目的 探讨改良版患者主观整体评估( mPG-SGA)、患者主观整体评估( PG-SGA) 与全球领导人营养不良倡议 (GLIM)3 种评估工具评定消化系统肿瘤患者营养不良的能力,并探讨 mPG-SGA 与常用临床检验指标在评定新入院消化系 统肿瘤患者营养不良的相关性。 方法 前瞻性收集于 2025 年 1 月至 4 月在中国人民解放军西部战区总医院肿瘤科入院≤48 h 的消化系统肿瘤患者作为研究对象,收集患者入院后常用血液学检验指标、人口统计学资料,并使用 mPG-SGA、PG-SGA 和 GLIM 工具分别进行营养不良的评估,比较三者营养不良评估的效果,并进一步运用 Spearman 相关性分析比较 mPG-SGA 的 评价结果与营养指标的相关性。 结果 入组的 105 例消化系统肿瘤患者的平均年龄为(58. 50± 13. 94) 岁,其中男性 71 例 (67. 62%)。 mPG-SGA、PG-SGA 和 GLIM 评价/ 诊断营养不良的发生率分别为 40. 00%、43. 81%和 60. 00%。 以 PG-SGA 为参 照,mPG-SGA 和 GLIM 检测营养不良的敏感度分别为 100. 00%和 82. 61%,特异度分别为 93. 65%和 57. 63%,曲线下面积 (AUC)值分别为 0. 957 和 0. 701。 mPG-SGA 与 PG-SGA、mPG-SGA 与 GLIM、PG-SGA 与 GLIM 的 Kappa 系数分别为 0. 922 (P<0. 001)、0. 432(P<0. 001)和 0. 387(P<0. 001)。 相关性结果显示淋巴细胞总数、血红蛋白和骨骼肌指数与 mPG-SGA 呈正 相关,相关系数分别为 0. 246 (P= 0. 011)、0. 226(P= 0. 020)、0. 373(P<0. 001)。 结论 mPG-SGA 评估消化系统肿瘤患者营养 不良发生率与 PG-SGA 和 GLIM 相近,且与部分营养指标呈现相关性。

Objective This study aimed to investigate the comparative efficacy of mPG - SGA PG - SGA and GLIM tools in assessing malnutrition among digestive system tumor patients and analyze correlations between mPG-SGA and hematological parameters in newly admitted patients. Method 105 digestive system tumor patients admitted within ≤48 hours to the department of oncology of our hospital between January and April 2025 were enrolled. Demographic data and hematological parameters were collected routinely. Nutritional status was assessed using mPG-SGA PG-SGA and GLIM criteria. Comparative efficacy was analyzed through Spearman correlation between mPG-SGA scores and nutritional biomarkers. Result The mean age of all subjects in this cohort n = 105 was 58. 50±13. 94 years old with male accounted for 67. 62% 71 cases . Prevalence of malnutrition detected by 3 tools was 40. 00% mPG-SGA 43. 81% PG-SGA and 60. 00% GLIM . Using PG-SGA as reference mPG-SGA showed superior sensitivity 100. 00% vs 82. 61% and specificity 93. 65% vs 57. 63% with higher AUC 0. 957 vs 0. 701 . The Kappa coefficients of mPG-SGA vs PG-SGA mPG-SGA vs GLIM and PG-SGA vs GLIM were κ = 0. 922 κ = 0. 432 and κ = 0. 387 respectively all P <0. 001 . Significant positive correlations emerged between mPG-SGA and lymphocytes r = 0. 246 P = 0. 011 hemoglobin r = 0. 226 P= 0. 020 and skeletal muscle index r = 0. 373 P<0. 001 . Conclusion mPG-SGA demonstrated comparable efficacy to PG-SGA and GLIM in malnutrition assessment with correlation with some nutrition indexes.