目的：通过全自动尿液分析仪细菌通道参数与细菌培养的分析比较，从而建立一种快速准确的儿童尿路感染初筛方法。方法选择北京儿童医院2015年1至5月疑似尿路感染患儿中段尿649例，同时进行细菌培养和UF-1000i尿沉渣定量分析。将尿液细菌培养结果作为金标准，与尿液分析仪通道细菌参数[前向散射光（ B-FSC），荧光强度（ B-FLH）]进行分析比较，通过ROC曲线分析，计算出诊断儿童尿路感染的最佳临界值。结果①624例患儿标本中细菌培养阳性220例，阳性率35．3％，共分离细菌238株，其中革兰阴性菌153株，革兰阳性菌73株，真菌12株；②建立尿液分析仪白细胞和细菌参数的最佳临界值43／L和474．4／L，其诊断尿路感染的灵敏度76．6％、83．3％，特异性75．5％、79．2％，约登指数51．2％、62．5％，漏诊率23．4％、16．7％，误诊率25．4％、20．8％，阳性预测值64．5％、71．4％，阴性预测值74．6％、88．4％，阳性似然比3．12、4．00，阴性似然比0．31、0．21；③革兰阴性菌B-FSC均值与革兰阳性菌均值差异均有显著性意义（t＝2．183，P＜0．05）。 B-FLH在区别革兰阴性菌和革兰阳性菌差异无统计学意义（P＞0．05）；④以B-FSC＜37ch和B-FSC／B-FLH＜0．4为临界值，判断革兰阴性菌的灵敏度分别为85．5％、91．6％，特异性86．4％、89．8％，约登指数71．9％、81．4％，漏诊率14．5％、8．4％，误诊率13．6％、10．2％，阳性预测值93．3％、95．2％，阴性预测值72．9％、82．8％，阳性似然比6．29、8．98，阴性似然比0．17、0．09。结论在儿童尿路感染中，细菌参数B-FSC能有效地区别革兰阴性菌和革兰阳性菌，B-FSC／B-FLH＜0．4区别革兰阴性菌和革兰阳性菌优于B-FSC＜37ch。建立实验室最佳临界值可帮助快速诊断儿童尿路感染及鉴别细菌类型。%Objective To establish a rapid and accurate method for preliminary screening of pediatric urinary tract infection (UTI) by comparing the bacteria channel parameters of automated urine analyzer ( AUA) and results of bacteria culture .Methods Midstream urines of 649 cases of children with suspected UTI were collected from Beijing Children ’s Hospital from January to May in 2015.Bacteria culture of the urine and quantitative analysis of urine sediment were performed simultaneously .Results of bacteria culture were set as the golden standard and were compared with bacteria channel parameters ( B-FSC and B-FLH) obtained by Urine Analyzer .The optimal threshold value for the diagnosis of UTI in children was calculated through receiver operating characteristic (ROC) curve analysis.Results Among 624 cases of samples, positive results of bacterial culture were found in 220 samples (35.3%).In total, 238 bacteria strains were isolated, including 153 strains of Gram-negative bacteria, 73 strains of Gram-positive bacteria and 12 strains of fungi.With the aid of AUA, results showed that the optimal threshold value of white blood cells and bacterial parameter was 43/l and 474.4/l, respectively. Their sensitivity for the diagnosis of UTI was 76.6% and 83.3%, specificity was 75.5% and 79.2%, Youden index was 51.2% and 62.5%, respectively.In addition, the rate of missed diagnosis was 23.4% and 16.7%, misdiagnosis rate was 25.4% and 20.8%, positive predictive value was 64.5%and 71.4%, respectively, negative predictive value was 74.6%and 88.4%, positive likelihood ratio was 3.12 and 4.00, and negative likelihood ratio was 0.31 and 0.21, respectively.Significant difference was identified between Gram-positive bacteria and Gram-negative bacteria in the mean value of B-FSC (t=2.183, P<0.05).No statistically significant difference was noted between Gram-positive bacteria and Gram-negative bacteria in B-FLH (P>0.05).When setting B-FSC<37ch and B-FSC/B-FLH<0.4 as threshold values , the sensitivity for Gram-negative bacteria was 85.5% and 91.6%, specificity was 86.4% and 89.8%, Youden index was 71.9% and 81.4%, rate of missed diagnosis was 14.5% and 8.4%, misdiagnosis rate was 13.6% and 10.2%, positive predictive value was 93.3%and 95.2%, negative predictive value was 72.9% and 82.8%, positive likelihood ratio was 6.29 and 8.98, and negative likelihood ratio was 0.17 and 0.09, respectively.Conclusion The bacterial parameter B-FSC can effectively differentiate Gram-negative bacteria from Gram-positive bacteria in children with UTI .The threshold value B-FSC/B-FLH<0.4 is more sensitive than B-FSC<37ch in the differentiation of Gram-negative bacteria from Gram-positive bacteria.Setting up a laboratory optimal threshold is helpful to the quick diagnosis of UTI in children and the identification of the types of bacteria .
展开▼
