Gencer Batuhan, Arzu Ufuk, Orhan Serdar Satılmış, Dinçal Turgut, Ekinci Mehmet
Department of Orthopedics and Traumatology, Marmara University Pendik Training and Research Hospital, 34890 Istanbul, Turkey.
J Clin Med. 2025 Apr 24;14(9):2942. doi: 10.3390/jcm14092942.
: Despite the rising tendency to depend on ChatGPT for medical counselling, it is imperative to evaluate ChatGPT's capacity to address sensitive subjects that patients often hesitate to discuss with their physicians. The objective of this study was to evaluate the recommendations provided by ChatGPT for sexual activity subsequent to total hip arthroplasty (THA) by orthopaedic surgeons with varying degrees of experience, as well as using standardized scoring systems. : Four patient scenarios were developed, reflecting different ages and indications for THA. Twenty-four questions were asked to ChatGPT 4.0, and responses were evaluated by three different orthopaedic surgeons. All responses were also scored using defined standardized scales. : No response was found to be 'faulty' or 'partial' by any of the observers. While the lowest mean score was attributed by the orthopaedic surgeon with less than five years of experience, the highest mean score was attributed by the orthopaedic surgeon with more than 15 years of experience but not actively working in the field of arthroplasty. An analysis of the data across scenarios revealed that in general, the scores decreased in the more specialized scenarios ( > 0.05). : ChatGPT shows potential as a supplementary resource for addressing sensitive postoperative questions such as sexual activity after THA. However, its limitations in providing nuanced, patient-specific recommendations highlight the need for further refinement. While ChatGPT can support general patient education, expert clinical guidance remains essential for addressing complex or individualized concerns.
尽管越来越倾向于依靠ChatGPT进行医疗咨询，但评估ChatGPT处理患者通常不愿与医生讨论的敏感话题的能力至关重要。本研究的目的是评估ChatGPT为不同经验程度的骨科医生提供的关于全髋关节置换术（THA）后性活动的建议，并使用标准化评分系统进行评估。：设计了四种患者场景，反映了不同的年龄和THA适应症。向ChatGPT 4.0提出了24个问题，并由三名不同的骨科医生对回答进行评估。所有回答也使用定义的标准化量表进行评分。：没有任何一位观察者发现回答有“错误”或“不完整”。经验少于五年的骨科医生给出的平均分数最低，而经验超过15年但未积极从事关节置换领域工作的骨科医生给出的平均分数最高。对不同场景下的数据进行分析后发现，一般来说，在更专业的场景中分数会降低（>0.05）。：ChatGPT显示出作为解决诸如THA后性活动等敏感术后问题的补充资源的潜力。然而，其在提供细致入微、针对患者具体情况的建议方面的局限性凸显了进一步改进的必要性。虽然ChatGPT可以支持一般的患者教育，但专家临床指导对于解决复杂或个性化问题仍然至关重要。
