Cliff Rosendahl, Martelle Coetzer-Botha
In their elegantly designed and executed study on training for skin tumour triage for primary-care physicians (PCPs),1 Harkemanne et al. highlight what is arguably obvious: Any structured training in dermatoscopy for PCPs is likely to be beneficial.
This study, involving 216 PCPs, 40% trainees, 87% younger than 46 years and 73% female, was adequately powered, rendering statistically significant findings. The training methods were based on a course previously designed for non-experts,2 notably with an initial emphasis on recognizing the dermatoscopic patterns of benign lesions, before considering dermatoscopic features pointing to malignancy. The methodology of assessment, presenting standardized sets of images from selected benign and malignant categories, was consistent with other studies of dermatoscopic diagnostic accuracy, including methods employed in the assessment of both machine and human diagnostic performance.3 The authors concluded that a short online dermatoscopic training session of 4 h was non-inferior to a longer 12-h training suite, which included the content of the first session. PCPs who completed all of four, monthly, 30-min refresher training sessions showed the best overall final tested performance (p < 0.001).1
