文献速递关节外科

评估人工智能护理辅助工具在关节置换术围手术期患者咨询中的实用性与可接受性

Evaluating the Utility and Acceptability of an Artificial Intelligence-Powered Care Aid for Perioperative Patient Inquiries in Joint Arthroplasty Surgery.
PubMed 文献 · J Arthroplasty · 2026-04-21

这项前瞻性观察性试点研究采用两阶段序贯设计,先纳入10例患者预试验,再对30例初次全髋或全膝关节置换患者实施。32例完成术前调查的患者中,84.8%成功使用定制人工智能平台,66.7%认为平台有临床价值,84.4%报告回答可理解性高,75.0%报告准备度提升。结论认为该平台作为术前教育辅助工具可行且有效,但需更大样本对照研究验证。

要点

  • 84.8%患者成功使用定制人工智能平台,56.3%将其作为信息来源。
  • 66.7%用户认为平台有临床价值,84.4%报告回答可理解性高。
  • 75.0%报告准备度提升,80.4%表示愿意推荐该平台。

核心结论

  • 定制人工智能平台作为全关节置换术前教育辅助工具可行,多数使用者报告回答可理解性高、感知回答完整且准备度提升。
  • 需要更大样本、设对照组的比较研究来确定该平台相对于标准教育的影响并评估下游临床结局。

研究设计与样本

  • 前瞻性观察性试点研究,采用两阶段序贯设计:先10例预先队列,再对30例额外初次全髋或全膝关节置换患者实施。
  • 32例患者完成术前调查,其中膝关节50.0%、髋关节50.0%;术后6周随访时24/32(75.0%)完成术后调查。

主要发现

  • 84.8%成功使用定制人工智能平台;59.4%提交1至4个问题,25.1%提交5个或以上问题。
  • 84.4%以外科医生提供的教育材料为主要信息来源,56.3%以平台参与为信息来源。
  • 66.7%用户认为平台有临床价值(评分≥4),84.4%报告回答可理解性高,75.1%表示手术疑问得到满意解决,75.0%报告准备度提升,80.4%表示愿意推荐平台。
  • 主要查询领域包括围手术期流程(22.8%)、术后关注(22.0%)和活动参数(12.9%)。
  • 术后6周,24/32(75.0%)完成调查,22/24(91.7%)报告成功访问平台;7/24(29.2%)提交零个问题,5/24(20.8%)提交超过10个问题;7/24(29.2%)报告所有问题被清晰回答(评分5),7/24(29.2%)报告平台未回答其问题(评分1);14/24(58.3%)报告康复指导非常容易理解(评分5)。

临床怎么用

  • 该平台可作为全关节置换术前患者教育的辅助工具,帮助解答围手术期流程、术后关注和活动参数等常见问题。
  • 使用中需注意参与度异质性,部分患者可能不提问或提问超过10个,术后问题解决满意度存在分化。

局限性

  • 为单中心试点研究,样本量小,缺乏对照组,无法确定平台相对于标准教育的效果。
  • 术后随访仅24例完成调查,存在失访偏倚,且未评估下游临床结局。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Conventional educational materials for total joint arthroplasty frequently exhibit deficiencies in readability and personalization. This investigation assessed a custom-tailored artificial intelligence (CTAI) platform to evaluate its efficacy in enhancing preoperative patient education within arthroplasty surgery. METHODS: This prospective, observational pilot study employed a two-phase sequential design: a 10-patient a priori cohort followed by implementation with 30 additional participants undergoing primary total hip or knee arthroplasty. Participants accessed the platform during the preoperative period and early postoperative recovery. There were 32 patients (50.0% knee, 50.0% hip) who completed the preoperative survey. Outcomes were captured using structured surveys administered at the presurgical screening visit and again at the 6-week follow-up using a phase-specific postoperative instrument that retained core items and incorporated recovery-focused domains. Measures included feasibility (e.g., platform access and engagement) and patient-reported experience, including usefulness, understandability of responses, perceived completeness of answers, and perceived preparedness/readiness. RESULTS: The majority (84.8%) successfully engaged with the CTAI; 59.4% submitted one to four queries, while 25.1% submitted five or more. Surgeon-provided educational materials constituted the predominant information source (84.4%), followed by CTAI engagement (56.3%). The platform was deemed clinically valuable (≥ 4 rating) by 66.7% of users, with 84.4% reporting high comprehensibility of responses and 75.1% indicating satisfactory resolution of surgical inquiries. Furthermore, 75.0% reported enhanced preparedness, and 80.4% expressed intention to recommend the platform. Predominant query domains included perioperative logistics (22.8%), postoperative concerns (22.0%), and activity parameters (12.9%). At six weeks postoperatively, 24 of 32 (75.0%) completed the postoperative survey, and 22 of 24 (91.7%) reported successful platform access. Engagement was heterogeneous, with seven of 24 (29.2%) submitting zero questions and five of 24 (20.8%) submitting greater than 10. For postoperative question resolution, seven of 24 (29.2%) reported all questions were answered clearly (score 5), whereas seven of 24 (29.2%) reported the CTAI did not answer their questions (score 1). Recovery instruction comprehensibility remained high, with 14 of 24 (58.3%) reporting the instructions were very easy to understand (score 5). CONCLUSIONS: Preliminary findings demonstrate that the CTAI platform is feasible to deploy as an efficacious adjunctive tool for preoperative total joint arthroplasty education. Most participants who accessed the platform reported favorable understandability and perceived completeness of responses and reported increased preparedness for surgery. The diversity of inquiry domains underscores the platform's capacity for addressing varied patient concerns. Larger comparative studies with appropriate control groups are needed to determine the platform's impact relative to standard education and to evaluate downstream clinical outcomes.

来源:J Arthroplasty,查看原文。版权归原作者所有。

引用

  1. Zepeda KE, Burgio C, Yared T, et al. Evaluating the Utility and Acceptability of an Artificial Intelligence-Powered Care Aid for Perioperative Patient Inquiries in Joint Arthroplasty Surgery. J Arthroplasty. 2026. PMID: 42025854. doi:10.1016/j.arth.2026.04.049

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。