E-patients in Oncology: a corpus-based characterization of medical terminology in an online cancer forum

Abstract

This study aimed to characterize medical terms in an online cancer forum, with particular focus on specialization and semantic features. A three-step analysis was carried out on a 60-million-word corpus to detect and characterize the most typical medical terms used in a cancer forum by means of (1) keywords contrastive, (2) co-text-guided, and (3) semantic analyses. More than half of the 1000 words analysed were medical terms according to the co-text-guided analysis carried out. Most of them (73%) were dictionary-defined medical terms, followed by co-text-defined terms (9%) and medical initialisms (8.5%). The semantic analysis showed a higher number of terms within the fields of Anatomy, Treatment, Hospital and Symptoms. Our findings suggest that medical terms are commonly used in cancer forums, especially to share e-patients’ concerns about treatment, symptoms and hospital environment. The method followed is efficient and could be applied in future studies. Altogether, this article contributes to characterizing medical terms used by e-patients in online cancer forums