文献速递消化、代谢与肾脏急症重症综合管理与镇静镇痛

不同诊断标准下重症监护病房患者再喂养低磷血症发生率差异及影响因素分析

[Differences in the incidence of refeeding hypophosphatemia in intensive care unit patients under different diagnostic criteria and an analysis of influencing factors].
PubMed 文献 · Zhonghua Wei Zhong Bing Ji Jiu Yi Xue · 2026-01-01
一句话结论ICU营养支持患者中,再喂养低磷血症发生率随诊断标准不同达38.2%–65.0%,早期血磷动态下降幅度是各标准下的一致预测因素。

该前瞻性观察研究纳入哈尔滨医科大学附属第二医院ICU接受营养支持超过72小时的310例成人患者,最终分析283例,按三种诊断标准判定再喂养低磷血症(RH)。联合性RH(CRH)、绝对性RH(ARH)、相对性RH(RRH)的发生率分别为38.2%、46.3%、65.0%。多因素Logistic回归显示,初始至喂养前血磷下降幅度、喂养前至喂养后第1天血磷下降幅度在三种标准下均为影响因素,饮酒史为CRH和ARH的影响因素,利尿剂使用为RRH的影响因素。所构建的三个预测模型AUC分别为0.787、0.784、0.781,作者认为早期血磷动态变化对预测RH有临床指导意义。

要点

  • 三种标准下RH发生率差异大:CRH 38.2%、ARH 46.3%、RRH 65.0%,床旁判定RH时须注明所用标准。
  • 喂养前至喂养后第1天血磷下降幅度是CRH(OR=3.806)、ARH(OR=2.302)、RRH(OR=20.193)的共同影响因素。
  • 初始至喂养前血磷下降幅度亦为三种标准下RH的影响因素,CRH OR=2.027、ARH OR=1.645、RRH OR=3.154。
  • 饮酒史为CRH(OR=2.229)和ARH(OR=2.121)的影响因素;利尿剂使用为RRH的影响因素(OR=0.314)。
  • 三个预测模型AUC为0.781–0.787,敏感度0.731–0.861、特异度0.603–0.731,作者认为可为优化高危患者电解质监测策略提供依据。

适用人群与场景

  • 2024年5月至2025年3月在哈尔滨医科大学附属第二医院ICU住院、接受营养支持治疗超过72小时的成人患者,共310例入选、283例纳入分析。

诊断标准与关键阈值

  • 联合性再喂养低磷血症(CRH):营养支持后72小时内血磷水平下降≥0.16 mmol/L且血磷水平<0.65 mmol/L。
  • 绝对性再喂养低磷血症(ARH):营养支持后72小时内血磷水平<0.65 mmol/L。
  • 相对性再喂养低磷血症(RRH):无论基线值如何,营养支持后72小时内血磷水平下降≥0.16 mmol/L。

研究设计与关键结果

  • 前瞻性观察研究,入选310例、分析283例,按三种标准判定RH:CRH 108例(38.2%)、ARH 131例(46.3%)、RRH 184例(65.0%)。
  • CRH影响因素:初始至喂养前血磷下降幅度OR=2.027(95%CI 1.528–2.687,P<0.001)、喂养前至喂养后第1天血磷下降幅度OR=3.806(95%CI 2.504–5.785,P<0.001)、饮酒史OR=2.229(95%CI 1.089–4.563,P=0.028)。
  • ARH影响因素:初始至喂养前血磷下降幅度OR=1.645(95%CI 1.284–2.109,P<0.001)、喂养前至喂养后第1天血磷下降幅度OR=2.302(95%CI 1.641–3.229,P<0.001)、饮酒史OR=2.121(95%CI 1.119–4.023,P=0.021)。
  • RRH影响因素:初始至喂养前血磷下降幅度OR=3.154(95%CI 2.046–4.860,P<0.001)、喂养前至喂养后第1天血磷下降幅度OR=20.193(95%CI 9.244–44.113,P<0.001)、利尿剂使用OR=0.314(95%CI 0.142–0.691,P=0.004)。
  • 预测模型区分度:CRH模型AUC=0.787(95%CI 0.735–0.840),敏感度0.731、特异度0.731;ARH模型AUC=0.784(95%CI 0.730–0.838),敏感度0.833、特异度0.611;RRH模型AUC=0.781(95%CI 0.728–0.833),敏感度0.861、特异度0.603。

证据类型、等级与局限

  • 证据类型:前瞻性观察研究。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为,不同诊断标准下早期血磷水平动态变化是预测ICU患者发生RH的重要危险因素之一。
  • 作者认为,据此构建的相关风险预测模型有临床指导意义,可为优化高危患者电解质监测策略提供依据。

提炼说明

  • 本条依据论文摘要整理;摘要未涉及「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开

OBJECTIVE: To compare the occurrence of refeeding hypophosphatemia (RH) in critically ill patients under different diagnostic criteria, explore its influencing factors, and establish a corresponding risk prediction model.

