Diagnostic value of flow cytometric urine analysis in urinary tract infections of newborns and infants Authors: Sinem Gülcan Kersin, Özgür Yanılmaz, Zeynep Arzu İlki, Hülya Selva Bilgen Published: 2025 ABSTRACT AbstractThis study aimed to compare the results of flow cytometric urine analysis and urine culture for the rapid diagnosis of urinary tract infections and to determine the leukocyte and bacterial cutoffs that could achieve maximum sensitivity and negative predictive value (NPV) during the newborn and infant periods. This cross-sectional observational study included urine samples submitted for culture from the neonatal intensive care unit and outpatient clinic between April 2019 and December 2021. The urine samples were first inoculated for culture and then analyzed for leukocyte and bacterial counts using a flow cytometric method (Sysmex UF-1000i, Kobe, Japan). Maximum sensitivity and NPV values ​​were determined by examining different bacterial and leukocyte cut-offs with Sysmex UF1000i. A total of 138 urine samples from 103 patients were included in the study. The reasons for urine examinations were prolonged hyperbilirubinemia, sepsis, and insufficient weight gain. Although no association was observed with leukocyte count, a bacterial cut-off of 83.4/µL was identified as the optimal threshold. Diagnostic values, including sensitivity (91.6%), specificity (88.6%), NPV (98.5%), and positive predictive value (PPV) (55%), were satisfactory. Based on current data, we suggest that the Sysmex UF-1000i is a valuable screening tool for detecting UTIs in asymptomatic newborns. Integrating this technology into clinical practice could enhance early diagnosis and treatment, ultimately improving neonatal care outcomes.Keywords: Newborn, Infant, Urinary tract infection, SysmexSubject terms: Clinical microbiology, Paediatric research Introduction Urinary tract infections (UTIs) are among the most common bacterial infections, affecting approximately 150 million people worldwide1. Although rare in the first 3 days of life, UTIs are common in newborns and infants2. The prevalence of UTI ranges from 0.1% to 1% in term newborns and from 4% to 25% in premature neonates, with studies indicating that diagnosing UTI in newborns is challenging3. In a meta-analysis of the prevalence of UTI in childhood, the highest prevalence, approximately 7%, was reported in uncircumcised boys less than 3 months and girls less than 12 months4. Rapid diagnosis and appropriate treatment of UTI are of great importance, especially in the neonatal period, due to its progression to sepsis and meningitis. Despite several screening methods, including dipstick urinalysis, automated urinalysis, and microscopic sediment urinalysis, urine culture remains the gold standard5. In critically ill neonates, treatment plans and approaches are guided by established protocols, with empiric antibiotic therapy initiated pending the results of traditional culture methods, which require at least 24–48 h. However, in well-appearing infants with conditions such as prolonged jaundice and/or insufficient weight gain, the use of a rapid and reliable screening method may help prevent unnecessary hospitalizations and antibiotic use. Fluorescence flow cytometry has become a valuable screening tool due to its improved analytical quality and specialized bacterial detection capabilities6. While reliability and cost studies using the flow cytometry method for urinary tract infections have been conducted across a wide age range, most studies focus on the adult population7. In our study, we aimed to evaluate urine samples using the fluorescent flow cytometry method and compare the results with the conventional culture method to save time in well-appearing infants. Materials and methods This cross-sectional observational study was conducted at a training and research hospital between April 2019 and December 2021, in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Marmara University Ethics Committee (protocol code: 09.2020.1190) for studies involving human participants. Infants who were evaluated for poor weight gain, hyperbilirubinemia, or suspected sepsis and for whom a decision was made to collect a urine sample were included in the study if their families signed a written informed consent form. Samples that could not be delivered to the microbiology department on weekdays between 08:00 and 16:00, or those with insufficient volume allowing only culture analysis, were excluded, as flow cytometry could not be performed in such cases (Fig. 1). Urine samples collected from the neonatal intensive care unit and the neonatal outpatient clinic were processed in the clinical microbiology laboratory using conventional culture (CC) and fluorescence flow cytometry (Sysmex UF-1000i, Sysmex Corporation, Kobe, Japan). Urine sample inclusion process. Urine samples were collected using a catheter or a urine bag. For catheter sampling, the periurethral area was cleansed with an antiseptic solution, after