侵袭性真菌感染风险评分系统的建立及效能分析

Establishment and effectiveness analysis of a risk evaluation system for invasive fungal infections

目的:建立一个识别侵袭性真菌感染(invasive fungal infection,IFI)高风险人群的风险评分系统,并验证其有效性。方法:对模型组103例病人的病例资料进行单因素及logistic多重回归分析,筛选出IFI的危险因素并分别赋值,建立IFI风险评分系统。运用该评分系统计算41例IFI病人的综合得分,比较各得分组IFI实际发生率的差异,运用接收者工作特征曲线(receiver operating characteristic curve,ROC)评价该评分系统的效能。结果:糖尿病、胃肠道手术、血液系统恶性肿瘤、机械通气治疗>3 d、广谱抗生素治疗>5 d、深静脉导管植入是IFI的危险因素。模型组和验证组中得分越高的病例发生IFI的几率越大,且高、中、低得分组的IFI发生率的差异有统计学意义(P<0.05)。模型组ROC曲线下面积为0.816,验证组ROC曲线下面积为0.811。结论:该评分系统能够有针对性地帮助临床医师识别IFI高风险人群,指导对病人的及时治疗。

Objective： To establish a risk scoring system for identifying patients in high risk of invasive fungal infection（IFI） and analyze its effectiveness.Methods： For the clinical data of 103 patients in the model group,univariate and multivariate logistic regression were used to identify risk factors of IFI.The risk factors were all scored and the IFI risk scoring system（IFIRSS） was established.IFIRSS was used to evaluate the scores of 41 patients in validation group.The evaluation result and actual incidence of IFI in each group were compared.The effectiveness of IFIRSS was analyzed by the receiver operating characteristic curve（ROC）.Results： Diabetes mellitus,gastrointestinal operations,hematological malignant tumors,mechanical ventilation longer than 3 days,broad-spectrum antibiotic therapy longer than 5 days and implantation of deep veins catheter were identified as independent risk factors of IFI.Patients with higher scores in model group and validation group were at higher risk of IFI.The difference of IFI incidence between high,middle,and low score groups was significant（P0.05）.Area under ROC was 0.816 in model group and 0.811 in validation group.Conclusion： IFIRSS can help clinicians to identify patients at high risk of IFI and guide treatment of patients in time.