The prevalence of hip fractures is increasing due to increased number of aging people and availability of modern transportation. Elderly patients with hip fracture tend to have many basic diseases and poor prognosis, with 1-year postoperative mortality as high as 15%-25%. Timely screening of those at high risk of hip fractures with predictive tools and providing them with interventions can improve the prognosis. But there are rare studies and no appropriate objective indicators regarding predicted 1-year postoperative mortality in older hip fracture patients.

To assess the predictive association of Controlling Nutritional Status (CONUT) score and other possible associated factors for 1-year postoperative mortality in elderly patients with hip fracture.

A total of 399 patients with unilateral hip fracture (including femoral neck and intertrochanteric fractures) hospitalized in Xuanwu Hospital Capital Medical University from January 2013 to March 2016 were selected. General data, routine blood test results, biochemical and coagulation indices, and calculated CONUT score, geriatric nutritional risk index (GNRI) and neutrophil-to-lymphocyte ratio (NLR) were collected. All the patients received internal fixation or femoral head replacement procedure, and 1-year postoperative clinic- and telephone-based follow-up with death as an endpoint. Multivariate Logistic regression was used to identify influencing factors of 1-year postoperative mortality. Receiver operating characteristic (ROC) curve analysis was used to analyze the predictive value of CONUT score for 1-year postoperative mortality.

Among the 399 patients, 47 died and 352 survived at the end of follow-up. Multivariate Logistic regression analysis found that age〔OR=1.093, 95%CI (1.040, 1.148) 〕, old cerebral infarction〔OR=0.353, 95%CI (0.169, 0.737) 〕, serum creatinine〔OR=1.006, 95%CI (1.002, 1.010) 〕, CONUT score〔OR=1.261, 95%CI (1.005, 1.583) 〕 and NLR〔OR=1.049, 95%CI (1.003, 1.098) 〕 were associated with 1-year postoperative mortality (P<0.05). In predicting 1-year postoperative mortality, the area under the curve of CONUT score was 0.681〔95%CI (0.590, 0.771) 〕 with 4.5 as the optimal cutoff point, 48.9% sensitivity and 83.5% specificity. The area under the curve of NLR was 0.611〔95%CI (0.523, 0.699) 〕, with 7.055 μg/L as the optimal cutoff point, and 47.8% sensitivity and 75.9% specificity.

Age, old cerebral infarction, serum creatinine, CONUT score and NLR were associated with 1-year postoperative mortality in elderly patients with hip fracture. Preoperative CONUT score greater than 5 may be associated with worse postoperative nutritional status and higher risk of death. The CONUT score could be used for postoperative risk assessment in these patients.

由于人口老龄化及现代交通工具的应用普及，髋部骨折发病率逐渐升高。老年髋部骨折患者基础疾病多、预后差，术后1年病死率可高达15%~25%。若使用预测工具及时发现高危患者并干预，可改善患者预后，但目前预测老年髋部骨折患者术后1年死亡的研究相对较少，缺乏合适的客观指标预测患者术后1年死亡情况。

探讨术前控制营养状况（CONUT）评分预测老年髋部骨折患者术后1年死亡的价值，并分析患者术后1年死亡的其他相关危险因素。

选取2013年1月至2016年3月于首都医科大学宣武医院住院治疗的一侧髋部（包括股骨颈及股骨粗隆间）骨折患者399例为研究对象。收集患者的一般资料、血常规、生化指标及凝血指标等，并计算术前CONUT评分、老年营养风险指数（GNRI）、中性粒细胞/淋巴细胞比值（NLR）。患者行内固定或人工股骨头置换术，术后对患者进行门诊和电话随访，随访1年，并根据术后1年是否死亡，将患者分为死亡组和存活组。采用多因素Logistic回归分析探讨老年髋部骨折患者术后1年死亡的影响因素，并进一步绘制受试者工作特征（ROC）曲线，分析术前CONUT评分预测老年髋部骨折患者术后1年死亡的价值。

399例患者中，死亡组47例，存活组352例。多因素Logistic回归分析结果显示，年龄〔OR=1.093，95%CI（1.040，1.148）〕、陈旧性脑梗死〔OR=0.353，95%CI（0.169，0.737）〕、肌酐〔OR=1.006，95%CI（1.002，1.010）〕、术前CONUT评分〔OR=1.261，95%CI（1.005，1.583）〕、NLR〔OR=1.049，95%CI（1.003，1.098）〕是老年髋部骨折患者术后1年死亡的影响因素（P<0.05）。术前CONUT评分预测老年髋部骨折患者术后1年死亡的ROC曲线下面积为0.681〔95%CI（0.590，0.771）〕，临界值为4.5分，灵敏度为48.9%，特异度为83.5%；NLR预测老年髋部骨折患者术后1年死亡的ROC曲线下面积为0.611〔95%CI（0.523，0.699）〕，临界值为7.055 μg/L，灵敏度为47.8%，特异度为75.9%。

年龄、陈旧性脑梗死、肌酐、术前CONUT评分、NLR是老年髋部骨折患者术后1年死亡的影响因素。术前CONUT评分≥5分时，患者术后的营养状况更差，死亡风险升高，可用于髋部骨折患者术后风险评估。