目的 探讨血清乳酸脱氢酶（LDH）在中晚期肝癌患者接受靶向联合免疫治疗后的预后预测价值。方法 选取2022年1月—2024年8月在莆田学院附属医院肿瘤内科经病理和影像学检查确诊的中晚期肝癌患者作为研究对象。从医院的电子病历系统中收集患者的基线资料，随访截止2025年8月，并记录随访结果，包括患者的疾病缓解情况和死亡情况，以及无疾病进展生存期（PFS）、总生存期（OS）。采用Kaplan-Meier方法绘制不同基线LDH水平患者的OS生存曲线，并通过Log-rank检验比较生存曲线。同时，运用多因素Cox比例风险回归分析探讨影响中晚期肝癌患者在接受靶向联合免疫治疗后OS的相关因素。结果 结果显示，在50例肝癌患者中，基线LDH低于200 U/L的有15例，而高于200 U/L的有35例。与基线LDH＜200 U/L组相比，基线 LDH≥200 U/L患者PFS、OS更短，差异均有统计学意义（χ2分别为5.51、15.6，P值分别为0.019、0.017）。治疗8周后，与LDH降低患者相比，LDH升高患者OS更短，差异有统计学意义（χ2=13.2，P=0.04）。多因素Cox比例风险回归分析结果表明，基线LDH水平超过200 U/L是中晚期肝癌患者接受靶向联合免疫治疗后OS的影响因素［P=0.035，HR（95%CI）=5.03（1.12，22.54）］。结论 基线LDH水平较低的患者表现出更好的OS。基线LDH水平可以作为预测中晚期肝癌患者在接受靶向联合免疫治疗时预后的指标。
Objective To evaluate the prognostic significance of serum lactate dehydrogenase（LDH）levels in patients with advanced hepatocellular carcinoma（HCC）undergoing targeted therapy combined immunotherapy．Methods Patients diagnosed with advanced HCC were selected in Putian College Affiliated Hospital from January 2022 to August 2024，diagnosed with pathological and imaging examinations results．Patient baseline data were collected from the hospital’s electronic medical records，with follow-up extending until August 2025．We documented outcomes such as disease response and mortality，along with progression-free survival（PFS）and overall survival（OS）．Kaplan-Meier survival curves were constructed based on baseline LDH levels，and the Log-rank test was employed for comparison．Additionally，multivariate Cox proportional hazards regression analysis was conducted to identify factors influencing OS in patients receiving targeted therapy combined immunotherapy．Results Among the 50 patients，15 had baseline LDH levels below 200 U/L，while 35 had levels above．Patients with baseline LDH≥200 U/L had significantly shorter PFS and OS than those with baseline LDH ＜200 U/L（χ2=5.51 and 15.6 for PFS and OS，respectively；P=0.019 and 0.017，respectively）．After 8 weeks of treatment，patients with increased LDH had significantly shorter OS compared with patients with decreased LDH（χ2=13.2，P=0.04）．Multivariate Cox proportional hazards regression analysis indicated that a baseline LDH level exceeding 200 U/L is an independent prognostic factor for OS in patients with intermediate to advanced HCC receiving targeted therapy combined with immunotherapy（P=0.035，HR 5.03［1.12，22.54］）．Conclusions Patients with lower baseline LDH levels demonstrated better OS，suggesting that baseline LDH can serve as an important prognostic indicator for advanced HCC patients undergoing targeted combined immunotherapy．
目的：描述发病48 h内急性缺血性卒中（acute ischemic stroke，AIS）患者依达拉奉右莰醇真实世界用药特征，探讨48 h内不同启动时间与住院期间神经功能改善及短期预后的关系。方法：回顾性连续筛选2023年12月1日至2026年4月30日本院诊断为AIS并使用依达拉奉右莰醇的住院患者，药学部基于药学信息系统、住院医嘱及病历记录提取资料。初筛147例，排除35例，纳入112例发病至首次用药时间（onset-to-treatment time，OTT）≤48 h者；按预设24 h界值分为24 h内用药组（n=67）和24～48 h用药组（n=45）。主要结局为住院期间美国国立卫生研究院卒中量表（National Institutes of Health Stroke Scale，NIHSS）变化值（ΔNIHSS=入院NIHSS评分－出院前NIHSS评分）；次要结局包括出院前NIHSS评分、出院改良Rankin量表（modified Rankin Scale，mRS）评分、显著神经功能改善、住院时间、出院去向及安全性事件。采用多因素线性回归分析24 h内用药与ΔNIHSS的相关性，并行简化模型、排除再灌注治疗患者、完整疗程人群、进一步调整大血管闭塞及OTT连续变量模型等敏感性分析。结果：总体OTT为19.65（14.97，34.68）h，疗程12.00（11.00，13.00）d，完成相对完整疗程比例89.3%。两组入院NIHSS评分差异无统计学意义［8.00（6.00，9.00）分 vs 8.00（6.00，10.00）分，P=0.447］；24 h内用药组出院前NIHSS评分更低［5.00（4.00，6.50）分 vs 6.00（5.00，8.00）分，P=0.025］，ΔNIHSS更高［3.00（2.00，3.00）分 vs 2.00（1.00，2.00）分，P＜0.001］，显著神经功能改善率更高（77.6% vs 60.0%，P=0.045）。多因素校正后，24 h内用药仍与更大的ΔNIHSS相关（β=0.768，95%CI：0.377～1.159，P＜0.001）；Logistic探索性分析显示其与显著神经功能改善发生可能性较高相关（OR=2.475，95%CI：1.047～5.853，P=0.039）。两组出血转化、症状性颅内出血、药物相关不良反应及院内死亡差异均无统计学意义。结论：本单中心真实世界队列中，发病48 h内依达拉奉右莰醇治疗疗程完成比例较高、短期安全性事件发生率低；24 h内启动治疗与住院期间NIHSS改善幅度较大及显著神经功能改善率较高相关。由于为回顾性观察性研究，结果应解释为关联性证据，需前瞻性、多中心研究及长期功能结局验证。
