Effectiveness and diagnostic accuracy of teledermatology for the assessment of skin conditions
Nikesh Patel, Nada Aboukhatwah, Ben Esdaile
Abstract
BACKGROUND: Teledermatology provides a platform for swift specialist advice without the potential need for face-to-face review. Our objectives were to investigate the effectiveness, accuracy and diagnostic concordance of the platform with regard to the remote management of skin conditions. METHODS: We undertook a single-centre, retrospective chart review over a 1-year period, comprising a total of 1703 teledermatology referrals. Two physicians independently assessed the diagnostic concordance between telederm diagnosis (TD), in-person diagnosis (ID) and histopathological diagnosis (HD). RESULTS: There were a total of 1703 TD referrals, of which 341 were rejected, leaving 1362 referrals for evaluation. Sixty-five per cent of these referrals were managed remotely and discharged with advice, although 4.6% of these were later re-referred for an in-person review. A total of 20% of referrals were rejected, of which the majority was due to a lack of appropriate imaging. The total concordance of TD compared to ID was 76.4%. When comparing the TD and ID/HD, we obtained a Kappa value of 0.636 indicating substantial agreement. In terms of accuracy, there were 49 biopsy-proven skin cancers picked up by the service in this cohort of data. Of these, 61.2% were given an accurate diagnosis on first impression via teledermatology, 14.3% were given a different diagnosis but correctly categorised as skin cancer and 24.5% could not be assessed; however, they were triaged and escalated based upon clinical suspicion. CONCLUSION: Our study demonstrates that teledermatology is an effective platform in terms of diagnosis and remote management, with adequate diagnostic accuracy and concordance to in-person diagnosis.