目的  探讨影响维持性血液透析(maintenance hemodialysis，MHD)患者发生中心静脉狭窄(central venous stenosis，CVS)的危险因素，并评估其预测能力。 方法 回顾性分析2022年1月—2023年9月于赣南医科大学第一附属医院接受治疗的MHD患者临床资料。将确诊CVS的患者设为观察组，无CVS症状的设为对照组。收集一般资料及实验室检查指标，采用组间差异分析及Logistic回归分析筛选独立危险因素，使用绘制受试者工作特征曲线下面积(area under curve，AUC)评估预测效能。 结果 共纳入146例患者，其中观察组75例(初诊狭窄组54例，再狭窄组21例)，对照组71例。组间差异分析结果显示：与对照组相比，观察组淋巴细胞计数降低(Z=-3.350，P＜0.001)，中性粒细胞-淋巴细胞比值(neutrophil-lymphocyte ratio，NLR)(Z=2.878，P=0.004)和血小板体积相关指数(platelet volume correlation indices,PVIs)(Z=3.500，P＜0.001)增高。与初诊狭窄组相比，再狭窄组的平均血小板体积(mean platelet volume，MPV)增大(t=-2.152，P=0.037)，是CVS介入手术后再狭窄的独立危险因素(OR=1.821，95%CI：1.020～3.247，P=0.043)。多因素Logistic回归分析表明：淋巴细胞计数(OR=0.102，95% CI：0.016～0.653，P=0.016)和PVIs(OR=1.343，95% CI：1.015～1.778，P=0.039)是MHD患者发生CVS的独立危险因素，PVIs(AUC=0.849)预测效能优于淋巴细胞计数(AUC=0.798),二者联合预测CVS的AUC值为0.854(敏感性80.8%，特异性78.4%)。 结论 PVIs及淋巴细胞计数是MHD患者发生CVS的独立危险因素，二者联合对CVS具有良好的预测价值，可作为CVS早期筛查与风险评估的生物标志物。

Objective  To investigate the risk factors for the occurrence of central venous stenosis (CVS) in patients undergoing maintenance hemodialysis (MHD) and to evaluate their predictive capabilities.  Methods   A retrospective study was conducted on the clinical data of 146 MHD patients treated at the First Affiliated Hospital of Gannan Medical University between January 2022 and September 2023. Seventy-five patients diagnosed with CVS were assigned to the observation group (comprising 54 cases of primary stenosis and 21 cases of restenosis), and 71 patients without CVS symptoms served as the control group. General clinical data and laboratory parameters were collected. Intergroup differences and univariate and multivariate logistic regression were used to identify the independent risk factors. The predictive performance was estimated using the area under the receiver operating characteristic curve (AUC) and other relevant indices.  Results  Intergroup analysis showed that the observation group had lower lymphocyte count (Z=-3.350, P＜0.001) and higher neutrophil-to-lymphocyte ratio (NLR) (Z=2.878, P=0.004) and platelet volume related indices (PVIs) (Z=3.500, P＜0.001), as compared those with the control group. Compared with the primary stenosis group, the restenosis group had bigger mean platelet volume (MPV) (t=-2.152, P=0.037), which was an independent risk factor for restenosis following CVS intervention (OR=1.821, 95% CI: 01.020～3.247, P=0.043). Multifactorial logistic regression indicated that lymphocyte count (OR=0.102, 95% CI: 0.016～0.653, P=0.016) and PVIs (OR=1.343, 95% CI: 1.015～1.778, P=0.039) were the independent risk factors for CVS in MHD patients. The predictive performance of PVIs (AUC=0.849) was superior to that of the lymphocyte count (AUC=0.798). The predictive value for CVS by combined both factors yielded an AUC of 0.854, with a sensitivity of 80.8% and a specificity of 78.4%.  Conclusion  PVIs and lymphocyte count are independent risk factors for the development of CVS in MHD patients. Combination of these two parameters obtains a higher predictive value and can serve as effective biomarkers for early screening and risk assessment of CVS.