Teledermatology for the Diagnosis and Management of Skin Cancer: A Systematic Review.

Importance: As technology becomes more commonplace in dermatological practice, it is essential to continuously review the accuracy of teledermatology devices and services compared with in-person care. The last systematic review was conducted over 5 years ago. Objective: To synthesize and assess the quality of the evidence to address 3 research questions: (1) How accurate is teledermatology for skin cancer diagnosis compared with usual care (face-to-face [FTF] diagnosis)? (2) Does teledermatology save clinician and/or patient time, compared with usual care? (3) What are the enablers and barriers to adoption of teledermatology in clinical practice for the diagnosis of skin cancer? Evidence Review: The review protocol was registered in the PROSPERO database. Six databases (Cochrane, PubMed, Medline, Science Direct, Embase, and Web of Science) were searched for studies investigating the diagnostic accuracy and concordance, management accuracy and concordance, measures of time (waiting times, delay to diagnosis), and enablers and barriers to implementation. Potentially eligible articles were screened by 2 reviewers. The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool was used to evaluate the risk of bias and applicability of individual studies assessing diagnostic accuracy. Findings: Twenty-one studies were reviewed. The diagnostic accuracy (defined as agreement with histopathology for excised lesions or clinical diagnosis for nonexcised lesions) of FTF dermatology consultation remains higher (67%-85% agreement with reference standard, Cohen κ, 0.90) when compared with teledermatology (51%-85% agreement with reference standard, κ, 0.41-0.63), for the diagnosis of skin cancer. However, some studies do report high accuracy of teledermatology diagnoses. Most studies of diagnostic accuracy and concordance had significant methodological limitations. Studies of health service outcomes found teledermatology reduced waiting times and could result in earlier assessment and treatment. Patients reported high satisfaction and were willing to pay out of pocket for access to such services. Conclusions and Relevance: Robust implementation studies of teledermatology are needed, paying careful attention to reducing risk of bias when assessing diagnostic accuracy. Teledermatology services consistently reduced waiting times to assessment and diagnosis, and patient satisfaction was high.