Positive Predictive Value of Melanoma Diagnosis in Store-and-Forward Teledermatology.

ABSTRACT

Background: Although store-and-forward teledermatology (SFT) has demonstrated good diagnostic sensitivity for melanoma, little is known about the diagnostic precision (positive predictive value [PPV]). Introduction: We conducted this investigation to ascertain the PPV of melanoma diagnosis among teledermatology readers. Materials and Methods: We reviewed teledermatology consultations on 8,706 patients completed during the period February 1, 2015-January 31, 2016. Melanoma was included in the differential diagnosis of 551 conditions. We conducted a chart review of each condition to determine the final diagnosis. Results: We ascertained a final diagnosis in 503 conditions. Sixty-nine conditions were ultimately diagnosed as melanoma, a PPV of 13.7%. There was considerable variability in PPV among readers. Image quality was associated with higher PPV. Discussion: Overall, SFT program PPV compared favorably with that found in two published studies of face-to-face dermatology clinic care. Conclusion: To increase the diagnostic precision of SFT program melanoma diagnosis, efforts should be directed toward improving selected individual reader's PPV and image quality.