AI辅助全景片检测根尖周炎的成本效果分析
本研究基于一项AI辅助检测根尖病变的随机研究数据,构建决策分析模型,从德国混合支付方视角评估AI辅助全景片检测根尖周炎的成本效果。结果显示AI支持降低了每例平均成本并提高了治疗有效性,主要归因于特异性改善减少了假阳性检出;效果因检查者经验而异,初级临床医生获益最大。
要点
- AI辅助可降低每例平均成本并提高治疗有效性,优于无辅助检查者。
- 成本效果增益主要源于特异性改善,减少假阳性检出。
- 初级临床医生成本节省和有效性提升最大,高年资检查者灵敏度降低。
核心结论
- AI辅助诊断通过减少过度治疗具有改善成本效果的显著潜力,在经验较少的从业者中获益最明显。
研究设计与样本
- 基于一项AI辅助检测根尖病变的随机研究数据,建立决策分析模型,从德国混合支付方视角分析成本与效果。
主要发现
- AI支持降低了每例平均成本并提高了治疗有效性,优于无辅助检查者表现。
- 增益主要驱动因素为特异性改善,减少了假阳性检出。
- 效果因检查者经验而异:初级临床医生成本节省和有效性增益最大;高年资检查者灵敏度降低且有效性略低,成本相近。
临床怎么用
- AI辅助诊断可优先用于经验较少的临床医生,以减少过度治疗并改善成本效果。
- 可根据检查者经验水平调整AI系统,以进一步增强临床和经济影响。
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: Artificial intelligence (AI) is transforming medical imaging, yet its economic impact in dentistry remains largely unexplored. AIM: This study evaluated the cost-effectiveness of AI-assisted detection of apical periodontitis on panoramic radiographs, including downstream clinical decision-making. MATERIAL AND METHODS: Using data from a randomised study on AI-assisted detection of apical lesions, a decision-analytic model was established to analyse costs and effectiveness from a German mixed-payer perspective. RESULTS: AI support reduced average costs per case and increased treatment effectiveness, outperforming unaided examiner performance. These gains were primarily driven by improved specificity, reducing false-positive detection. However, effects varied by examiner experience; junior clinicians achieved the greatest cost savings and effectiveness gains, whereas senior examiners showed reduced sensitivity and slightly lower effectiveness at similar costs. CONCLUSION: AI-assisted diagnostics offer significant potential to improve cost-effectiveness by reducing overtreatment, with benefits being most pronounced among less experienced practitioners. Adapting AI systems to individual examiners or experience levels might further enhance clinical and economic impact.
引用
- Benz L, Pul U, Brock T, et al. Cost-Effectiveness of AI-Assisted Detection of Apical Periodontitis on Panoramic Radiographs. Int Endod J. 2026. PMID: 41826269. doi:10.1111/iej.70142
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。