Epidural-related maternal fever (ERMF) heightens the risk of intrapartum fever, whereas effective prevention and treatment in clinical practice are currently lacking. Rapid and sensitive screening tools for ERMF are urgently needed to advance relevant research. In response to this challenge, we devise and craft porous Co₃O₄/CuO hollow polyhedral nanocages with p-p heterojunctions derived from metal–organic frameworks. We employ these p-p heterojunctions in conjunction with high-throughput mass spectrometry to conduct metabolic analysis of substantial plasma samples, with only about 0.03 μL per sample. Leveraging these p-p heterojunctions, metabolic signals from complex plasma can be amplified, with great reproducibility. By harnessing the power of machine learning on these metabolic signals, we are able to achieve advanced warning of ERMF with an area under the curve (AUC) of 0.887–0.975 by the differentially metabolic analysis of plasma samples collected upon admission. Furthermore, we can accurately diagnose ERMF with an AUC of 0.850–1.000 by analyzing plasma samples collected at the time of delivery from individuals who have received epidural analgesia. These breakthroughs offer invaluable insights for clinical decision making during labor and have the potential to significantly reduce the incidence of ERMF.

中文翻译：

---

通过 p-p 异质结辅助激光解吸/电离质谱法测定血浆代谢谱，用于硬膜外相关产妇发热的高级预警和诊断


硬膜外相关母热 （ERMF） 增加了产时发热的风险，而目前临床实践中缺乏有效的预防和治疗。迫切需要快速、灵敏的 ERMF 筛查工具来推进相关研究。为了应对这一挑战，我们设计并制作了多孔 Co₃O₄/CuO 空心多面体纳米笼，其 p-p 异质结源自金属有机框架。我们将这些 p-p 异质结与高通量质谱法结合使用，对大量血浆样品进行代谢分析，每个样品仅约 0.03 μL。利用这些 p-p 异质结，可以放大来自复杂血浆的代谢信号，具有极强的重现性。通过利用机器学习对这些代谢信号的力量，我们能够通过对入院时收集的血浆样本进行差异代谢分析，实现曲线下面积 （AUC） 为 0.887-0.975 的 ERMF 高级预警。此外，我们可以通过分析在分娩时从接受硬膜外镇痛的个体那里收集的血浆样本来准确诊断 AUC 为 0.850-1.000 的 ERMF。这些突破为分娩期间的临床决策提供了宝贵的见解，并有可能显著降低 ERMF 的发生率。