UTI Complicated: Diagnosis

Urine for both urinalysis and urine culture should be collected with suspected acute complicated UTI (Hooten & Gupta, 2021; Meyrien, 2021). Urinalysis results inform the diagnosis. Since pyuria (i.e., urine containing white blood cells or pus) is present in almost all patients with UTI, its absence suggests an alternative diagnosis, particularly in patients with nonspecific symptoms. As a result, pyuria is consistent with, but not diagnostic of, UTI and pyuria in the bacteriuric patient do not identify symptomatic infection.

The ideal voided urine sample for UTI evaluation accurately represents the bladder bacterial count with minimal contamination by bacteria colonizing the distal urethra and genital mucosa. Theoretically, this would be a clean-catch, midstream sample of the first micturition of the day. But in fact, there is no clinical evidence that this ideal specimen yields more accurate results.

The likelihood of detecting a bladder bacteriuria by voided urine culture is highest if urine is collected on arising. This sample is likely to be most concentrated, and bacteria in the bladder will have had time to multiply overnight. However, this ideal sample is not practical since most cultures are obtained when the patient is seeing the clinician. With samples collected later in the day, a more dilute urine and partial bacterial washout due to multiple voids may lower the colony count below the accepted definition for bacteriuria diagnostic of a UTI. Collection of a midstream urine, with or without cleaning of the urethral meatus, at the time of clinical evaluation likely produces a reasonable specimen for analysis.

An optimal clean-catch, midstream urine is collected through the following steps in attempts to minimize the degree of contamination with bacteria from the urethra (Ali & Wesonga, 2021; Hooten & Gupta, 2021; Meyrien, 2021):

1. Local disinfection of the meatus and adjacent mucosa should be performed with a non-foaming antiseptic solution, such as Dakin's solution (diluted sodium hypochlorite solution 0.5%).
2. This region should then be dried with a sterile swab to avoid a mixture of the antiseptic with urine.
3. Spread the labia or pull back the foreskin to avoid contact of the urinary stream with the mucosa.
4. The initial stream of the voided specimen should be discarded since the initial urine flushes urethral contaminants. It is the subsequent midstream sample that should be sent to the laboratory. However, clinical studies do not demonstrate that cleaning the meatus is associated with lower contamination rates. Thus, for patients in whom the cleaning step may be impractical or difficult, midstream urine collected (with the labia spread, for females) is likely an appropriate sample.

For males with suspected chronic bacterial prostatitis, evaluation of the last few drops of urine after the prostatic massage is indicated (Ali & Wesonga, 2021; Hooten & Gupta, 2021; Meyrien, 2021). Firm prostatic massage per rectum, from lateral to midline on each side, causes the contents of the prostatic ducts to be expressed. Vertical strokes in the midline will then project the secretions into the urethra and permit the counting of leukocytes. This maneuver, however, should be avoided when acute bacterial prostatitis is suspected because of the risk of bacteremia.

When a urine sample is collected for culture, it should be sent immediately to the laboratory since bacteria will continue to multiply in the warm medium of newly voided urine, leading to increased bacterial counts. If such immediate referral is not possible, the container should be transported in iced water and stored in a refrigerator at 4°C.

When a midstream urinary sample is obtained at home and brought into the laboratory, the patient should be instructed to keep the tightly closed container in a plastic bag or package containing cold water and ice cubes until the sample is delivered. Cooling stops bacterial growth, but the bacteria can still grow on a culture medium the following day (Ali & Wesonga, 2021). However, refrigeration of urine could alter urinary leukocytes and thus affect the urinalysis interpretation.

Dipsticks to evaluate for urine leukocyte esterase and nitrite can be used as a screening tool to diagnose UTIs (Hooten & Gupta, 2021; Meyrien, 2021). Dipsticks are generally performed whenever a UTI is suspected. Notably, urine collected by dipstick is not routinely obtained in young nonpregnant females with a history suggestive of a UTI (i.e., typical symptoms without vaginal discharge or irritation), as the dipstick generally does not provide additional useful information. In general, dipsticks to evaluate for leukocyte esterase and nitrite should not be performed in patients without any symptoms consistent with a UTI, as a positive dipstick, which would denote the presence of pyuria or bacteriuria, does not indicate a UTI in an asymptomatic patient.

Dipsticks detect the presence of leukocyte esterase and nitrite in the urine. Leukocyte esterase corresponds to pyuria. Leukocyte esterase may detect >10 leukocytes per high power field.

A positive nitrite test is a reliable index of significant bacteriuria, although a negative test does not exclude bacteriuria. Normally no detectable nitrite is present.

Most bacteria that cause UTIs (E. coli, Proteus mirabilis, and Klebsiella) produce nitrate reductase, an enzyme that transforms urine nitrate, a metabolite of alimentary origin, into nitrite that is detected by dipsticks. When bacteriuria with one of these organisms is significant, the test is positive in about 80% of cases in which the urine has incubated for at least four hours in the bladder.

Possible causes for a negative nitrite test despite bacteriuria include (Hooten & Gupta, 2021; Meyrien, 2021):

• Decreased urine pH (due to cranberry juice, vitamin C, or other dietary supplements)
• Early phase of infection
• High (>5 g/L) glycosuria
• High urine specific gravity (>1.02)
• Insufficient bladder incubation time for conversion of nitrate to nitrite
• Low urinary excretion of nitrate
• Other potential urinary pathogens such as Staphylococcus, Pseudomonas, Group B Streptococcus, Acinetobacter, Enterococcus faecalis, and fungi which do not produce a nitrate reductase
• Proteinuria >1 g/L

False-positive nitrite tests can occur with substances that turn the urine red, such as the bladder analgesic phenazopyridine or the ingestion of beets (Hooten & Gupta, 2021; Meyrien, 2021). Most studies have supported the diagnostic value of the nitrite and leukocyte esterase results, particularly when they are concordant. A meta-analysis including 70 publications concluded that the urine dipstick test alone is useful when nitrite and leukocyte esterase tests are positive (Ali & Wesonga, 2021).