Objective: To describe real-world treatment characteristics of edaravone dexborneol in acute ischemic stroke (AIS) patients treated within 48 hours of onset and to explore the association between initiation time and short-term in-hospital outcomes. Methods: This single-center retrospective real-world study screened hospitalized AIS patients with edaravone dexborneol records from December 1, 2023 to April 30, 2026. Medication data were extracted from the pharmacy information system, inpatient orders, and medical records. After 35 exclusions, 112 of 147 patients with onset-to-treatment time (OTT) ≤48 h were included and classified by a prespecified 24-hour cutoff into a within-24-hour group (n=67) and a 24-48-hour group (n=45). The primary outcome was in-hospital National Institutes of Health Stroke Scale (NIHSS) change (ΔNIHSS=admission NIHSS minus pre-discharge NIHSS); secondary outcomes included pre-discharge NIHSS, discharge modified Rankin Scale (mRS), marked neurological improvement, length of stay, discharge destination, and safety events. Multivariable linear regression assessed the association between within-24-hour treatment and ΔNIHSS, with sensitivity analyses using simplified adjustment, exclusion of reperfusion-treated patients, the complete-course population, additional adjustment for large-vessel occlusion, and continuous OTT modeling. Results: Median OTT was 19.65 (14.97, 34.68) h, treatment duration was 12.00 (11.00, 13.00) days, and 89.3% completed a relatively complete course. Baseline NIHSS was comparable [8.00 (6.00, 9.00) vs 8.00 (6.00, 10.00), P=0.447]. The within-24-hour group had lower pre-discharge NIHSS [5.00 (4.00, 6.50) vs 6.00 (5.00, 8.00), P=0.025], greater ΔNIHSS [3.00 (2.00, 3.00) vs 2.00 (1.00, 2.00), P<0.001], and a higher marked improvement rate (77.6% vs 60.0%, P=0.045). After adjustment for age, admission NIHSS, reperfusion therapy, and atrial fibrillation, within-24-hour treatment remained associated with greater ΔNIHSS (β=0.768, 95% CI: 0.377-1.159, P<0.001). Exploratory logistic regression showed a higher likelihood of marked improvement (OR=2.475, 95% CI: 1.047-5.853, P=0.039). Hemorrhagic transformation, symptomatic intracranial hemorrhage, drug-related adverse reactions, and in-hospital death did not differ significantly. Conclusion: In this real-world cohort, edaravone dexborneol within 48 h of AIS onset showed a high complete-course proportion and low short-term safety event incidence. Initiation within 24 h was associated with greater in-hospital NIHSS improvement and a higher marked improvement rate than initiation at 24-48 h. These findings are associative and need prospective multicenter validation with long-term functional outcomes.