which a sterile catheter was inserted into the infant’s bladder to obtain the urine sample. The catheter was removed immediately following the procedure. For bag sampling, after cleansing and air-drying the genital area, a pediatric urine collection bag was securely attached. Once the infant voided, the urine was promptly transferred into a sterile container. Samples were delivered to the laboratory within one hour at the latest and processed for conventional culture. For culture, samples were inoculated onto 5% Sheep Blood Agar or MacConkey Agar (Biomerieux, France) and incubated at 37 °C. After 24 h of incubation, the Petri dishes were evaluated. If there was no growth, the incubation was extended to 48 h and the evaluation was repeated. Pathogens in samples with growth were identified using MALDI-TOF MS (Biomerieux, France). UTI was defined as more than 10.000 CFU (colony-forming units)/mL of a single pathogen obtained via bladder catheterization or greater than or equal to 100.000 CFU/mL in bag-collected urine. Values ​​below this ​​or the non-dominant growth of more than two microorganisms were defined as contamination. Following the culture process, fluorescence flow cytometry was performed using the Sysmex UF-1000i, which provides two distinct channels: one dedicated to bacterial analysis and the other to sediment particles (leucocyte, yeast, etc.). After the nucleic acids (DNA and RNA) in the urine sample are labeled with fluorescent dyes, each cell passes individually in front of a red laser. The signals from the components, including size information, granularity information, and nucleic acid information, are collected on scattergrams. The analyzer software automatically counts the components in the urine, and provides the user with quantitative values ​​in microliters8. Sysmex UF-1000i can detect potential urinary tract infections in about a minute instead of hours, and at a rate of up to 100 samples per hour. The system includes daily low and high-level quality control materials, which test the accuracy and analytical sensitivity of the counted components in the urine. In this study, the Sysmex UF-1000i white blood cell (WBC) and bacteria (BACT) count results were compared with the urine culture results using ROC curve analysis. Specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) were evaluated. Results A total of 138 urine samples from 103 patients were analyzed. Of the infants, 43.7% were born via normal spontaneous vaginal delivery, and 63.1% were male. The mean gestational week and birth weight were 34.4 ± 4.9 weeks and 2526 ± 1056 g, respectively. As shown in Table 1, the median sampling day was 14 (4–174). Pyuria was observed in six urine tests with a median of 10.5 WBC/HPF. Out of 138 results, 57% (n = 79) showed no growth, 34% (n = 47) were reported as contamination, and 9% (n = 12) showed bacterial growth (Table 2). Table 1Descriptive characteristics and distribution of participants and samples.Descriptive characteristics% (N)Delivery type (NSVD)43.7 (45)Sex of newborn (male)63 (65)Gestational week (mean ± sd)34.4 ± 4.9Birth weight (g) (mean ± sd)2526 ± 1056Mechanical ventilation36 (37)Parenteral nutrition48.5 (50)Sampling age (day) (median)14 (4–174)Sampling with a catheter38.8 (40)Indwelling catheter5.8 (6)Blood leucocyte count (/mm3)(min-max)13,466 (2800–76000)Thrombocyte count (x103/mm3) (min-max)273 (51–856)CRP positivity16.5 (17)Blood culture positivity6.8 (7)Lumber puncture culture positivity3.8 (4)Pyuria5.8 (6)Leukocyte esterase6.8 (7)Nitrite2 (2)Reason for urine sampling Prolonged hyperbilirubinemia48.5 (50) Suspected sepsis42.8 (44) Insufficient weight gain8.7 (9) Performing urinary ultrasonography15.5 (16) Descriptive characteristics and distribution of participants and samples. Table 2Comparison of conventional culture methods and fluorescence flow cytometry for samples revealing bacteria and/or leukocytes.Sysmex UFConventional cultureGrowthNo growth/contaminationTotalPositive127284Negative05454Total12126138 Comparison of conventional culture methods and fluorescence flow cytometry for samples revealing bacteria and/or leukocytes. In descending order of prevalence, the identified bacteria were Enterococcus faecalis, Klebsiella pneumoniae, Escherichia coli, Acinetobacter baumanii complex, Candida albicans, and Raoultella ornithinolytica. Leukocytosis, leukopenia, thrombocytopenia, or elevated CRP were not detected in any of the patients with positive urine culture. Urinary ultrasonography was performed in 16 infants with positive infection markers and suspected urinary tract infection, and grade 2 hydronephrosis was found in only two patients. Bayesian Kendall’s test was used to evaluate the probability of H1 (the association hypothesis) and H0 (the independence/no association hypothesis), with a stretched beta prior width set to 1. BF10 values of 1–3 represent anecdotal evidence for the correlation hypothesis, 3–10 represent moderate evidence, 10–30 represent strong evidence, and > 30 