%0 Journal Article
%T 低血压预测指数应用的局限和不足研究进展<br>Progress of the Limitation and Inadequacy of Hypotension Prediction Index in Clinical Application
%A 周函
%A 朱汝佳
%A 蒋宗明
%J Advances in Clinical Medicine
%P 1456-1464
%@ 2161-8720
%D 2024
%I Hans Publishing
%R 10.12677/acm.2024.1441181
%X 低血压预测指数是一种将处于专利保护期内的算法规则嵌入商用的围术期血流动力学监测系统中进行血压预测，主要依赖动脉血压波形的多种特征性参数预测低血压出现的可能性。目前证据表明低血压预测指数能减少低血压发生率并降低其严重程度，但因模型构建时数据选择的系统性偏差会造成预测低血压不准确或过度估计。本文将对模型在数据选择偏倚、数据泄露、对术中和术后低血压预测的不足以及需要改进的方面等进行综述，为临床准确应用提供借鉴和指导。<br />
The hypotension prediction index (HPI) is a proprietary prediction model incorporated into a commercially available intraoperative hemodynamic monitoring system. The HPI uses multiple features of the arterial blood pressure waveform to predict hypotension. The current available evidence demonstrated that HPI lowers the incidence and severity of hypotension, whereas due to systematic bias in data selection that resulted in an overestimation or inaccuracy of the current MAP value’s ability to predict future hypotension. In this review, we will discuss its problem in data selection, its misconceptions in preventing upcoming hypotension events, inadequacy for the patients’ prognosis in an attempt to provide guidance for clinical use.
%K 低血压预测指数，低血压，偏倚，数据泄露<br>Hypotension Prediction Index
%K Hypotension
%K Bias
%K Data Leakage
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=84978