目的 探讨急性心肌梗死（AMI）患者经皮冠状动脉介入（PCI）术后6个月内预后不良的影响因素及术前血清髓过氧化物酶（MPO）、淀粉样蛋白A（SAA）、肌酸激酶同工酶（CK-MB）水平联合检测对预后不良的预测效能。方法 前瞻性选取2023年1月～2025年1月许昌市人民医院诊治的204例AMI患者作为AMI组，另选取同期102例健康志愿者作为对照组。比较两组血清MPO、SAA、CK-MB水平。AMI组患者予以PCI术治疗，依据PCI术后6个月内（失访8例）预后情况将分为预后不良亚组（42例）和预后良好亚组（154例），比较不同预后AMI患者临床资料及术前血清MPO、SAA、CK-MB水平。分析AMI患者PCI术后6个月内预后不良的影响因素；分析术前血清MPO、SAA、CK-MB水平联合检测对预后不良的预测效能。结果 AMI组血清MPO、SAA、CK-MB水平高于对照组（P＜0.05）；预后不良亚组多支病变占比、Killip分级Ⅲ级占比、支架置入数、术前血清MPO、SAA、CK-MB水平高于预后良好亚组（P＜0.05）；校正病变支数、Killip分级、支架置入数后，术前血清MPO、SAA、CK-MB水平是AMI患者PCI术后6个月内预后不良的独立影响因素（P＜0.05）；术前血清MPO、SAA、CK-MB联合预测预后不良的AUC值明显高于各血清指标单度指标预测（P＜0.05）。结论 AMI患者血清MPO、SAA、CK-MB水平明显升高，且是AMI患者PCI术后预后不良的独立影响因素，联合检测其水平对预后不良具有较高的预测效能。
【摘要】 目的 探讨基于峰值呼气流速（peak expiratory flow rate，PEFR）的呼吸性肌肉力量减低对I-IIIA期非小细胞肺癌（non-small cell lung cancer，NSCLC）患者术后预后的影响。方法 回顾性分析我院2020年1月-2025年11月接受根治性手术切除的I-IIIA期NSCLC患者临床及影像资料，包括基于肺功能的PEFR（呼吸性肌肉力量指标）及胸部CT的胸肌质量指数（pectoralis muscle index, PMI）。分别采用Jonckheere-Terpstra检验、Spearman’s相关分析比较PEFR与PMI随年龄的变化规律及二者的相关性。低PEFR定义为小于PEFR的性别特异性下四分位数，进一步采用单、多因素Cox回归分析探讨PEFR及PMI对NSCLC患者术后结局的影响。结果 共纳入102例患者，中位年龄62岁（53-67岁），男性65例（63.7%），低低PEFR组24例（23.5%）。低PEFR组在年龄、FEVI、DLCO、FEV1/FVC、FVC、血清白蛋白及随访时间等方面均与正常组间存在显著差异（P＜0.05）。在男、女性患者中，PEFR均表现为随年龄增长逐渐下降的趋势；且与PMI具有较好的相关性（r=0.25，P=0.001）。单因素及多因素Cox回归分析显示，低PEFR是影响NSCLC患者术后无进展生存期（progression free survival, PFS）的独立危险因素（HR=1.57，95%CI：1.03-2.39；P=0.036）结论 呼吸性肌肉力量减低是NSCLC患者术后PFS的独立危险因素，有望成为NSCLC术后复发的早期生物学标志物。
【Abstract】 Objective To investigate the impact of reduced respiratory muscle strength, assessed by peak expiratory flow rate (PEFR), on postoperative outcomes in patients with stage I-IIIA non-small cell lung cancer (NSCLC). Methods Clinical and imaging data of patients with stage I-IIIA NSCLC who underwent radical resection at our hospital from January 2020 to November 2025 were retrospectively analyzed, including PEFR (an indicator of respiratory muscle strength) based on pulmonary function tests and the pectoralis muscle index (PMI) derived from chest CT. The Jonckheere-Terpstra test and Spearman’s correlation analysis were used to evaluate age-related changes in PEFR and PMI and their correlation, respectively. Low PEFR was defined as values below the sex-specific lower quartile of PEFR. Univariate and multivariate Cox regression analyses were performed to assess the impact of PEFR and PMI on postoperative prognosis in NSCLC patients.Results A total of 102 patients were enrolled, with a median age of 62 years (range 53-67 years); 65 patients (63.7%) were male, and 24 (23.5%) were classified into the low PEFR group. The low PEFR group showed significant differences from the normal PEFR group in age, FEV1, DLCO, FEV1/FVC, FVC, serum albumin, and follow-up duration (all P < 0.05). In both male and female patients, PEFR progressively decreased with age and was positively correlated with PMI (r = 0.25, P = 0.001). Univariate and multivariate Cox regression analyses identified low PEFR as an independent risk factor for postoperative progression-free survival (PFS) in NSCLC patients (HR = 1.57, 95% CI: 1.03–2.39; P = 0.036).Conclusion Reduced respiratory muscle strength is an independent risk factor for postoperative PFS in NSCLC patients and may serve as an early biomarker for postoperative recurrence.
