Lin Hui, Huang Xiaofang, Sheng Yaying, Tang Ning, Lian Hengli, Zhang Wenjie, Zhao Lvjun, Zhu Hanqing, Chang Pingjun, Guo Yingxuan
J Patient Saf. 2025 Mar 1;21(2):62-68. doi: 10.1097/PTS.0000000000001295. Epub 2025 Feb 18.
With the development of day surgery, the characteristics of "short, frequent and fast" ophthalmic surgery are becoming more prominent. However, nurses are not efficient in verifying patients' surgical information, and problems such as patient privacy leakage are becoming more prominent. To improve the situation, we developed a new augmented reality (AR)-based tool for visual recognition and artificial intelligent (AI) interpretation of the pattern and location of patient surgical skin markings for the verification of the correct surgical site and procedure. The tool can also display a variety of other verbally requested patient information. The purpose of this proposal is to evaluate its feasibility of use by surgical nurses in a real clinical setting.
We developed a tool with image recognition technologies to interpretation patient surgical skin markings and match the information obtained with the patients surgical records, thus, verify the patient's surgical information. Verification includes the proper surgical site and type of procedure to be performed. Nurses can interact with the device through its speech recognition capabilities and the device provides them with a variety of other requested patient information via a heads-up display. Three hundred patients in an outpatient ophthalmology clinic were divided into an AR intelligent verification experimental group and a manual verification control group. The accuracy of information verification, work time consumption, and economic cost data were compared between the 2 groups to evaluate the effectiveness of the AR Surgical Information Intelligent Verification Tool in clinical patient surgical information verification.
There was no statistically difference in the correct rates of patient surgical information review between the experimental group (95.33%) and the control group (98.67%) (χ 2 = 2.934, P = 0.087). The median time for information verification was 10.00 (10.00, 11.00) seconds in the experimental group and 21.00 (20.00, 24.00) seconds in the control group, a statistically difference (Z = 0.000, P < 0.001). The experimental group saved 11 seconds per patient per review compared with the control group. Considering 10,531 surgeries in 2023, printing 1 page of surgical information per 9 patients and requiring 4 copies, 4680 pages of printing paper could be saved.
The AR Surgical Information Intelligent Verification Tool has advantages in assisting medical staff in patient surgical information verification, improving nursing efficiency, preventing surgical mark errors or nonstandardization, protecting patient privacy, and saving costs. It has certain research and application value in the scenario of patient surgical information verification in ophthalmic day ward.
随着日间手术的发展，眼科手术“短、频、快”的特点日益突出。然而，护士在核对患者手术信息时效率不高，患者隐私泄露等问题也日益凸显。为改善这种情况，我们开发了一种基于增强现实（AR）的新工具，用于对患者手术皮肤标记的图案和位置进行视觉识别及人工智能（AI）解读，以核实正确的手术部位和手术程序。该工具还能显示各种其他口头要求的患者信息。本提议的目的是评估其在实际临床环境中供外科护士使用的可行性。
我们利用图像识别技术开发了一种工具，用于解读患者手术皮肤标记，并将获取的信息与患者手术记录进行匹配，从而核实患者的手术信息。核实内容包括正确的手术部位和拟进行的手术类型。护士可通过其语音识别功能与该设备交互，设备通过平视显示器为他们提供各种其他要求的患者信息。眼科门诊的300例患者被分为AR智能核实实验组和人工核实对照组。比较两组之间信息核实的准确性、工作耗时和经济成本数据，以评估AR手术信息智能核实工具在临床患者手术信息核实中的有效性。
实验组（95.33%）和对照组（98.67%）患者手术信息核对的正确率差异无统计学意义（χ² = 2.934，P = 0.087）。实验组信息核实的中位时间为10.00（10.00，11.00）秒，对照组为21.00（20.00，24.00）秒，差异有统计学意义（Z = 0.000，P < 0.001）。与对照组相比，实验组每位患者每次核对节省11秒。考虑到2023年有10531例手术，每9例患者打印1页手术信息且需要4份副本，则可节省4680页打印纸。
AR手术信息智能核实工具在协助医护人员进行患者手术信息核实、提高护理效率、防止手术标记错误或不规范、保护患者隐私以及节约成本等方面具有优势。它在眼科日间病房患者手术信息核实场景中具有一定的研究和应用价值。