METHODS: A prospective observational study was adopted. A total of 310 adult patients who were hospitalized in the intensive care unit (ICU) of the Second Affiliated Hospital of Harbin Medical University from May 2024 to March 2025 and received nutritional support therapy for more than 72 hours were selected. Demographic data, relevant scores and laboratory test data within 24 hours of ICU admission were collected, and the dynamic monitoring results of serum electrolytes before and after nutritional support therapy were recorded simultaneously. Patients were determined to have RH or not according to the following three diagnostic criteria, and divided into the RH group and the non-RH group. 1) combined refeeding hypophosphatemia (CRH): serum phosphorus level decreased by ≥0.16 mmol/L and the serum phosphorus level <0.65 mmol/L within 72 hours after nutritional support; 2) absolute refeeding hypophosphatemia (ARH): serum phosphorus level was <0.65 mmol/L within 72 hours after nutritional support; 3) relative refeeding hypophosphatemia (RRH): regardless of the baseline value, serum phosphorus level decreased by ≥0.16 mmol/L within 72 hours after nutritional support. Multivariate Logistic regression was performed to analyze the risk factors for RH in ICU patients under different diagnostic criteria, and corresponding risk prediction models were constructed respectively. Receiver operator characteristic curve (ROC curve) was used to evaluate the risk discrimination ability of each model.

RESULTS: A total of 283 patients were enrolled. Among them, 108 patients (38.2%), 131 patients (46.3%), and 184 patients (65.0%) were diagnosed as having CRH, ARH, and RRH, respectively, according to the corresponding diagnostic criteria. Multivariate Logistic regression analysis showed that the decrease amplitude of serum phosphorus level from initial to pre-feeding [odds ratio (OR)=2.027, 95% confidence interval (95%CI) was 1.528-2.687, P<0.001], the decrease amplitude of serum phosphorus level from pre-feeding to the first day post-feeding (OR=3.806, 95%CI was 2.504-5.785, P<0.001), and a history of alcohol consumption (OR=2.229, 95%CI was 1.089-4.563, P=0.028) were identified as influencing factors for the occurrence of CRH. The decrease amplitude of serum phosphorus level from initial to pre-feeding (OR=1.645, 95%CI was 1.284-2.109, P<0.001), the decrease amplitude of serum phosphorus level from pre-feeding to the first day post-feeding (OR=2.302, 95%CI was 1.641-3.229, P<0.001), and a history of alcohol consumption (OR=2.121, 95%CI was 1.119-4.023, P=0.021) were also identified as influencing factors for the occurrence of ARH. The decrease amplitude of serum phosphorus level from initial to pre-feeding (OR=3.154, 95%CI was 2.046-4.860, P<0.001), the decrease amplitude of serum phosphorus level from pre-feeding to the first day post-feeding (OR=20.193, 95%CI was 9.244-44.113, P<0.001), and the use of diuretics (OR=0.314, 95%CI was 0.142-0.691, P=0.004) were identified as influencing factors for the occurrence of RRH. ROC curve analysis demonstrated that the areas under the curve (AUC) of the prediction models for CRH, ARH, and RRH were 0.787 (95%CI was 0.735-0.840), 0.784 (95%CI was 0.730-0.838), and 0.781 (95%CI was 0.728-0.833), respectively; and the sensitivities were 0.731, 0.833, and 0.861; while the specificities were 0.731, 0.611, and 0.603, respectively.

CONCLUSIONS: Under different diagnostic criteria, the dynamic changes of early serum phosphorus levels are one of the important risk factors for predicting the occurrence of RH in ICU patients. The relevant risk prediction models constructed on this basis have clinical guiding significance and can provide a basis for optimizing the electrolyte monitoring strategy for high-risk patients.

来源:Zhonghua Wei Zhong Bing Ji Jiu Yi Xue,查看原文。版权归原作者所有。

参考来源

  1. Zhang X, Fan Y, Yu H, et al. [Differences in the incidence of refeeding hypophosphatemia in intensive care unit patients under different diagnostic criteria and an analysis of influencing factors]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2026. PMID: 42281543. doi:10.3760/cma.j.cn121430-20250918-00487

供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。