目的：分析急性有机磷农药中毒（AOPP）引发缺血缺氧性脑病预后相关因素，建立相关的预后预测模型。方法：回顾性分析90例（33例预后不良、57例预后良好）AOPP致HIE患者（2022年3月～2025年8月）的临床资料、中毒指标和血清学指标，独立危险因素用Logistic回顾分析筛选，并构建预后不良预测模型，采用ROC工具对模型效能进行验证。结果：Logistic 回归分析显示，年龄≥60岁、重度中毒、中毒至就诊时间、LAC水平、CHE水平、CRP水平及NSE水平均为患者预后不良的独立危险因素（P＜0.05）；AUC、灵敏度、特异度为0.943、90.91%、87.72%。结论：高龄、中毒程度高及中毒至就诊时间长等因素可导致AOPP致HIE患者出现不良结局，据此构建风险预测模型可有效预测预后不良的发生风险。
To determine the key impacting factors for hypoxic ischemic encephalopathy (HIE) caused by acute organophosphorus pesticide poisoning (AOPP) and build a prediction model. Methods: The clinical data, poisoning indicators and serological indicators of 90 patients (33 cases with poor prognosis and 57 cases with good prognosis) with HIE caused by AOPP (from March 2022 to Aug 2025) were analyzed. Independent risk factors were screened using logistic retrospective analysis, and a poor prognosis prediction model was constructed. The model efficiency was verified by the receiver operating curve (ROC). Results: Logistic regression analysis showed that age ≥ 60 years, severe poisoning, time from poisoning to treatment, LAC level, CHE level, CRP level, and NSE level were all risk factors for the prognosis in patients (P < 0.05). The AUC, sensitivity, and specificity were 0.943, 90.91%, and 87.72%.Conclusion: Factors such as advanced age, high degree of poisoning, and long time from poisoning to treatment can lead to adverse outcomes in patients with HIE caused by AOPP. Based on this, building a risk prediction model can effectively predict the risk of poor prognosis.
虽然高效抗逆转录病毒治疗已经把艾滋病变成了能够长期管理的慢性传染病，可是患者预后的个体差异比较明显。近些年不良环境暴露被发现是影响艾滋病生存质量、免疫重建、疾病进展的重要外源性因素。环境污染物作为广泛存在而且有可预防的外部风险因素，有着低剂量、长时程、多途径暴露的特点，能够通过免疫毒性、氧化应激、慢性炎症激活、代谢紊乱等多种通路，干扰艾滋病患者的免疫重建进程和病毒抑制效果，进而影响其疾病进展、远期生存结局。本文系统综述了大气污染物、重金属、黄曲霉素、多环芳烃等典型环境污染物与艾滋病患者发病进展及预后转归的关联，深入分析人群易感性差异和当前研究存在的局限，可为优化HIV/AIDS患者的健康管理策略、降低环境相关健康风险、改善患者远期预后给予理论参考和实践依据。
Although highly active antiretroviral therapy has turned AIDS into a chronic infectious disease that can be managed for a long time, the individual differences in the prognosis of patients are obvious. In recent years, adverse environmental exposure has been found to be an important exogenous factor affecting the quality of life, immune reconstruction and disease progression of AIDS. As an external risk factor that exists and has potential intervention, environmental pollutants have the characteristics of low-dose, long-term and multi-channel exposure. They can interfere with the immune reconstruction process and virus inhibition effect of AIDS patients through various pathways such as immunotoxicity, oxidative stress, chronic inflammation activation, and metabolic disorders, thereby affecting their disease progression and long-term survival outcomes. This article systematically reviews the association between typical environmental pollutants such as air pollutants, heavy metals, aflatoxins, and polycyclic aromatic hydrocarbons and the progression and prognosis of AIDS patients. In-depth analysis of the differences in population susceptibility and the limitations of current research can provide theoretical reference and practical basis for optimizing the health management strategies of HIV / AIDS patients, reducing environmental-related health risks, and improving the long-term prognosis of patients.
