Abbott Jessica, Teleni L, McKavanagh D, Watson J, McCarthy A L, Isenring E
Gold Coast University Hospital, Southport, QLD, 4215, Australia.
Princess Alexandra Hospital, Woolloongabba, QLD, 4102, Australia.
Support Care Cancer. 2016 Sep;24(9):3883-7. doi: 10.1007/s00520-016-3196-0. Epub 2016 Apr 19.
In the oncology population where malnutrition prevalence is high, more descriptive screening tools can provide further information to assist triaging and capture acute change. The Patient-Generated Subjective Global Assessment Short Form (PG-SGA SF) is a component of a nutritional assessment tool which could be used for descriptive nutrition screening. The purpose of this study was to conduct a secondary analysis of nutrition screening and assessment data to identify the most relevant information contributing to the PG-SGA SF to identify malnutrition risk with high sensitivity and specificity.
This was an observational, cross-sectional study of 300 consecutive adult patients receiving ambulatory anti-cancer treatment at an Australian tertiary hospital. Anthropometric and patient descriptive data were collected. The scored PG-SGA generated a score for nutritional risk (PG-SGA SF) and a global rating for nutrition status. Receiver operating characteristic curves (ROC) were generated to determine optimal cut-off scores for combinations of the PG-SGA SF boxes with the greatest sensitivity and specificity for predicting malnutrition according to scored PG-SGA global rating.
The additive scores of boxes 1-3 had the highest sensitivity (90.2 %) while maintaining satisfactory specificity (67.5 %) and demonstrating high diagnostic value (AUC = 0.85, 95 % CI = 0.81-0.89). The inclusion of box 4 (PG-SGA SF) did not add further value as a screening tool (AUC = 0.85, 95 % CI = 0.80-0.89; sensitivity 80.4 %; specificity 72.3 %).
The validity of the PG-SGA SF in chemotherapy outpatients was confirmed. The present study however demonstrated that the functional capacity question (box 4) does not improve the overall discriminatory value of the PG-SGA SF.
在营养不良患病率较高的肿瘤患者群体中，更多具有描述性的筛查工具能够提供更多信息，以辅助进行分诊并捕捉急性变化。患者主观全面评定简表（PG-SGA SF）是一种营养评估工具的组成部分，可用于描述性营养筛查。本研究的目的是对营养筛查和评估数据进行二次分析，以确定对PG-SGA SF最具相关性的信息，从而以高灵敏度和特异性识别营养不良风险。
这是一项对澳大利亚一家三级医院300例接受门诊抗癌治疗的成年患者进行的观察性横断面研究。收集了人体测量和患者描述性数据。计分的PG-SGA得出营养风险评分（PG-SGA SF）和营养状况总体评定。绘制受试者工作特征曲线（ROC），以确定根据计分的PG-SGA总体评定，预测营养不良时具有最高灵敏度和特异性的PG-SGA SF各部分组合的最佳截断分数。
第1 - 3部分的累加分数具有最高灵敏度（90.2%），同时保持了令人满意的特异性（67.5%），并显示出较高的诊断价值（AUC = 0.85，95% CI = 0.81 - 0.89）。纳入第4部分（PG-SGA SF）作为筛查工具并未增加更多价值（AUC = 0.85，95% CI = 0.80 - 0.89；灵敏度80.4%；特异性72.3%）。
PG-SGA SF在化疗门诊患者中的有效性得到了证实。然而，本研究表明功能能力问题（第4部分）并未提高PG-SGA SF的总体鉴别价值。
