In the present study, we confirmed that UC-3500 and UF-5000 urine pipeline could quickly and effectively exclude the possibility of bacterial UTI. The best cut-off value of WBC and BACT in UF-5000 were determined at the same time. We further chose the most suitable scheme for screening UTI (only BACT parameter was positive). Rapid screening of procedures can reduce unnecessary urine cultures and minimize the number of missed detections. Furthermore, the screening program effectively reduces waiting times for urine cultures.
Out of 2,600 urine specimens evaluated, 425 bacterial culture were positive(16.4%) for UTI and 2175 specimens were negative(83.6%).Of the 425 positives cultures,347 samples were Gram-negative bacteria(81.65%). Escherichia coli. was the most frequently identified Gram-negative bacteria(228/425,53.65%). Among the Gram-positive bacteria, Enterococcus faecalis. was the most frequently identified. 9 specimens were marked as contamination because both Gram-negative bacteria and Gram-positive bacteria were identified. These results indicated that Escherichia coli. were the major causative agent of UTI in this area. And this is consistent with mohnarin’s annual report on continuous monitoring of bacterial resistance in urine samples in recent years. There was no significant difference in the distribution of positive bacterial infections. (16, 17).It is different from Sysmex urine analyzer UF-500i and UF-1000i. The UF-5000 has been developed for improved performance for classification of Gram-Neg, Gram-Pos, Gram-Neg / Gram-Pos, and unclassified. UF-1000i classified bacteria as either “ords” or “cocci/mixed”, not Gram-Neg, Gram-Pos bacteria. Compared with UF-1000i, the ability of the new BACT Info flag of UF-5000 is more significant for clinical treatment.
The thresholds of WBC and BACT Info flag information with UTI provided by the Sysmex corporation were 10 /µL and100 /µL. However, these thresholds were lower than the reference range of the instrument(18). The laboratory can modify and set the threshold independently, so our laboratory set higher thresholds of Info flag information(WBC,30 /µL, BACT,300 /µL). Therefore, in this study, we evaluated the accuracy of Gram-positive bacteria and Gram-negative bacteria instead of the accuracy of Info flag information in instrument to determine UTI or not. The results showed that the consistency between Info flag information of instrument and positive results of bacterial culture was fair(Kappa = 0.339).Thus, UF-5000 Info flag information can’t be used as a clinical reference for bacterial classification. This conclusion is different from that of Ren, Du et al. They believe that the consistency between Info flag information of instrument and bacterial culture's positive results was substantial (Kappa = 0.775 or 0.724)(13–20).We analyze the reasons for the different conclusions in recent research, the instrument's status, the number of specimens, and the distribution of bacterial species in other laboratories, making it difficult to interpret two different conclusions. However, few studies focus on this question, so this conclusion needs to be further analyzed by expanding the specimens.
Urine culture was considered as the gold standard, the accuracy of LEU and NIT were evaluated, the results showed that LEU and NIT had good negative predictive value(LEU,96.03%, NIT,91.08%). This conclusion suggested that the negative results of LEU and NIT were helpful for clinicians to exclude urinary tract infection quickly. Sagbo and Pandey have the same viewpoint in their researches. Other parameters showed the PPV(30.32%) and coincidence rate(kappa = 0.277) of LEU, the sensitivity(49.84%) and coincidence rate(kappa = 0.566) of NIT were unsatisfactory. Therefore, it is thought that a single parameter (LEU or NIT) was not an effective detection parameter to judge UTI or not. In clinical diagnosis and treatment, they need to be combined with UF-5000 WBC and BACT to judge UTI more accurately.
According to the number of colonies, the positive results can be divided into104 CFU/ml and 105 CFU/ml. Therefore, in this experiment, the urine samples were divided into three groups (A. a negative growth, B. bacterial growth = 104CFU/ml, C. bacterial growth = 105CFU/ml). Compared the level of WBC and BACT in three groups, data showed that there were significant differences among all the groups(P༜0.01). Further analyses indicated that the lowest level of WBC and BACT was found in the A group, while the highest level was found in the C group. The level of WBC and BACT increases with the number of bacterial colonies. This conclusion is similar to previous studies (Kim SY. 2018; Liao.2020) (11, 21). The results showed a high consistency among WBC, BACT, and the number of bacterial colonies. Clinicians can preliminarily judge the bacterial content of UTI by the level of WBC and BACT, and these results may act as a reference for clinical anti-infection treatment.
Different reference ranges of WBC and BACT were set for males and females by different urine sediment analyzers and conventional microscopic examinations. It could probably be due to the physiological structure is different in males and females (18, 19, 22).
As a consequence, we also constructed ROC curves using only information regarding sex (12, 23). Furthermore, different cut-offs were used for judging UTI, the cut-off values of WBC were 163.25 /µL for males (AUC,0.859, 95%CI,0.826–0.891) and 50.20 /µL for females (AUC,0.796, 95%CI,0.764–0.829), the cut off value of BACT were 104.80/µ L for males (AUC,0.956, 95%CI,0.936–0.976) and 705.50/µL for females (AUC,0.940, 95%CI,0.920–0.959). Two of them have a high predictive value for UTI, but the Youden index, sensitivity, and specificity of BACT were relatively high, which were sufficient for diagnosing infection. The cut-off value of BACT was different from Rosa’s research. Cut off = 58 /µL was used to screen urinary tract infection (12). Participants in the study believed that instrument status, regional differences contributed to different conclusions. Consequently, different laboratories should set their cut-off value.
As mentioned earlier, LEU, NIT in combination with WBC, BACT in order to successfully determine whether UIT. In our study, the value of WBC and BACT in UF-5000 were regarded as positive values when they were more significant than the cut-off value. Thus, we established six UTI screening programs. The bacterial culture results were used for performance evaluation. All schemes had good negative predictive value and coincidence rate. Each evaluating indicator of Single index screening strategy② was better than strategy①, the PPV of strategy③-⑥(66.99%-91.49%) was better than strategy②. The PPV, NPV, and coincidence rate of strategy ⑤ and strategy ⑥ were about 90%. The specificity achieved 99.19%, but the result showed low sensitivity for screening UTI of only 41.35%. For clinical work, excluding patients with suspected UTI quickly and effectively could be more valuable than screening out specimens of UTI. Concerning this, PPV is the primary index of exclude patients with suspected UTI. So, in our opinion, although the strategy⑤ and strategy⑥ can screen UTI patients nearly 100%, it is not suitable for clinical work. strategy②is more suitable (PPV,97.54%). However, other more appropriate screening programs may be chosen dependent on different clinical needs.