【摘要】目的：基于潜类别增长模型（LCGM）探讨脑出血患者神经功能恢复轨迹及不同轨迹对预后的影响。方法：回顾性采集360例自发性脑出血患者（2023年6月～2025年6月）的临床资料及神经功能评分[美国国立卫生研究院卒中量表（NIHSS）]，并采用LCGM识别神经功能恢复轨迹的潜在类别，分析影响恢复不良型轨迹的危险因素，对比不同轨迹的预后情况[改良Rankin量表（mRS）、格拉斯哥预后评分（GOS）]。结果：LCGM模型拟合结果显示，3类轨迹为最优拟合模型，可将360例自发性脑出血患者分为快速恢复型139例（38.61%）、稳定恢复型154例（42.78%）、恢复不良型67例（18.61%）；入院格拉斯哥昏迷量表（GCS）评分、初始NIHSS评分、机械通气、血管活性药物使用及血肿体积是神经功能恢复不良的独立影响因素（P＜0.05）；预后方面，三组患者mRS、GOS评分存在显著差异（P＜0.05）。结论：基于LCGM可有效识别脑出血患者神经功能恢复的异质性轨迹，同时还能明确影响患者神经功能修复的独立危险因素及不同神经功能恢复轨迹与预后的关联。
Abstract Objective: To explore the trajectory of neurological recovery in patients with cerebral hemorrhage and the impact of different trajectories on prognosis based on latent class growth model (LCGM). Methods: The clinical data and neurological function scores [National Institutes of Health Stroke Scale (NIHSS)] of 360 patients with spontaneous cerebral hemorrhage (June 2023 to June 2025) were retrospectively collected, and LCGM was used to identify potential categories of neurological recovery trajectories, analyze risk factors affecting poor recovery trajectories, and compare the prognosis of different trajectories [modified Rankin Scale (mRS), Glasgow Outcome Score (GOS)]. Results: The LCGM model fitting results showed that the three types of trajectories were the optimal fitting model, and 360 patients with spontaneous cerebral hemorrhage could be divided into 139 cases (38.61%) of rapid recovery type, 154 cases (42.78%) of stable recovery type, and 67 cases (18.61%) of poor recovery type; admission to Glasgow Coma Scale (GCS) score, initial NIHSS score, mechanical ventilation, use of vasoactive drugs and hematoma volume are independent influencing factors of poor neurological recovery (P＜0.05); in terms of prognosis, there were significant differences in mRS and GOS scores among the three groups of patients (P＜0.05).Conclusion: Based on LCGM, it is possible to effectively identify the heterogeneous trajectories of neurological function recovery in patients with intracerebral hemorrhage (ICH), while also identifying independent risk factors influencing neurological function repair and establishing associations between different recovery trajectories and prognosis.
【摘要】目的：探讨营养控制状态（CONUT）评分与急性心肌梗死（AMI）患者长期预后的相关性。方法：纳入90例AMI患者，于2021年1月～2023年12月入院，随访2年，分为主要心血管不良事件（MACEs）组（32例）与非MACEs组（58例），回顾性分析并对比两组基线资料，并分析CONUT评分与MACEs的相关性及MACEs的影响因素，评估CONUT评分对MACEs的预测效能。结果：90例AMI患者2年MACEs发生率为35.56%；相较于非MACEs组，MACEs组年龄、糖尿病、血肌酐、C反应蛋白（CRP）、降钙素原（PCT）、N末端脑钠肽前体（NT-proBNP）、KillipⅢ~Ⅳ级、CONUT评分更高，白蛋白、血钙、左心室射血分数（LVEF）更低（P＜0.05）；CONUT评分与MACEs发生正相关（P＜0.05）；高CONUT评分是MACEs的独立危险因素（P＜0.05）；MACEs预测中，CONUT评分的灵敏度为93.75%，特异度为93.10%，曲线下面积（AUC）为0.854。结论：CONUT评分与AMI患者长期预后密切相关，营养不良程度越重，MACEs发生风险越高。
Abstract Objective: To explore the correlation between nutritional control status (CONUT) score and long-term prognosis of patients with acute myocardial infarction (AMI). Methods: 90 AMI patients were included, admitted to the hospital from January 2021 to December 2023, and followed up for 2 years. They were divided into major adverse cardiovascular events (MACEs) group (32 cases) and non-MACEs group (58 cases). The baseline data of the two groups were retrospectively analyzed and compared, and the correlation between CONUT score and MACEs and influencing factors of MACEs were analyzed to evaluate the predictive efficacy of CONUT score for MACEs. Results: The 2-year incidence rate of MACEs in 90 AMI patients was 35.56%; compared with the non-MACEs group, the MACEs group had higher age, diabetes, serum creatinine, C-reactive protein (CRP), procalcitonin (PCT), N-terminal pro-brain natriuretic peptide (NT-proBNP), Killip grades III to IV, and CONUT score. Albumin, blood calcium, and left ventricular ejection fraction (LVEF) were lower (P<0.05); CONUT score was positively correlated with the occurrence of MACEs (P<0.05); high CONUT score was an independent risk factor for MACEs (P<0.05); in the prediction of MACEs, the sensitivity of CONUT score was 93.75 %, the specificity was 93.10 %, and the area under the curve ( AUC ) was 0.854. Conclusion: CONUT score is closely related to the long-term prognosis of AMI patients. The more severe the malnutrition, the higher the risk of MACEs.
