Tang Wanyun, Yao Wei, Wang Wei, Lv Qiaomei, Ding Wenbo, He RenJian
Department of Orthopedics, Zigong First People's Hospital, Zigong, China.
Department of Orthopedics, Dandong Central Hospital, China Medical University, Dandong, China.
Front Med (Lausanne). 2024 May 30;11:1333472. doi: 10.3389/fmed.2024.1333472. eCollection 2024.
This study aims to discern the significance of common hematological and biochemical parameters for predicting urinary tract infections in geriatric patients with hip fractures.
Multivariable logistic regression and propensity score-matched analyses were conducted to calculate adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for UTIs. The abilities of these parameters to predict UTIs were evaluated by receiver operating characteristic (ROC) curves. Dose-response relationships were assessed by categorizing hematological and biochemical parameters into quartiles. Subgroup analyses were further explored to investigate the relationship between these parameters and urinary tract infections.
Out of the 1,231 participants, 23.2% were diagnosed with UTIs. Hyperglycemia, hypoproteinemia and hyperglobulinemia were risk factors for UTIs in multivariate analysis. After propensity score matching, hyperglycemia (OR 2.14, 95% CI 1.50-3.05, < 0.001), hypoproteinemia (OR 1.75, 95% CI 1.18-2.63, = 0.006), and hyperglobulinemia (OR 1.38, 95% CI 0.97-1.97, = 0.074) remained significantly associated with increased odds of urinary tract infections. ROC curve analyses showed moderate predictive accuracy of blood glucose, albumin and globulin for UTIs, with areas under the curves of 0.714, 0.633, and 0.596, respectively. Significant dose-response relationships were observed between these parameters and UTIs. The associations were consistent in subgroup analyses.
Blood glucose, albumin and globulin levels can facilitate early identification of geriatric hip fracture patients at high risk of UTIs. These easily obtainable hematological and biochemical parameters provide a practical clinical prediction tool for individualized UTI prevention in this population.
本研究旨在探讨常见血液学和生化参数对预测老年髋部骨折患者尿路感染的意义。
进行多变量逻辑回归和倾向评分匹配分析，以计算尿路感染的调整优势比（OR）和95%置信区间（CI）。通过受试者工作特征（ROC）曲线评估这些参数预测尿路感染的能力。通过将血液学和生化参数分为四分位数来评估剂量反应关系。进一步进行亚组分析，以研究这些参数与尿路感染之间的关系。
在1231名参与者中，23.2%被诊断为尿路感染。多变量分析显示，高血糖、低蛋白血症和高球蛋白血症是尿路感染的危险因素。倾向评分匹配后，高血糖（OR 2.14，95%CI 1.50 - 3.05，<0.001）、低蛋白血症（OR 1.75，95%CI 1.18 - 2.63，=0.006）和高球蛋白血症（OR 1.38，95%CI 0.97 - 1.97，=0.074）与尿路感染几率增加仍显著相关。ROC曲线分析显示，血糖、白蛋白和球蛋白对尿路感染的预测准确性中等，曲线下面积分别为0.714、0.633和0.596。这些参数与尿路感染之间存在显著的剂量反应关系。亚组分析中这些关联是一致的。
血糖、白蛋白和球蛋白水平有助于早期识别老年髋部骨折患者中尿路感染的高危人群。这些易于获得的血液学和生化参数为该人群个体化预防尿路感染提供了一种实用的临床预测工具。
