中国全科医学 ›› 2024, Vol. 27 ›› Issue (11): 1349-1355.DOI: 10.12114/j.issn.1007-9572.2023.0410

• 论著 • 上一篇    下一篇

四种营养评价方法对腰椎退行性疾病术后并发症的预测价值比较研究

曾藜1,2, 王惠惠1, 王晓宇1, 龚渭轶1, 胡纯秋1,*()   

  1. 1.230032 安徽省合肥市,安徽医科大学公共卫生学院
    2.410008 湖南省长沙市,中南大学湘雅医院脊柱外科
  • 收稿日期:2023-05-10 修回日期:2023-09-26 出版日期:2024-04-15 发布日期:2024-01-23
  • 通讯作者: 胡纯秋

  • 作者贡献:曾藜提出研究目标,负责文章的构思与设计、收集数据、撰写论文;王惠惠、王晓宇负责数据录入及统计分析;龚渭轶负责文章的审校;胡纯秋负责文章的质量控制。
  • 基金资助:
    国家自然科学基金资助项目(81402676); 安徽省高等学校科学研究项目(2022AH040094); 2022年安徽医科大学公共卫生与预防医学高峰学科建设项目

Comparison of the Predictive Value of Four Nutritional Evaluation Methods for Postoperative Complications of Lumbar Degenerative Diseases

ZENG Li1,2, WANG Huihui1, WANG Xiaoyu1, GONG Weiyi1, HU Chunqiu1,*()   

  1. 1. School of Public Health, Anhui Medical University, Hefei 230032, China
    2. Spine Surgery Department, Xiangya Hospital of Central South University, Changsha 410008, China
  • Received:2023-05-10 Revised:2023-09-26 Published:2024-04-15 Online:2024-01-23
  • Contact: HU Chunqiu

摘要: 背景 腰椎退行性疾病(LDD)术后并发症的发生与营养不良密切相关,而目前缺乏能有效、客观、全面评估LDD患者营养状态并预测术后并发症发生的营养评价方法。目的 比较术前预后营养指数(PNI)、营养控制状态评分(CONUT)、那不勒斯预后评分(NPS)和高敏改良格拉斯哥预后评分(HS-mGPS)四种营养评价方法对LDD患者术后发生并发症的预测价值,以便制定精准营养干预方案,有效预防并发症的发生。方法 选取2021年12月—2022年7月在中南大学湘雅医院脊柱外科确诊为LDD且接受开放手术的患者201例为研究对象,分别应用PNI、CONUT、NPS及HS-mGPS四种营养评价方法对患者进行术前营养状况评估,观察患者术后至出院并发症发生情况,并根据术后有无并发症发生分为并发症组和非并发症组,分析比较两组患者临床特征资料是否存在差异;采用受试者工作特征(ROC)曲线计算不同营养评价方法的最佳截断值,比较不同评价方法的预测效能,运用MedCalc软件进行ROC曲线下面积(AUC)比较;多因素Logistic回归模型分析不同营养评价方法与术后并发症发生的相关性。结果 术后发生并发症患者有60例,发生率为29.8%。临床特征资料比较发现,并发症组与非并发症组年龄、性别、术中出血量、手术节段、PNI、CONUT及NPS评分比较,差异有统计学意义(P<0.05)。ROC曲线计算PNI、CONUT、NPS、HS-mGPS预测LDD术后发生并发症的最佳截断值分别为47.5、2分、1分、1分,依据截断值将不同营养评价方法进行组内比较,结果显示,低PNI、高CONUT评分和高NPS评分与术后并发症相关(P<0.01),而高HS-mGPS评分与术后并发症无关(P=0.815)。AUC结果显示CONUT与HS-mGPS(Z=4.961,P<0.000 1)、CONUT与NPS(Z=2.534,P=0.011 3)、CONUT与PNI(Z=3.455,P=0.000 6)、HS-mGPS与NPS(Z=3.349,P=0.000 8)、NPS与PNI(Z=2.025,P=0.042 9)AUC两两比较,差异均有统计学意义;HS-mGPS与PNI(Z=1.594,P=0.110 9)AUC两两比较,差异无统计学意义。多因素Logistic回归分析结果显示,CONUT(OR=2.431,95%CI=1.376~4.295,P=0.002)、手术节段(OR=11.924,95%CI=5.491~25.893,P<0.001)是LDD术后发生并发症的独立影响因素。结论 用于预测LDD患者术后发生并发症的四种营养评价方法中,CONUT效能最高,可作为LDD术后并发症发生的独立预测因子。

