Background

Despite its efficacy, only about 1% of those who qualify for bariatric surgery undergo it. Herein, we use search analytics to determine the top questions regarding bariatric surgery and evaluate AI-generated answers to these queries, to ascertain potential barriers and assess the quality of information patients find online. The 'People Also Ask' function of Google was queried for the top 100 questions related to the term'weight loss surgery'. These were sorted with Rothwell's classification and related sub-categories. The top 10 unique questions were fed to an AI-powered search engine which generated original responses and sources for user reference. These were graded according to JAMA benchmark criteria for source quality. The answers were then graded by four bariatric surgeons. 50% of questions were fact-based with the most common sub-category being cost, followed by technical details of the surgery. 23% of questions were value-based and pertained to evaluation of the surgery. 17% of questions were policy-based and pertained to indications for surgery. None related to risks or pain. Regarding source quality, 44% presented an author, 59% cited sources, 82% presented publication dates, and 33% contained ownership or sponsorship disclosures. The average quality score was 2.1/4. For the AI-generated answers, these received 4.73/6 for accuracy and 2.20/3 for completeness. Individuals asking questions online about bariatric surgery are most concerned about paying and qualifying for these procedures. AI-generated responses to these questions provided fair results, although the sources to which patients are directed are less reliable.