Multidimensional indicators of scholarly impact in the skin oncology literature: is there a correlation between bibliometric and altmetric profiles?

Introduction: Bibliometric and altmetric analyses are used to identify landmark publications in their respective research field. We hypothesised that highly cited skin oncology articles correlate positively with the Oxford Evidence Based Medicine scoring level, altmetric score (AS) and rank within the top 100 manuscripts. Methods: Thomson Reuter's Web of Science citation indexing database was searched to identify all English-language skin oncology full-text articles in the last 75 years. The top 100 articles with the highest citation count were analysed by subject matter, publishing journal, author, year, institution, individual and five-year impact factor, AS and Oxford EBM level. Results: 180,132 articles were identified. The most cited article (Hodi et al.) demonstrated improved survival with ipilimumab in patients with metastatic melanoma (7894 citations). The article with the highest AS was Esteva et al. (AS = 576.7, 'dermatologist-level classification of skin cancer with deep neural networks'). No difference was found between evidence level and citation count (r = −0.1239, p = 0.2291), but a significant difference was seen for AS (r = −0.3024, p = 0.0028). AS scores increased over time, whereas bibliometrics did not. Conclusion: This work highlights the most influential work in the skin oncology field in the last 75 years. We have identified a differential relationship between commonly used metrics and evidence level in the field of skin oncology. As the digitalisation of research output and consumption increases, both bibliometric and altmetric analyses need to be considered when an article's impact is being assessed.