represent very strong evidence. Values of 0.33–1.33 represent anecdotal evidence against the no association hypothesis, 0.1–0.33 represent moderate evidence against the no association hypothesis, 0.03–0.1 represent strong evidence against the no association hypothesis, and < 0.03 represent very strong evidence against the no association hypothesis. We assessed the correlation of culture results with WBC and BACT. As seen in Table 3, WBC was found to be independent of culture results and had moderate evidence. When we performed the ROC analysis with BACT, the best result was 83.4/µL (Table 4). Table 3The correlation of culture results with WBC and BACT.Sysmex-UF1000i r BF₁₀UF-1000i WBC/µl0.0540.181UF-1000i BACT/µl0.427> 100Bayesian Kendall’s tau test. WBC white blood cell, BACT bacteria. The correlation of culture results with WBC and BACT. Bayesian Kendall’s tau test. WBC white blood cell, BACT bacteria. Table 4Sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV) and Youden index results for different threshold values.CutpointSensitivity (%)Specificity (%)PPV (%)NPV (%)Youden’s indexAUCMetric score26.7100%60.76%27.91%100%0.60760.9431.6175.391.67%87.34%52.38%98.57%0.79010.9431.7983.491.67%88.61%55.00%98.59%0.80270.9431.80188.483.33%92.41%62.50%97.33%0.75740.9431.76342.283.33%93.67%66.67%97.37%0.77000.9431.77385.583.33%94.94%71.43%97.40%0.78270.9431.78PPV positive predictive value, NPV negative predictive value, AUC area under curve. Sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV) and Youden index results for different threshold values. PPV positive predictive value, NPV negative predictive value, AUC area under curve. Discussion Early and rapid diagnosis of UTIs in newborns and infants is essential. It is also important that the diagnostic method used is fast, safe, reliable, and cost-effective. In infants, UTIs may present with non-specific symptoms such as poor feeding, insufficient weight gain, vomiting, irritability, neonatal jaundice, or fever alone9. Although the infant may appear well, further advanced examinations should be conducted. Urine culture is the gold standard test for diagnosing UTIs, but it has a turnaround time of up to 48 h. During this period, these babies are at risk of developing urosepsis. Empirically initiating unnecessary treatments due to the risk of urosepsis may lead to increased resistance and higher healthcare costs10. However, it is also known that in a high percentage of cases, 40% to 70%, urine cultures are negative11. The decision to hospitalize symptomatic infants is generally well-justified and made without confusion. Conversely, not hospitalizing at-risk asymptomatic infants may result in delayed diagnosis and treatment. Therefore, a method is needed to identify the infection early without delaying hospitalization. Can urinalysis provide important clues about the presence of a UTI? Although it is known that higher pyuria, leukocyte esterase, and nitrite positivity are more predictive for UTI in adults, the value of pyuria in newborns and the definition of UTI remain unclear, and the sensitivity of these markers is unknown12. The definition of pyuria varies in the literature. Some studies suggest that 10 WBC/mL in an uncentrifuged urine sample is a more sensitive indicator of a UTI; in contrast, other studies use a threshold of more than 5 WBC/HPF13. In a study conducted on 27,000 infants with a median age of 1.7 months, 3 WBC/HPF in dilute urine and 6 WBC/HPF pyuria in concentrated urine were significant for possible UTI14. Conversely, Rahman et al. reported that a threshold of greater than 20 WBC/HPF may offer some benefit in diagnosing UTI in newborns. In our study, the leukocyte esterase test was positive in 4 of 6 urine tests in which pyuria was detected, and no growth was seen in any of them15. Additionally, no infection marker abnormalities were detected in patients with positive urine cultures. Nevertheless, symptomatic cases, regardless of CRP status, should be managed as potential urosepsis. Since tests such as procalcitonin and IL-6, which can be used for infection evaluation, are not routinely performed, no assessment or comparison was made in this regard. One of the underlying causes of UTI in newborns is prolonged jaundice, with a prevalence of 6% to 18%, which may be the first indication of an underlying renal abnormality2. In a study, it was reported that abnormal urinalysis and microscopy results were seen in 42% of infants diagnosed with UTI, and renal abnormalities were detected, similar to the 30–55% frequency reported in the literature16. In our study, the most common reason for urine testing was prolonged hyperbilirubinemia, followed by the evaluation for neonatal sepsis and insufficient weight gain. Urinary ultrasonography was conducted on 16 infants, and grade 2 hydronephrosis was detected only in two patients. Given the challenges, such as the low sensitivity of urinary markers in detecting infections and the asymptomatic neonates with UTI, searching for faster and more reliable biomarkers to detect urinary tract infections continues until CC results are available. The rapid screening test, Fluorescence Flow Cytometry (Sysmex), can determine whether to initiate treatment before conventional method results are available. Previous studies have suggested various bacterial and leukocyte cut-offs for different populations. In a study conducted on hospitalized adult patients, a BACT cut-off value of 38.7/mL for UTI was established17. 