目的 探讨血清白蛋白（ALB）、纤维蛋白原（FIB）、乳酸脱氢酶（LDH）水平联合检测对急性心肌梗死患者短期预后的预测价值，以期为临床早期制定相应干预方案、改善患者预后提供参考。 方法 回顾性选取我院2023年1月~2025年1月收治的132例急性心肌梗死患者，入院后均接受经皮冠脉介入术（PCI）术治疗，根据术后6个月是否发生心血管主要不良事件（MACE）分为预后不良组（n=34）、预后良好组（n=98），比较两组临床资料及入院时血清ALB、FIB、LDH水平，Logistic回归方程分析血清ALB、FIB、LDH水平与急性心肌梗死患者PCI术后预后的关系，受试者工作特征（ROC）曲线分析其预测价值。 结果 两组病变血管支数、Killip分级及入院时血清ALB、FIB、LDH水平比较，差异具有统计学意义（P＜0.05）；入院时血清ALB、FIB、LDH水平与PCI术后预后显著相关，均为其独立影响因素（P＜0.05）；绘制ROC曲线结果显示，入院时血清ALB、FIB、LDH对于PCI术后预后不良的预测AUC分别为0.795、0.856、0.832，联合预测预后不良的AUC为0.927（95%CI：0.892~0.963），大于各方案单独预测，敏感度为91.18%，特异度为89.80%；危险度分析显示，血清ALB低水平患者预后不良风险是高水平患者的6.127倍，血清FIB、LDH高水平患者预后不良风险是低水平患者的10.493、8.170倍（P＜0.05）。 结论 血清ALB、FIB、LDH水平联合检测对急性心肌梗死患者PCI术后预后具有较高预测价值，临床可根据其早期评估患者预后不良风险，以制定个体化干预方案，改善患者预后。
目的 探讨症状性颅内动脉重度狭窄（sICAS）患者接受自膨式支架成形术后预后不良的相关因素，并构建预测模型。方法 回顾性连续纳入2023年1月至2025年6月于本院脑血管病科行自膨式支架联合经皮腔内血管成形支架置入术（PTAS）的重度sICAS患者96例。收集患者一般临床资料、影像学特征、手术相关资料及随访结局。以术后1年内发生主要终点事件（包括缺血性卒中复发、颅内出血、死亡或症状性支架内再狭窄）定义为预后不良。采用单因素及多因素Logistic回归分析筛选独立危险因素，并构建列线图预测模型。通过受试者工作特征曲线（ROC）及校准曲线评估模型效能。结果 96例患者中，术后1年共发生预后不良事件22例（22.9%），其中缺血性卒中复发12例（12.5%），症状性支架内再狭窄8例（8.3%），颅内出血2例（2.1%）。多因素Logistic回归分析显示，糖尿病（OR = 3.21，95% CI：1.28～8.05，P = 0.013）、术前狭窄长度≥10 mm（OR = 2.89，95% CI：1.15～7.28，P = 0.024）、Mori C型病变（OR = 4.12，95% CI：1.52～11.16，P = 0.005）及术后即刻残余狭窄率≥20%（OR = 2.67，95% CI：1.06～6.72，P = 0.037）是预后不良的独立危险因素。基于上述因素构建的预测模型AUC为0.84（95% CI：0.76～0.92），校准曲线显示模型一致性良好。结论 糖尿病、长病变、复杂Mori分型及术后残余狭窄率高是自膨式支架成形术后预后不良的独立预测因素，所构建的预测模型具有较好的区分度与校准度，可用于个体化风险评估。
