目的 探讨预后营养指数与尿素/肌酐比值在食管癌放化疗患者营养状态评估中的应用价值研究。 方法 筛选接受放化疗的食管癌患者138例，采用NRS 2002营养筛查为标准，应用患者客观指标的预后营养指数（PNI）及尿素/肌酐比值（UCR）对食管癌放化疗患者营养不良进行评价，分析两种客观性营养筛查工具与各营养指标的相关性，采用ROC曲线评估比较不同营养筛查工具的诊断效能。 结果 食管癌放化疗患者的生化指标和血液学指标呈现一定特征，其中尿素、肌酐、UCR、白蛋白（ALB）、前白蛋白（PALB）、总蛋白（TP）、TC、TG、HDL和LDL水平均在正常范围内波动。在营养风险筛查方面，NRS 2002评分识别出57.97%的患者存在营养风险，而PNI和UCR预测存在营养风险的患者比例分别为31.16%和40.58%。一致性检验显示，PNI与NRS 2002评分的Kappa值为0.460（P < 0.05），UCR与NRS 2002评分的Kappa值为0.522（P < 0.05），表明两者与NRS 2002评分在营养风险筛查上具有一定的相关性。相关性分析表明，NRS 2002评分与UCR、ALB、PALB、PBL、Cr、HDL、TP存在显著相关性（P < 0.05），PNI与ALB、PALB、PBL、肌酐（Cr）、HDL、TP、尿素（UREA）存在显著相关性（P < 0.05），UCR与PALB、UREA、Cr存在显著相关性（P < 0.05）。Logistic单因素回归分析进一步证实了这些相关性，并揭示了营养不良与ALB、PALB、TP、淋巴细胞、Cr、BMI指标成负相关，与UREA呈正相关。然而，多因素回归分析未发现这些相关因素与两种客观评价指标之间存在统计学差异。诊断效能评价结果显示，PNI和UCR评估预测放化疗食管癌营养不良的ROC曲线下面积分别为0.779和0.736，差异有统计学意义（P < 0.001）。其中，UCR显示出较高的灵敏度，而PNI表现出较高的特异度，但两者在诊断价值上无显著差异。 结论 PNI和UCR作为营养评估工具在食管癌放化疗患者的营养风险筛查中具有一定的应用价值，但两者与NRS 2002评分在一致性方面有待提高。在评估患者营养状况时，应结合多种指标和方法，以获得更准确的诊断结果。

Objective To investigate the application value of prognostic nutritional index and urea/creatinine ratio in nutritional status assessment for patients with esophageal cancer undergoing radiotherapy and chemotherapy. Methods A total of 138 patients with esophageal cancer who received radiotherapy and chemotherapy were screened. NRS 2002 nutritional screening was used as the standard. Prognostic nutritional index （PNI） and urea/creatinine ratio （UCR） were used to evaluate the malnutrition of patients with esophageal cancer. ROC curve was used to evaluate and compare the diagnostic efficacy of different nutritional screening tools. Results The biochemical and hematological indexes of patients with esophageal cancer showed certain characteristics， in which the levels of urea， creatinine， UCR， ALB， PALB， TP， TC， TG， HDL and LDL all fluctuated within the normal range. In terms of nutritional risk screening， the NRS 2002 score identified nutritional risk in 57.97% of patients， while the PNI and UCR predicted nutritional risk in 31.16% and 40.58% of patients， respectively. Consistency test showed that the Kappa value of PNI and NRS 2002 scores was 0.460 （P < 0.05）， and the Kappa value of UCR and NRS 2002 scores was 0.522 （P < 0.05）， indicating that there was a certain correlation between them and NRS 2002 scores in nutritional risk screening. Correlation analysis showed that NRS 2002 score was significantly correlated with UCR， ALB， PALB， PBL， Cr， HDL and TP （P < 0.05）， while PNI was significantly correlated with ALB， PALB， PBL， Cr， HDL， TP and UREA （P < 0.05）. There were significant correlations between UCR and PALB， UREA and Cr （P < 0.05）. Logistic univariate regression analysis further confirmed these correlations and revealed that malnutrition was negatively correlated with albumin， prealbumin， total protein， lymphocytes， creatinine and BMI， and positively correlated with urea. However， multivariate regression analysis did not find any statistical difference between these related factors and the two objective evaluation measures. The results of diagnostic efficacy evaluation showed that the area under ROC curve of PNI and UCR evaluation to predict esophageal cancer malnutrition was 0.779 and 0.736， respectively， with statistical significance （P < 0.001）. Among them， UCR showed higher sensitivity and PNI showed higher specificity， but there was no significant difference in diagnostic value between them. Conclusion PNI and UCR as nutritional assessment tools have certain application value in the nutritional risk screening of patients with esophageal cancer， but their consistency with NRS 2002 score needs to be improved. When assessing the nutritional status of patients， a combination of various indicators and methods should be used to achieve more accurate diagnostic results.