Studies in which the diagnostic performance of ultrasonography and palpation during initial investigations or follow-up of patients with American Joint Committee on Cancer (AJCC) stage I-III cutaneous or mucous-membrane melanoma were eligible for inclusion. Studies had to have at least 30 paired palpation-ultrasound lymph-node examinations. Characteristics of included patients needed to be well described with sensitivity and specificity values that were or could be calculated. Case reports and studies on ocular melanoma, development of ultrasound criteria and of primary melanoma lesions before excision were excluded.
The included studies were mostly from European radiology departments or dermatology departments that specialised in patients with melanoma. The primary sites of melanoma were the limb, trunk, head and neck or other. Where stated, the interval between palpation and ultrasonography ranged from less than 24 hours to 14 days. Frequency of the ultrasound probe varied between studies, as did the type of transducer used. Ultrasonography was undertaken by experienced operators in most studies. Where stated, individual melanomas had Breslow thicknesses of greater than 1mm; one study had more than 50% of patients with very thin melanoma lesions (<0·76mm). Most studies included patients with high-risk melanoma for whom early diagnosis of nodal invasion was justified. One study included a small number of patients with stage IV melanoma. Mean patient age ranged from 47 to 58 years of age.
Three authors independently selected studies for inclusion in the review; disagreements were resolved by consensus.