評估IRIS合併尿液生化與尿沉渣之全尿液自動化分析系統應用於預測泌尿道感染

Evaluation of IRIS Automated Urine Chemistry and Sediment Analyzers for Total Urinalysis in Prediction of UTI

最近IRIS自動化儀器整合了尿液生化與尿沉渣分析儀器，組成一套完整的自動化尿液分析系統；串聯iChem Velocity生化分析與iQ200 ELITE尿沉渣分析，可以縮短處理時間並且交叉比對檢驗結果，但這套系統的執行成效仍值得進一步確認。本篇研究目標為評估IRIS全尿液分析系統對於尿液常規檢驗的執行，分析系統應用於泌尿道感染的預測及與泌尿道感染致病菌特性的相關性。本研究在成大醫院內連續收樣315位懷疑泌尿道感染的病患，包含132位男性與183位女性（平均年齡：54.3歲），實施回溯性試驗分析尿液常規檢驗數據及尿液培養結果。尿液培養結果為98個泌尿道感染陽性（31.1％）與217個泌尿道感染陰性（68.9％）；女性泌尿道感染比率高於男性約3倍（p＜0.0001）。將iChem與iQ200的檢驗結果進行比較，兩者具有顯著的相關性，包括白血球酯酶對應白血球數（Spearman's rank correlation coefficient R＝0.854，p＜0.001）、亞硝酸鹽對應細菌數（R＝0.537，p＜0.001）及血紅素對應紅血球（R＝0.548，p＜0.001）。系統預測泌尿道感染的最好指標為合併iChem白血球酯酶與iQ200細菌數，陰性概似比為0.06，陰性預測值為0.98，偽陰性率為3.1％及診斷勝算比為39.2。泌尿道感染病患尿液培養分離的菌叢以E. coli（59.4％）佔大宗，K. pneumoniae（9.4％）及Enterococcus（9.4％）次之。此外，感染革蘭氏陰性菌的病患相較於感染革蘭氏陽性菌的病患，具較高的iQ200白血球數（p＝0.012）與細菌數（p＝0.023）；而感染帶有廣效性乙型內醯胺酶表現型細菌的病患相較於感染無帶此抗生素表現型細菌的病患，具較高的細菌數（p＝0.047）。綜合以上結果，IRIS全尿液分析系統的執行成效令人滿意，並可提供預測泌尿道感染的附加價值，並且與致病菌的特性相關。

Automated IRIS instruments for urine chemistry and urine sediment were integrated as total urinalysis system recently. The advantages of shortening turnaround time and cross-checking results are expected between tandem iChem Velocity for chemistry and iQ200 ELITE for sediment, but the performance of this system has not yet been examined. In this study, the total IRIS urinalysis system was evaluated in routine testing for a purpose to predict urinary tract infection (UTI) and correlate with characterizations of UTI etiology. In National Cheng Kung University Hospital, 315 suspected UTI patients including 132 males and 183 females (mean age: 54.3 yr) were recruited consecutively and a retrospective survey of urinalysis data warehouse and urine culture was conducted. The urine culture yielded 98 (31.1%) positive and 217 (68.9%) negative for UTI, with the positive rate for female over that for male by ~3 folds (p<0.0001). The comparisons of the results by iChem and iQ200 revealed significant correlations for leukocyte esterase vs. white blood cell count (Spearman's rank correlation coefficient R=0.854, p<0.001), nitrite vs. bacteria count (R=0.537, p<0.001) and hemoglobin vs. red blood cell count (R=0.548, p<0.001). Combined items of iChem leukocyte esterase and iQ200 bacteria count served as the best rule-out criteria of UTI (negative likelihood ratio=0.06), with a negative predictive value of 0.98, false negative rate of 3.1% and diagnostic odds ratio of 39.2. The most frequently isolate was E. coli (59.4%), followed by K. pneumoniae (9.4%) and Enterococcus (9.4%). Furthermore, the Gram-negative isolates exhibited higher iQ200 WBC count (p=0.012) and bacteria count (p=0.023) than did Gram-positive isolates, whereas the extended-spectrum beta-lactamase phenotype exhibited a higher bacteria count than did those without this antibiotics resistance phenotype (p=0.047). In conclusion, the performance of IRIS total urinalysis system was satisfactory which might provide additional values in predictive of UTI and causative etiology.