No open-access location is known for this record yet.
No open-access location is known for this record yet.
BACKGROUND: Plastic surgeons and patients increasingly use social media. Despite evidence implicating its importance in plastic surgery, the large volume of data has made social media difficult to study. OBJECTIVES: The aim of this study was to provide a comprehensive assessment of plastic surgery social media content worldwide by utilizing techniques for analyzing large-scale data. METHODS: The hashtag "#PlasticSurgery" was used to search public Instagram posts. Metadata were collected from posts between December 2018 and August 2020. In addition to descriptive analysis, 2 instruments were created to characterize textual data: a multilingual dictionary of procedural hashtags and a rule-based text classification model to categorize the source of the post. RESULTS: Plastic surgery content yielded more than 2 million posts, 369 million likes, and 6 billion views globally over the 21-month study. The United States had the most posts of 182 countries studied (26.8%, 566,206). Various other regions had substantial presence including Istanbul, Turkey, which led all cities (4.8%, 102,208). The classification model achieved high accuracy (94.9%) and strong agreement with independent raters (κ = 0.88). Providers accounted for 40% of all posts (847,356) and included the categories physician (28%), plastic surgery (9%), advanced practice practitioners and nurses (1.6%), facial plastics (1.3%), and oculoplastics (0.2%). Content between plastic surgery and non-plastic surgery groups demonstrated high textual similarity, and only 1.4% of posts had a verified source. CONCLUSIONS: Plastic surgery content has immense global reach in social media. Textual similarity between groups coupled with the lack of an effective verification mechanism presents challenges in discerning the source and veracity of information.
Loading…
Stored only when both papers are in the corpus — 34 citations exist upstream.
Papers that cite the same references — the relational fallback while embeddings are still backfilling.
Imported from OpenAlex (CC0) — nothing here is generated.