Evaluation of a new automated urine cell analyser (Sysmex UF-1000i) for bacteriological urinalysis

Abstract number: 1733_646

von Wulffen H.

Objectives: We wanted to assess to what extent the UF-1000i (SYSMEX) automated urine cell analyser can support bacteriological urinalysis, in particular to find answers to the following questions: can it replace Gram stain? Can we dispense with culture in case of negative results for bacteria, leukocytes and yeast-like cells? Is there a suitable cut-off for number of bacteria to select urine specimens for direct susceptibility testing?

Methods: Urine specimens with a request for urine culture were processed using the UF-1000i in parallel with Gram stain and culturing overnight of 0.001 mL undiluted urine each onto UTI, CNA, and Sabouraud agar using calibrated loops. Specimens were also tested for inhibitory activity using spore plates.

Results: To select out negative specimens UF-1000i results were considered positive if white blood cells exceeded 20/ml and/or bacterial particles were counted over 50/ml and/or yeast-like cells exceeded 25/ml. In comparison, cultures yielding organisms at 103 CFU/mL or higher were considered positive. On this basis out of 363 urine specimens evaluated 42% of the UF-1000i results were true positive, 30% true negative, 22% false positive, and 6% false negative. However, only 2/24 false negative results yielded 106 CFU/mL, all others had 104 CFU/mL or lower, most of them with mixed flora, thus reducing significantly false negative results to less than 0.6%. Only looking at the specimens without inhibitory activity (n = 232) and using a cut-off of 105 CFU/mL or higher for predicting a urinary tract infection, the UF-1000i yielded a sensitivity of 97.6%, a specificity of 86.7%, a positive predictive value of 80.0%, and a negative predictive value of 98.5% when using a cut-off of 100 bacterial particles/ml. This cut-off also proved useful for selecting those specimens suitable for direct susceptibility testing.

Conclusions: For our purposes the UF-1000i could replace Gram stain of urine specimens and thereby reduce time-consuming microscopy. The low rate of truly false negative results around 0.6% appears to be tolerable and would allow dispensing with culture in case of negative UF-1000i results and thus further reduce workload in the laboratory. Specimens from certain patients (e.g. haematology patients with cytopenia) or certain materials (e.g. suprapubic bladder aspirates) should be cultured in any case.

Session Details

Date: 31/03/2007
Time: 00:00-00:00
Session name: European Society of Clinical Microbiology and Infectious Diseases
Location: ICC, Munich, Germany
Presentation type:
Back to top