Abstract:
Background Studies have shown the benign to malignant ratio of excised pigmented skin lesions is suboptimal in primary care. Objectives To assess the impact of dermoscopy and short-term sequential digital dermoscopy imaging (SDDI) on the management of suspicious pigmented skin lesions by primary care physicians. Methods 63 primary care physicians were trained in the use of dermoscopy and SDDI (interventions) and then recruited 327 patients (374 pigmented lesions) requiring biopsy or referral in routine care by naked eye examination. They were then given a dermatosope and the option of a SDDI instrument, and change of diagnosis and management was assessed. Results Following the use of the interventions on 374 lesions a total of 163 lesions (43.6%) were excised or referred, representing a reduction of 56.4% (95% CI 51.2% to 61.5%). Of the 323 lesions confirmed to be benign, 118 (36.5%) were excised or referred, leading to a reduction of 63.5% (95% CI 58.0% to 68.7%; p<0.0005) in those requiring excision or referral. The baseline naked eye examination benign to melanoma ratio was 9.5:1 which decreased to 3.5:1 after the diagnostic interventions (p<0.0005). Of the 42 malignant lesions recruited (34 melanoma, 6 pigmented BCC and 2 Bowen's disease) only 1 in situ melanoma was incorrectly managed (patient to return if changes occur), resulting in the correct management of 97.6% (95 CI 87.4-99.9) and 97.1% (95% CI 84.7 to 99.9) of malignant pigmented lesions and melanoma respectively. The improved management of benign lesions occurred with no significant reduction in the correct management of melanoma (p=0.32) or all malignant lesions (p=0.15). When comparing the initial naked eye diagnosis with the final diagnosis (using dermoscopy with or without SDDI) there was an almost doubling of sensitivity for the diagnosis of melanoma (37.5% 95% CI 21.1% to 56.3% v 71.9% 95% CI 53.3% to 86.3%; p=0.006), a 16% increase in confidence of diagnosis (p=0.049) and an increase in the certainty of melanoma for the true melanoma lesions (p=0.02). While there was no significant difference in specificity for the diagnosis of melanoma following the interventions (84.6% 95% CI 80.0% to 88.5% v 86.6% 95% CI 82.2% to 90.3%; p=0.48) there was a significant improvement in the doctors' certainty of non-melanoma (p=0.02) and a 21% increase in the confidence of diagnosis (p<0.0005) in the true non-melanoma skin lesions. Conclusions In a primary care setting the combination of dermoscopy and short-term SDDI reduces the excision or referral of benign pigmented lesions by more than half while nearly doubling the sensitivity for the diagnosis of melanoma.