Diagnostic accuracy and cost-effectiveness of dermoscopy in primary care: a cluster randomized clinical trial

BackgroundThe dermoscope improves general practitioners' (GP) sensitivity for melanoma. However, diagnostic accuracy (DA) and cost-effectiveness of the dermoscope in primary care for the evaluation of all skin lesions suspected of malignancy remains unknown. ObjectivesTo determine the DA and cost-effectiveness of the dermoscope in primary care for skin lesions suspected of malignancy. MethodsIn a cluster randomized clinical trial, 48 Dutch general practices were randomized to either intervention group using a dermoscope or control group using only naked-eye examination. A total of 194 lesions from 170 patients in the intervention group and 222 lesions from 211 patients in the control group were analysed for DA and cost-effectiveness. ResultsThe percentage of correctly diagnosed lesions in intervention group and control group was 50.5% and 40.5% respectively. This was 61.5% and 22.2% for melanomas. In the intervention group, three malignancies were treated with the expectative treatment option compared to none in the control group. The odds ratio (OR) of a correct diagnosis in the intervention group, compared to control group, was 1.51 (95% CI: 0.96-2.37) P=0.07. Consequently, the relative risk was 1.25. The incremental cost-effectiveness ratio was Euro89 (95% CI -Euro60 to Euro598), indicating that using a dermoscope costs an additional Euro89 for one additional correctly diagnosed patient. Additional analyses showed better effects of dermoscopy compared to the control group for 98% of the bootstrap resamples. ConclusionsThe probability of a correct diagnosis was 1.25 times higher using a dermoscope than without a dermoscope. Although this difference is marginally not statistically significant, dermoscopy in general practice appears to be cost effective. We therefore think that GPs should be trained to use a dermoscope, although they should realize that even with the use of a dermoscope not all lesions will be diagnosed correctly.