1. Introduction

---

Urinary Tract Infection (UTI) which is one of the most common factors in visiting urology outpatient clinics can be non-complicated as in causing bladder irritation and dysuria symptoms, can also become complicated so much as to lead to hypertension in progressive stages, by causing renal parenchyma damage and renal scar tissue. Especially in female patients, 60-80% of the dysuria complaints are associated with bacteriuria [1-3]. Considering that uremia was not determined in 25-30% of symptomatic UTI patients, especially in noncomplicated UTI cases, ordering urine culture test has always been controversial.

Even though medical history and physical examination (FE) are valuable in UTI diagnosis, the most fundamental test to be ordered is TUA. In UTI patients the most frequent accompanying symptoms are dysuria, and urgent and frequent desire to void. Empiric antibiotic therapy is started in these patients before ordering urine culture test [4,5]. Urine culture test is an expensive, tedious and time consuming procedure. Bacterial growth in culture media requires 18-48 hours time and this increases hospital stay period as well as treatment costs. An inexpensive test that will decrease the need for ordering urine culture test, will also decrease economic burden and increase the efficacy of the empiric treatment to be initiated. High accuracy rate of TUA used for diagnostic purpose before empiric therapy increases the efficacy of the therapy and reduces the need for culture test. Thus, TUA which is an inexpensive and quick test, can help saving time and money, and protects public assets. That’s why, it is used before urine culture test routinely, for predictive purposes [6,7].

The aim of our study was, to determine the efficacy and power of TUA in predicting culture test results and so to make our contribution to the correction of general mistake of ordering both TUA and urne culture in our country, to give a support to reduction of economic loss and to increase success rates of empiric treatment.

2. Materials And Methods

---

The data of 228 patients were retrospectively analysed, who admitted to our urology clinic between January and March 2018 for dysuria and UTI symptoms and both TUA and urine culture tests were performed on the same day. After the patients were informed about appropriate wiping method to cleanse the urethral area, midstream urine sample was taken into two sterile urine cups (one for TUA the other for urine culture test). For TUA, macroscopic, dipstick and direct microscopic examination of urine were established. For the efficacy and predictivity of TUA, culture test results were regarded as reference line. In macroscopic observation, blurred view in urine and the colors other than light yellow; in microscopic observation >5/hpf value for erythrocyte and leucocyte, and >15/hpf value for epithelial cell were accepted as positive. Statistical correlation was analysed by chi-square test. No threshold value was taken for mucus parameter. (Statistical correlation was analysed by Mann-Whitney U test.) In dipstick test, nitrite (>0.06 mg/dL), hemoglobin (>0.01 mg/dL), leucocyte esterase (>25hc/µL), protein (>30 mg/dL), glucose (>15 mg/dL), bilirubin (>0.02 mg/dl), urobilinogen (>2mg/dL), ketone (>5 mg/dL acetoacetate) values were accepted as positive and the correlation with urine culture was analysed by chi-square test. Statistical correlation of pH and density values with urine culture was analysed by Mann-Whitney U test. For culture, urine sample inoculated ob blood agar was incubated at 35˚C for 18-24 hours. Subsequently, < 105 CFU/mL colony number was accepted as negative (group 1), and higher colony number was accepted as positive (group 2) Statistical correlation was evaluated by using IBM Statistical Package for Social Sciences (SPSS) Version 22.0 program. P< 0.05 values were accepted as statistically significant.

3. Results

---

Average age of 228 patients included in our study was 51.95±16.59 years (51.42±16.1 and 53.9±18.33 years in group 1 and group 2 respectively). Statistically significant difference was not determined between the groups with respect to age (p=0.360). The number of patients was 179 (78.6%) in Group 1 (culture negative) and 49 (21.4%) in group 2 (culture positive). Chi-square analysis was used in the statistical analysis of the parameters examined in urine microscopy (leucocyte, erythrocyte, epithelial cell) of culture result groups. The sensitivity of leucocyte in the microscopic analysis was determined as 53%, specificity as 81%, positive predictive value (PPV) as 44%, negative predictive value (NPV) as 86.3% , and Odds Rate (OR) as 5 (p=0.000). For erythrocyte in microscopic analysis, the parameters of sensitivity, specificity, PPV, NPV and OR were determined as 44.89%, 75.4%, 33.3%, 83.3% ve 2.5 respectively (p=0.005). The same parameters for epithelial cell were determined as 6.1%, 94.4%, 23%, 78.6% and 1.1 (p=0.886). For the correlation of urine appearance (blurred or clear) with culture positivity, the same parameters were determined as 8.16%, 99.4%, 4%, 79.8% and 15.8 respectively (p=0.008). We determined no statistically significant relation of urine color with culture positivity, amd the same parameters were determined as 6.1%, 92.7%, 18%, 78.3% and 0.83 respectively (p=0.537). The correlation between the presence of mucus cells and culture positivity was analysed by Mann-Whitney U test and no difference was determined (p=0.487)Considering the relation of the Dipstick parameters with culture results, for leucocyte esterase the same parameters above were, 69.3%, 77.6%, 45.9%, 90.2% and 7.8(p=0.000); for hemoglobin parameter, 67.3%, 63.6%, 33.6%, 87.6% and 3.6(p=0.000); for nitrite parameter, 32.6%, 96%, 69.5%, 83.9% and 11.9(p=0.000); and for protein parameter 53%, 81.5%, 41%, 84.8% and 3.91 (p=0.000) respectively.

