标题：术前GNRI预测老年胰十二指肠切除术患者术后并发症的价值

      作者：高灵花 1，陈颖 1，沈金花 2，俞华 3    1.复旦大学附属华山医院胰腺外科，上海 200040；2.复旦大学附属华山医院危重症室，上海 200040；3.上海市静安区闸北中心医院胰腺外科，上海 200070

      卷次： 2023年34卷19期

      【摘要】 目的 探究术前老年人营养风险指数(GNRI)预测老年胰十二指肠切除术(PD)患者术后并发症的价值。方法 回顾性分析2019年5月至2021年5月复旦大学附属华山医院收治的93例老年PD术患者的临床资料，根据术后并发症发生情况分组，其中57例术后未发生并发症患者纳入对照组，36例发生并发症患者纳入并发症组，比较两组患者的临床资料及GNRI，采用Logistic回归分析确定老年PD患者术后并发症发生的危险因素并采用受试者工作曲线(ROC)分析术前GNRI对术后并发症的预测价值。结果 并发症组患者的BMI为(21.73±0.95) kg/m2，明显低于对照组的(22.58±1.11) kg/m2，有糖尿病史患者占比为 66.67%，明显多于对照组的 42.11%，HSA水平为(34.15±5.49) g/L，明显低于对照组的(37.03±3.28) g/L，手术时间为(5.56±1.58) h，明显长于对照组的(4.83±1.46) h，GNRI为91.08±8.44，明显低于对照组的98.15±7.03，差异均有统计学意义(P<0.05)；经多因素Logistic回归分析结果显示，GNRI是患者术后并发症发生的危险因素(P<0.05)；经ROC分析结果显示：GNRI预测老年PD患者术后发生并发症的曲线下面积(AUC)值为0.702，敏感度、特异度分别为58.33%、75.44%。结论 GNRI为老年PD术患者术后并发症发生的危险因素且预测价值良好，医师可根据术前GNRI评估患者营养状态并及时采取针对性干预措施以减少老年患者PD术后并发症发生情况。
      【关键词】 胰十二指肠切除术；老年；老年人营养风险指数；营养风险评估；并发症；预测价值
      【中图分类号】 R656.6+4 【文献标识码】 A 【文章编号】 1003—6350（2023）19—2797—05

Value of preoperative geriatric nutritional risk index in predicting complications in elderly patients afterpancreaticoduodenectomy.
GAO Ling-hua 1, CHEN Ying 1, SHEN Jin-hua 2, YU Hua 3. 1. Department of PancreaticSurgery, Huashan Hospital, Fudan University, Shanghai 200040, CHINA; 2. Intensive Care Unit, Huashan Hospital, FudanUniversity, Shanghai 200040, CHINA; 3. Department of Pancreatic Surgery, Shanghai Jing'an District Zhabei CentralHospital, Shanghai 200070, CHINA
【Abstract】 Objective To explore the value of preoperative geriatric nutritional risk index (GNRI) in predictingcomplications in elderly patients after pancreaticoduodenectomy (PD). Methods The clinical data of 93 elderly pa-tients who underwent PD in Huashan Hospital, Fudan University between May 2019 and May 2021 were analyzed retro-spectively. According to the presence or absence of postoperative complications, the patients were divided in the controlgroup (57 patients without complications) and the complication group (36 patients with complications). Clinical data andGNRI of the two groups were compared. Logistic regression analysis was conducted to identify the risk factors for post-operative complications, and the predictive value of preoperative GNRI for postoperative complications was analyzed us-ing the receiver operating characteristic (ROC) curve. Results Body mass index (BMI) of the complication group was(21.73±0.95) kg/m2, significantly lower than (22.58±1.11) kg/m2 of the control group; and the proportion of patients withdiabetes history was 66.67%, significantly higher than 42.11% in the control group; human serum albumin (HSA) levelin the complication group was (34.15±5.49) g/L, significantly lower than (37.03±3.28) g/L in the control group; the oper-ation time was (5.56±1.58) h, significantly longer than (4.83±1.46) h of the control group; GNRI was 91.08±8.44, signif-icantly lower than 98.15±7.03 of the control group; the differences were statistically significant (P<0.05). Multivariate lo-gistic regression analysis showed that GNRI was the risk factor for postoperative complications (P<0.05). ROC curveanalysis showed that the area under the curve (AUC) value, sensitivity, and specificity of GNRI for predicting complica-tions in elderly patients after PD were 0.702, 58.33%, and 75.44%. Conclusion GNRI is a risk factor for complicationsin elderly patients after PD, with good predictive value. Physicians can evaluate nutritional status of patients accordingto preoperative GNRI and take timely targeted intervention measures to reduce postoperative complications.
      【Key words】 Pancreaticoduodenectomy; Elderly; Geriatric nutritional risk index; Nutritional risk assessment;Complication; Predictive value