 of 29 cases for which glass slides were received. Interobserver concordance was performed for two pathologists. Conclusion Static-image telepathology is a feasible means of rendering diagnoses on dermatopathology cases and is a cost-effective technology for obtaining much-needed second opinions in resource-poor settings.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    30
    References
    36
    Citations
    NaN
    KQI
    []