When the other dipstick parameters were considered, according to chi-square test, glucose, bilirubin, urobilinogen, ketone parameters were determined to be statistically not significant in culture predictivity (p=0.682, p=0.345, p=1.000, p=0.345 respectively). Also pH ve density values were determined to be insignificant in predicting urine culture results (Mann Whitney-U test; p= 0.110, p=0.374 respectively). (See Table 1 for all statistical results.)

Table 1

4. Discussion

---

In the studies on female patients, the prevalence of dysuria was determined as 2-5%, and the rate of bacteriuria in these patients was reported as 60-80% [1-3,8]. The tests implemented in order to determine bacteriuria can be inexpensive and fast tests but with less accuracy like TUA, as well as expensive and time taking tests but with high accuracy like urine culture tests. Urine culture test is gold standard in revealing bacteriuria but it is expensive and requires time. Furthermore, according to studies in the literature, 52-68.3% of culture tests are known to be negative. However, there are also studies with much higher number of cases, reporting much higher culture negativity 97.7% than these studies [7]. It is essential to predict the patients with probable culture negative results in order to prevent this situation causing increased cost [9-11]. Right at this point, urine dipstick test and urine microscopy (TUA) becomes important.

According to the results of our study leucocyte esterase, nitrite, protein and hemoglobin were found to be significant in regard to predicting urine culture positivity. In the meta-analysis, Deville et al [12] examined the effectivity of nitrite test in predicting culture result and the sensitivity and specificity were determined as 50% and 82% respectively in the general population (the data of 15 studies). The same parameters were determined to be 59% and 97% in the population of urology patients (data of 3 studies) as 59% and 97%. In the study of Kayaalp et al [7] these values were found to be 17.7% and 90.1 % in general population respectively. And in the study of Yüksel et al[11] these values were 36.1% and 95.4%. In our study however, these values were determined as 32.6% and 96%, and OR was 11.9. We can say that, in the cases where nitrite test was negative, urine culture was also negative at such a high rate of 96% (specificity).

In the meta-analysis of Deville et al [12] sensitivity and specificity rates for leucocyte esterase in general population was 62% and 70%, and for urologic population as 86% and 93%, respectively. These values were determined to be 71% and 83.6% respectively in general population, in the study of Kayaalp et al [7]. In the same study, OR value for leucocyte esterase was determined to be at a much higher rate with respect to nitrite (For nitrite and for LE as 2 and 12.5 respectively)[7]. And these values were 86.1% and 65.5% respectively in the study of Yüksel et al [11]. In our study however, these values were determined as 69.3% and 77.6%, and OR as 7.8. According to our results, although the sensitivity of leucocyte esterase was higher in comparison to nitrite test, OR value was determined to be lower (OR values for nitrite and LE as 11.9 and 7.8 respectively).

In the meta-analysis of Deville et al[12] nitrite and/or leucocyte esterase negative tests results in general population were determined to be very significant, for positive tests however, the cases with positive results in both test were considered as significant, and the cases with only one positive result, they reported that further assessments were required.

In the study of Yüksel et al [11] Hb and protein parameters among dipstick parameters were considered, and sensitivity and specificity rates were determined to be 70.3% ve 56.7% for HB, and 51.9% ve67.8% for protein, respectively.

In our study, Hb and protein parameters were found to be diagnostically significant, and sensitivity, specificity and OR values were 67.3%, 63.6% and 3.6 for Hb; and 53%, 81.5% and 3.91 for protein. In our study, the specificity of proteinuria in determining bacteriuria was found to be significantly higher with respect to Hb parameter. In the other dipstick tests, these values were 4%, 96.6% and 1.2 for glucose,; 2%, 98.3% and 1.2 for urobilinogen; and 0%, 96.8%, and 0.77 for ketone and bilirubin, respectively; and none of these were found to be statistically significant in predicting urine culture results.

In our study, >5 leucocyte or erythrocyte at every field was regarded as positive. In our study leucocyte and erythrocyte parameters analysed in urine microscopy were found to be statistically significant in regard to predicting culture test results. In the study of Kayaalp et al [7] the sensitivity, specificity and OR values of microscopic leucocyte parameter were found as 68.2%, 87.8% ve 15.5 respectively. In the study of Yüksel et al [11] however, sensitivity and specificity values for leucocyte were found as 88% and 61.8 %, and for erythrocyte as 74.1% and 54.6% respectively. In our study however, these values for leucocyte were determined to be 53%, 81% and OR as 5, and for erythrocyte 44.8%, 75.4% and OR as 2.5 respectively. According to our results, diagnostic value of leucocyte was higher with respect to erythrocyte. These values for epithelial cell were determined as 6.1%, 94.4% and 1.1, hence not statistically significant. Since there was no threshold value for mucus parameter, Mann-Whitney U test was implemented and found to be statistically insignificant in regard to predicting urine culture.

