A Machine Learning Analysis of Queries Related to Blepharoplasty. Issue 1 (March 2023)

• Record Type: Journal Article
• Authors: Dhanda, Aatin K.; Tseng, Christopher C.; Gao, Jeff; Talmor, Guy; Paskhover, Boris
• Is Part Of: American journal of cosmetic surgery. Volume 40:Issue 1(2023)
• Page Start: 52
• Page End: 57
• Publication Date: 2023-03
• Subjects: blepharoplasty -- facial cosmetic surgery -- facial plastic surgery -- general cosmetic surgery; Surgery, Plastic; Periodicals; 617.95
• DOI: 10.1177/07488068211052800
• Languages: English
• ISSNs: 0748-8068

Abstract

Introduction: Blepharoplasty is a commonly performed cosmetic procedure which can help reduce age-related changes around the eyes. Like other surgical procedures, patient satisfaction is highly correlated with physician responsiveness to questions. As social media becomes more popular, patients are increasingly turning to online sources of information such asRealself.com, a website where patients can ask questions to verified physicians. We extracted and analyzed blepharoplasty-related questions on Realself to gain greater understanding of commonly asked pre- and postoperative questions. Materials and Methods: A Web crawler was used to gather the text for each question under the search term "eyelid surgery." Questions were individually assigned by the authors to respective pre- or postoperative categories. Preoperative categories included eligibility for surgery, nonsurgical options, potential adverse effects, surgeon recommendations, surgical techniques and logistics, ability to pursue other surgeries following blepharoplasty, cost, and miscellaneous. Postoperative question categories included symptoms after surgery, appearance, behavior allowed/disallowed, options to revise surgery, and miscellaneous. Machine learning was then utilized to establish the most common pre- and postoperative questions. Results: 2009 blepharoplasty questions were extracted in total. A total of 60.93% of questions were preoperative related and 39.07% of questions were postoperative related. Preoperative questions were predominantly about eligibility for surgery (43.85%), while the majority of postoperative question were related to symptoms after surgery (17.42%) and appearance (10.80%). Machine learning analysis showed that most preoperative questions inquired about the possibility of correcting specific features, whereas most postoperative questions focused on resolving complications following surgery. Conclusion: As health information becomes more prevalent online, our results underscore the need for greater awareness about frequently asked questions to help surgeons preemptively address patient concerns about blepharoplasty. Our 10 most common questions can be used in the clinical setting as a patient educational handout to augment the preoperative experience.