人工智能辅助临床决策支持系统在远程伤口护理中的随机对照试验
一项随机对照试验纳入40例慢性溃疡患者,比较AI辅助远程医疗平台Ulcer Care与常规护理。
要点
- AI辅助远程伤口护理在12周愈合结局上不劣于常规护理。
- UC组门诊就诊次数显著少于对照组(P=0.009)。
- 照护者生活质量改善UC组有更大趋势但未达统计学显著(P=0.08)。
核心结论
- AI辅助远程医疗平台Ulcer Care在伤口愈合结局上不劣于常规伤口护理。
- 该系统显著改善照护者生活质量,并减少门诊就诊次数。
- AI在远程医疗中可作为慢性伤口护理的辅助手段,但需进一步开发AI准确性与系统可用性。
研究设计与样本
- 随机对照试验,于2023年7月至2024年2月开展,纳入40例慢性溃疡患者。
主要发现
- 两组伤口愈合率与伤口面积缩小百分比相似(64.9% vs 73.4%,P=0.88)。
- UC组门诊就诊次数显著少于对照组(5.5±2.6 vs 13.2±10.6,P=0.009)。
- 两组生活质量均改善,UC组照护者FDLQI改善更大(10.3±5.0 vs 5.7±5.9,P=0.08)。
临床怎么用
- 可将AI辅助远程医疗平台作为慢性伤口护理的辅助工具,以减少门诊就诊负担。
- 使用AI系统自动测量伤口大小、分类组织类型并对恶化伤口发出警报,结合医生反馈。
- 关注照护者生活质量改善,AI远程护理可能对此有潜在益处。
局限性
- AI准确性与系统可用性仍需进一步开发完善。
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: Chronic wounds are increasingly prevalent due to an aging population and rising chronic diseases. Effective wound care is often hindered by the inexperience of home caregivers, leading to suboptimal healing outcomes. Telemedicine has emerged as a solution to bridge care gaps, though manual wound assessment remains a significant burden. Artificial intelligence (AI) presents an opportunity to enhance wound evaluation and management. OBJECTIVE: To assess the effectiveness of an AI-assisted clinical decision support system integrated into a telemedicine platform, "Ulcer Care" (UC), in the management of chronic wounds, and to evaluate its impact on the psychosocial outcomes of both patients and caregivers. METHODS: A randomized controlled trial was conducted between July 2023 and February 2024, involving 40 patients with chronic ulcers who were randomly assigned to either the UC group (AI-assisted telemedical care) or the control group (conventional care). The UC system utilized AI to automatically measure wound size, classify tissue types, and issue alerts for deteriorating wounds with doctors' feedback. The primary endpoint was wound healing at 12 weeks, assessed by percentage reduction in wound area and healing status. Secondary endpoints included patient and caregiver quality of life (PU-QOL and FDLQI), clinical visits, and complications. RESULTS: Thirty-five patients completed the study (17 in the UC group, 18 in the control group). The wound healing rates and percentage reduction in wound area were similar between the groups (64.9% vs 73.4%, P = 0.88), but the UC group required significantly fewer clinic visits (5.5 ± 2.6 vs 13.2 ± 10.6, P = 0.009). Both groups showed improvements in quality of life, with the UC group demonstrating a greater improvement in caregiver quality of life (FDLQI improvement: 10.3 ± 5.0 vs 5.7 ± 5.9, P = 0.08). CONCLUSION: The AI-assisted telemedicine platform "Ulcer Care" was found to be noninferior to conventional wound care in terms of wound healing outcomes and significantly improved caregiver quality of life. While the system demonstrated potential for improving chronic wound management, further development of AI accuracy and system usability is needed. This study highlights the utility of AI in telemedicine as an adjunct to chronic wound care, offering potential benefits in reducing healthcare utilization and enhancing patient outcomes.
引用
- Chang DH, Nguyen DK, Nguyen TN, et al. Artificial Intelligence-Assisted Clinical Decision Support System in Telemedical Wound Care: A Randomized Controlled Trial. Ann Plast Surg. 2026. PMID: 41576249. doi:10.1097/SAP.0000000000004554
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。