Diagnostic accuracy in virtual dermatopathology

Abstract

Virtual microscopy is used for teaching medical students and residents and for in-training and certification examinations in the United States. However, no existing studies compare diagnostic accuracy using virtual slides and photomicrographs. The objective of this study was to compare diagnostic accuracy of dermatopathologists and pathologists using photomicrographs vs. digitized images, through a self-assessment examination, and to elucidate assessment of virtual dermatopathology. Forty-five dermatopathologists and pathologists received a randomized combination of 15 virtual slides and photomicrographs with corresponding clinical photographs and information in a self-assessment examination format. Descriptive data analysis and comparison of groups were performed using a chi-square test. Diagnostic accuracy in dermatopathology using virtual dermatopathology or photomicrographs was similar: 0.70 vs. 0.73 respectively, p = 0.465. Order of administration of virtual dermatopathology and photomicrographs did not affect diagnostic accuracy. The vast majority (93%) of the participants felt the virtual slides were adequate for diagnosis and that virtual dermatopathology represented a useful tool for learning; 90% felt that virtual dermatopathology is useful tool for teaching dermatopathology. No significant difference was observed in diagnostic accuracy using virtual dermatopathology compared to photomicrographs. Most participants felt virtual dermatopathology was adequate for diagnostic purposes and a useful tool for learning and teaching dermatopathology.