As for urine macroscopy, the sensitivity, specificity and OR values were 8%, 99.4% ve 15.8 for urine appearance parameter and found to bestatistically significant in predicting urine culture. For color parameter these were 6.1%, 92.7%% and 0.83, and not valuable in predicting urine culture.

In conclusion, nitrite, LE protein in dipstick tests; leucocyte and erythrocyte in microscopy parameters, and urine appearance parameter in macroscopy, were determined to be significant in predicting urine culture results It is clear that specificity values of TUA parameters were very high, sensitivity values however were relatively low. Final conclusion to be deduced from that is, it was not quite necessary to conduct urine culture in cases where TUA parameters were specified as negative, thus high cost and time loss due to urine culture could be reduced by the help of TUA, but in cases where TUA was positive, it was necessary to make verification by combined values (in conjunction with the other TUA parameters) or by urine culture.

We also would like to highlight that recent reports in the literature noted bacteria were isolated by PCR and bacterial genomic identification (16S rRNA sequencing) methods in the cases with negative urine culture and this led to the controvercy of whether urine samples were sterile or not [13,14]. It is clear that randomized and prospective studies with higher patient numbers are required in order to illuminate this issue which will be of a center of interest in the near future.

References

---

1. Naber KG, Schito G, Botto H, Palou J, Mazzei T. Surveillance study in Europe and Brazil on clinical aspects and Antimicrobial Resistance Epidemiology in Females with Cystitis (ARESC): implications for empiric therapy. Eur Urol 2008; 54:1164-1178.
2. De Backer D, Christiaens T, Heytens S, De Sutter A, Stobberingh EE, Verschraegen. G. Evolution of bacterial susceptibility pattern of Escherichia coli in uncomplicated urinary tract infections in a country with high antibiotic consumption: a comparison of two surveys with a 10 year interval. Journal of Antimicrobial chemotherapy 2008; 62:364-368.
3. Heytens S, Boelens J, Claeys G, DeSutter A, Christiaens T. Uropathogen distribution and antimicrobial susceptibility in uncomplicated cystitis in Belgium, a high antibiotics prescribing country: 20-year surveillance. European Journal of Clinical Microbiology & Infectious Diseases 2017; 36:105-113.
4. Little P, Moore MV, Turner S, et al. Effectiveness of five different approaches in management of urinary tract infection: randomised controlled trial. BMJ 2010; 340: c199.
5. Gupta K, Hooton TM, Naber KG, et al. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clinical infectious diseases 2011; 52: e103-e120.
6. Hiscoke C, Yoxall H, Greig D, Lightfoot NF. Validation of a method for the rapid diagnosis of urinary tract infection suitable for use in general practice. Br J Gen Pract 1990; 40:403-405.
7. Kayalp D, Dogan K, Ceylan G, Senes M, Yucel D. Can routine automated urinalysis reduce culture requests?. Clinical biochemistry 2013; 46:1285-1289.
8. Verheij RA, Van Dijk CE, Abrahamse H, et al. Landelijk informatienetwerk Huisartsenzorg. Feiten en cijfers over huisartsenzorg in Nederland. Utrecht/Nijmegen: NIVEL/IQ healthcare; 2009.
9. Okada H, Sakai Y, Miyazaki S, Arakawa S, Hamaguchi Y, Kamidono S. Detection of significant bacteriuria by automated urinalysis using flow cytometry. Journal of clinical microbiology 2010; 38: 2870-2872.
10. Martinez MHM, Bottini PV, Levy CE, Garlipp CR. UriSed as a screening tool for presumptive diagnosis of urinary tract infection. Clin Chim Acta 2013; 425: 77–79.
11. Yüksel H, Kaplan İ, Dal T, Kuş S, Toprak G, Evliyaoğlu O. İdrar kültürü testi gerekliliğini öngörmede tam otomatik idrar analizi sonuçlarının performansı. Journal of Clinical and Experimental Investigations 2014; 5.
12. Devillé WL, Yzermans JC, Van Duijn NP, Bezemer PD, Van Der Windt DA, Bouter LM. The urine dipstick test useful to rule out infections. A meta-analysis of the accuracy. BMC urology 2004; 4: 4.
13. Heytens S, De Sutter A, Coorevits L, et al. Women with symptoms of a urinary tract infection but a negative urine culture: PCR-based quantification of Escherichia coli suggests infection in most cases. Clinical Microbiology and Infection 2017; 23: 647-652.
14. Hilt EE, McKinley K, Pearce MM, et al. Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder. Journal of clinical microbiology 2014; 52:871-876.