A Comparison of the Concordance of Digital Images, Live Examinations, and Skin Biopsies for the Diagnosis of Hospitalized Dermatology Consultation Patients

Noah Scheinfeld, MD, JD, Jeremy Kurz, MD, Ellen Teplitz, MD


Skinmed. 2003;2(1) 

In This Article


Our basic measure of utility of digital consultation was a comparison of digital versus live diagnosis by a physician. Agreement between digital and live diagnoses of the resident in this study was 81% (38 of a total of 47). A list of the diagnoses that differed between digital and live evaluations and biopsy results, when available, is included in Table I . Because the resident discussed cases with the supervising board certified dermatologist after examination, it was impossible to have the supervising physicians conduct blinded digital examinations, so such data was not acquired. Interestingly, when comparing live resident diagnosis to live attending diagnosis (when available), the diagnoses were in agreement in 90% of the cases (18 out of a total of 20). Thus, it is reasonable to view resident diagnoses as meaningful in this setting.

As a measure of relative accuracy of residents' digital diagnoses, digital diagnoses were compared to available live diagnoses from supervising dermatologists and found to be in agreement in 89% of cases (17 out of 19). Others have found that the concordance of dermatologists' diagnoses is high.[2] In order to assess true accuracy, clinical examination, live or digital, must be compared to the definitive diagnostic exam. This was done when available in the form of a skin biopsy or potassium hydroxide preparation. The resident's digital diagnoses correlated with the definitive diagnosis 76% of the time (22 of 29). A list of the diagnoses that differed between digital and definitive evaluations is included in Table II . Live attending physician diagnoses agreed with the available definitive diagnoses 73% of the time. Finally, out of 14 cases where digital, attending, and definitive diagnoses were available, 71% of the cases were diagnosed similarly with all diagnostic means.

The comparison of digital and live clinical diagnoses could not be rigorously compared because the resident who reviewed the digital images subsequently examined patients in person. The results were not wholly confounded because even a live examination did not definitively declare a diagnosis. Skin biopsy results were not available to the resident or attending physician at the time of digital or live diagnosis. All skin biopsies provided definite diagnoses.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: