TY - JOUR
T1 - Diagnostic accuracy and interobserver concordance
T2 - teledermoscopy of 600 suspicious skin lesions in Southern Denmark
AU - Vestergaard, T.
AU - Prasad, S. C.
AU - Schuster, A.
AU - Laurinaviciene, R.
AU - Andersen, M. K.
AU - Bygum, A.
PY - 2020/7
Y1 - 2020/7
N2 - Background: Skin cancer incidences are increasing and early diagnosis, especially of malignant melanoma, is crucial. Teledermatology including teledermoscopy (TDS) can be used to triage referrals of suspicious skin lesions, however, this is not currently recommended in Denmark. Objectives: To evaluate diagnostic accuracy, sensitivity, specificity and interobserver concordance of TDS, and to evaluate the number of incidental lesions potentially missed by TDS. Methods: Fifty general practices were invited to send images of suspicious skin lesions for evaluation using smartphone TDS. Simultaneously, the patient was referred for a face-to-face (FTF) consultation. Images for TDS were independently evaluated by two dermatologists; a third dermatologist performed the FTF consultation. Diagnosis, management plan and level of diagnostic confidence were noted. For TDS photo quality was rated, and for FTF any incidental findings were described. Results: Six hundred lesions in 519 patients were included. The diagnostic accuracy was significantly higher on FTF evaluation than on TDS (P < 0.01). However, this was associated with a significant difference in specificity (P ≤ 0.012) whereas no significant difference was found in sensitivity. The concordance between FTF and TDS, and the interobserver concordance of two TDS evaluations was moderate to substantial (AC1 = 0.57–0.71). Incidental melanomas were found in 0.6% of patients on FTF evaluation, adding an extra 13% of melanomas. However, on TDS these patients’ photographed lesions all warranted FTF follow-up, where these melanomas would have been identified. Conclusion: In this large prospective study, no significant difference in sensitivity was observed between FTF and TDS, but specificity was lower on TDS than FTF. Taking management plans into account, we would, however, potentially have dismissed 2 of 23 melanomas, if only TDS had been used for assessment. One of these was a melanoma located on the scalp, an anatomic region less suitable for TDS.
AB - Background: Skin cancer incidences are increasing and early diagnosis, especially of malignant melanoma, is crucial. Teledermatology including teledermoscopy (TDS) can be used to triage referrals of suspicious skin lesions, however, this is not currently recommended in Denmark. Objectives: To evaluate diagnostic accuracy, sensitivity, specificity and interobserver concordance of TDS, and to evaluate the number of incidental lesions potentially missed by TDS. Methods: Fifty general practices were invited to send images of suspicious skin lesions for evaluation using smartphone TDS. Simultaneously, the patient was referred for a face-to-face (FTF) consultation. Images for TDS were independently evaluated by two dermatologists; a third dermatologist performed the FTF consultation. Diagnosis, management plan and level of diagnostic confidence were noted. For TDS photo quality was rated, and for FTF any incidental findings were described. Results: Six hundred lesions in 519 patients were included. The diagnostic accuracy was significantly higher on FTF evaluation than on TDS (P < 0.01). However, this was associated with a significant difference in specificity (P ≤ 0.012) whereas no significant difference was found in sensitivity. The concordance between FTF and TDS, and the interobserver concordance of two TDS evaluations was moderate to substantial (AC1 = 0.57–0.71). Incidental melanomas were found in 0.6% of patients on FTF evaluation, adding an extra 13% of melanomas. However, on TDS these patients’ photographed lesions all warranted FTF follow-up, where these melanomas would have been identified. Conclusion: In this large prospective study, no significant difference in sensitivity was observed between FTF and TDS, but specificity was lower on TDS than FTF. Taking management plans into account, we would, however, potentially have dismissed 2 of 23 melanomas, if only TDS had been used for assessment. One of these was a melanoma located on the scalp, an anatomic region less suitable for TDS.
U2 - 10.1111/jdv.16275
DO - 10.1111/jdv.16275
M3 - Journal article
C2 - 32031277
AN - SCOPUS:85082337138
SN - 0926-9959
VL - 34
SP - 1601
EP - 1608
JO - Journal of The European Academy of Dermatology and Venereology
JF - Journal of The European Academy of Dermatology and Venereology
IS - 7
ER -