关键词: 腰椎退行性疾病, 营养评价方法, 预后营养指数, 营养控制状态评分, 那不勒斯预后评分, 高敏改良格拉斯哥预后评分, 术后并发症

Abstract:

Background

The occurrence of postoperative complications in lumbar degenerative disease (LDD) is closely related to malnutrition, and currently there is a lack of nutritional evaluation methods that can effectively, objectively, and comprehensively assess the nutritional status of LDD patients and predict the occurrence of postoperative complications.

Objective

To compare the predictive value of four nutritional evaluation methods of Prognostic Nutritional Index (PNI), Controlling Nutritional Status (CONUT), Naples Prognostic Score (NPS) and High-sensitivity Modified Glasgow Prognostic Score (HS-mGPS) for postoperative complications in patients with LDD, so as to formulate a precise nutritional intervention plan and effectively prevent complications.

Methods

A total of 201 patients diagnosed with LDD and underwent open surgery in Spine Surgery Department of Xiangya Hospital of Central South University from December 2021 to July 2022 were selected as the study subjects, four nutritional evaluation methods of PNI, CONUT, NPS and HS-mGPS were applied to assess the preoperative nutritional status of the patients, respectively, and occurrence of postoperative complications from postoperative period to hospital discharge were observed. The patients were divided into the complication group and non-complication group according to the occurrence of postoperative complications to analyze and compare the difference in clinical characteristics between the two groups; the optimal cut-off values of different nutritional evaluation methods were calculated by plotting the receiver operating characteristic (ROC) curve, and the prediction efficiency of different evaluation methods was compared, the area under curve (AUC) was compared by using MedCalc software. Multivariate Logistic regression model was used to analyze the correlation between different nutritional evaluation methods and postoperative complications.

Results

Postoperative complications occurred in 60 patients with an incidence rate of 29.8%. The comparison of clinical characteristics data showed statistically significant differences in age, gender, intraoperative bleeding, surgical segment, PNI, CONUT, and NPS scores between the complication group and non-complication group (P<0.05). The optimal cut-off values of PNI, CONUT, NPS, and HS-mGPS for predicting postoperative complications of LDD were calculated by using ROC curve were 47.5, 2 scores, 1 score, and 1 score, respectively, and the different nutritional evaluation methods were compared within the group based on the cut-off values, showing that low PNI score, high CONUT score, and high NPS score were associated with postoperative complications (P<0.01), while high HS-mGPS score was not associated with postoperative complications (P=0.815). The comparison of AUC showed statistically significant differences between CONUT and HS-mGPS (Z=4.961, P<0.000 1), CONUT and NPS (Z=2.534, P=0.011 3), CONUT and PNI (Z=3.455, P=0.000 6), HS-mGPS and NPS (Z=3.349, P=0.000 8), NPS and PNI (Z=2.025, P=0.042 9) ; there was no statistically significant difference between HS-mGPS and PNI (Z=1.594, P=0.110 9). Multivariate Logistic regression analysis showed that CONUT (OR=2.431, 95%CI=1.376-4.295, P=0.002), surgical segment (OR=11.924, 95%CI=5.491-25.893, P<0.001) were independent influencing factors for postoperative complications after LDD.

Conclusion

Among the four nutritional evaluation methods used to predict postoperative complications of LDD patients, CONUT score is the most effective and can be used as an independent predictor of postoperative complications of LDD.

Key words: Lumbar degenerative disease, Nutritional evaluation method, Prognostic Nutritional Index, Controlling Nutritional Status, Naples Prognostic Score, High-sensitive Modified Glasgow Prognostic Score, Postoperative complications