Internet-based skin cancer screening using clinical images alone or in conjunction with dermoscopic images: A randomized teledermoscopy trial.

Abstract

Teledermoscopy involves the use of dermoscopic images for remote consultation and decision-making in skin cancer screening. We sought to analyze the potential benefits gained from the addition of dermoscopic images to an internet-based skin cancer screening system. A randomized clinical trial assessed the diagnostic performance and cost-effectiveness of clinical teleconsultations (CTC) and clinical with dermoscopic teleconsultations. A total of 454 patients were enrolled in the trial (n_CTC = 226, n_{clinical with dermoscopic teleconsultation} = 228). Teledermoscopy improved sensitivity and specificity (92.86% and 96.24%, respectively) compared with CTC (86.57% and 72.33%, respectively). Correct decisions were made in 94.30% of patients through clinical with dermoscopic teleconsultations and in 79.20% in CTC (P < .001). The only variable associated with an increased likelihood of correct diagnosis was management using teledermoscopy (odds ratio 4.04; 95% confidence interval 2.02-8.09; P < .0001). The cost-effectiveness analysis showed teledermoscopy as the dominant strategy, with a lower cost-effectiveness ratio (65.13 vs 80.84). Potentially, a limitation is the establishment of an experienced dermatologist as the gold standard for the in-person evaluation. The addition of dermoscopic images significantly improves the results of an internet-based skin cancer screening system, compared with screening systems based on clinical images alone.