40 WBC/µL and 460 BACT/µL were considered as cut-off points in a UF-1000i study involving nearly 2000 samples from 2 months to 100 years of age, with a mean age of 56 years18. A single cut-off value is not appropriate for UTI screening in all age groups1. Although there are many studies conducted with Sysmex, we could not find a study conducted in the neonatal and infant periods. Kim et al. reported that UF-1000i is not suitable for UTI screening under 15 years of age due to contamination1. İlki et al. reported that the cut-offs for leukocytes and bacteria were found to be 30 and 50/µL, respectively, and that Sysmex could not be used as a screening test for children aged 0–5 years19. In the literature, contamination of up to 63% of the urine bag has been reported; approximately one-third of the samples in our study were contaminated20. We believe that the high contamination rate in our study, may be attributed to the fact that 60% of the urine samples were collected using urine bags, and most likely due to improper aseptic technique by the parents during the procedure. None of the studies in the literature reached 100% sensitivity and NPV, which remained at 95.4% sensitivity and 98% NPV19. In this study, while we could not find a cut-off value for WBC, but we found that the bacterial cut-off value of 83.4/µL that corresponds to 91.6% sensitivity and 98.5% NPV in our study. Thus, in samples above this cut-off value, antibiotic treatment could be initiated 48 h earlier rather than waiting for the CC results. This reduction in time may be particularly crucial in the neonatal period due to the high incidence of urosepsis. The limitations of this study were its monocenter design, small sample size, and lack of a control group. In newborns, ideally, samples for urine culture should be collected suprapubically or via catheter. We did not establish a control group using suprapubic aspiration or catheterization for the diagnosis of urinary tract infection in healthy neonates, as these procedures would have been invasive and ethically inappropriate solely for the purpose of creating a control arm. High contamination rates and low positive CC results, combined with the limited repertoire of neonates and the potential emergence of other underlying conditions, present significant challenges in accurately diagnosing and managing infections in this population. Urosepsis is a condition with very high mortality and morbidity. The use of the cut-off value, we determined for Sysmex UF-1000i is valuable, as it enables not only the collection of urine cultures and immediate initiation of treatment in symptomatic infants but also the early diagnosis and timely treatment of asymptomatic infants at risk for urinary tract infections, such as those with prolonged jaundice. These infants are typically below the phototherapy threshold and are evaluated for the etiology of hyperbilirubinemia. Urine cultures are obtained, and follow-up is scheduled after two days in the outpatient clinic. Including flow cytometry in the diagnostic process for these infants may facilitate the timely and appropriate initiation of treatment, thereby optimizing clinical management. In conclusion, based on the available data, we suggest that the Sysmex UF-1000i can be used alongside urine culture in asymptomatic but at-risk babies for UTI. In babies above the threshold level, antibiotic treatment can be initiated without waiting for CC results, while in those below the cut-off level, treatment can be withheld until CC results are obtained. This strategy will help prevent treatment delays and allow for early initiation within 48 h. We consider it important for future studies to be repeated with a larger cohort of newborns to validate the findings and contribute to clinical practice. Acknowledgements We would like to thank Ibrahim Kandemir for performing the statistical analyses. Author contributions Conceptualization, S.G.K. and H.S.B.; methodology, S.G.K., O.Y., H.S.B.; software, validation, formal analysis, S.G.K.; writing—original draft preparation, S.G.K.; writing—review and editing, visualization, H.S.B. and Z.A.I. All authors have read and agreed to the published version of the manuscript. All authors reviewed the manuscript. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Informed consent Since urine tests and cultures were performed for diagnostic and therapeutic purposes, informed consent was obtained from all subjects’ legal guardians. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.