In truly infected patients, many leukocytes (>10/microL or 10,000/mL) should generally be present. Given the very high association between infection and pyuria, an absence of pyuria on microscopic assessment can suggest colonization instead of infection when there is bacteriuria (although bacteriuria and pyuria do not necessarily signify infection, particularly if there are no symptoms).

Hemocytometry is the preferred method of microscopic assessment for pyuria, but the less accurate method of counting leukocytes in the sediment of centrifuged urine is more commonly employed. In a study of females with symptoms of UTI, a urine leukocyte count ≥8 cells/microL, as measured by hemocytometry, was found in 61 of 62 females who had evidence of bladder bacteriuria through a culture of suprapubic aspiration or urethral catheterization samples (Hooten & Gupta, 2021; Meyrien, 2021). In contrast, only 7 of 34 asymptomatic females had ≥ 8 urine leukocytes/microL, of whom three were found to have asymptomatic Chlamydia trachomatis infection of the cervix. In a separate study of females with urinary symptoms and low count bacteriuria (>102 to 104 colony-forming units/mL), pyuria defined as ≥20 leukocytes/microL was associated with isolation of urinary pathogens (E. coli and Staphylococcus saprophyticus) in contrast to other organisms. The threshold of 8 cells/microL on hemocytometry corresponds to 2 to 5 leukocytes per high power field in centrifuged urine sediment (Ali & Wesonga, 2021). This method has some technical pitfalls. Two variables are contamination with vaginal secretions in females and the volume of supernatant in which the centrifuged pellet is resuspended, which will affect the leukocyte count.

White blood cell casts in the urine indicate kidney inflammation, which may reflect pyelonephritis or other renal conditions. In the absence of pyuria, especially when various strains are found, the presence of bacteria is usually due to contamination during sampling.

Finding various bacterial strains should not be considered a UTI when the patient is asymptomatic. However, a repeat urine sampling should be done in a symptomatic patient who has complaints consistent with cystitis and bacteriuria but does not have pyuria. If the pattern is the same, with bacteriuria but no pyuria and with persistent symptoms of urethral inflammation, this is consistent with "acute urethral syndrome," and antibiotic treatment should generally not be undertaken. However, this is a vague clinical entity that could be equivalent to bacteriuria at low counts.

True infection without pyuria is unusual. Pyuria can occur without apparent bacterial infection in patients who have already taken antimicrobials. Other causes of sterile pyuria include (Ali & Wesonga, 2021; Henri, 2017; Hooten & Gupta, 2021; Meyrien, 2021):

• Contamination of the urine sample by the sterilizing solution used to clean the meatus.
• Contamination of the urine sample with vaginal leukocytes from vaginal secretions.
• Chronic interstitial nephritis.
• Nephrolithiasis.
• Uroepithelial tumor:
  • In the case of uroepithelial tumors, both malignant cells and leukocytes are shed.
  • The tumor can cause inflammation that results in the appearance or exacerbation of sterile pyuria.
  • Malignant cells are difficult to distinguish from leukocytes on routine urinalysis.
• Painful bladder syndrome/interstitial cystitis.
• Intra-abdominal inflammatory process adjacent to the bladder.
• Infection with atypical organisms, such as Chlamydia, Ureaplasma urealyticum, or tuberculosis.
• Also, some females with UTIs may have low bacterial counts in their midstream urine. If the clinical laboratory does not quantify bacterial counts below a certain threshold, they may not be identified as having bacteriuria despite consistent symptoms and pyuria.

The role of pretreatment urine culture in the evaluation of suspected UTI is to confirm the presence of bacteriuria and to identify and provide antibiotic susceptibility information on the causative organism (often retrospectively, if treatment is empirically given) (Hooten & Gupta, 2021; Meyrien, 2021).

A urine Gram stain limits potential causative organisms and helps select antimicrobial treatment. Culture of a clean-catch voided urine specimen is warranted for most patients with suspected UTI, except for healthy, nonpregnant, young females with typical symptoms of acute simple cystitis (i.e., without fever or concern for ascending infection) who have no risk factors for resistant infection. In general, urine culture should not be performed in nonpregnant patients without symptoms consistent with a UTI, as bacteriuria does not indicate a UTI in an asymptomatic patient.

In asymptomatic patients, the standard threshold for bacterial growth on a midstream voided urine reflective of bladder bacteriuria as opposed to contamination is ≥105 colony forming units (CFU)/mL (Hooten & Gupta, 2021; Meyrien, 2021). However, in symptomatic females with pyuria, lower midstream urine counts (i.e., ≥102/mL) have been associated with the presence of bladder bacteriuria. Thus, in such instances, the findings of a colony count <105 but ≥102/mL may still indicate a UTI. Lower bacterial counts still representative of infection are also seen in males, in patients already on antimicrobials, and with organisms other than E. coli and Proteus species.

Normal values in a noninfected midstream, clean-catch sample is <105 CFU/mL, with bacterial growth primarily due to E. coli from fecal contamination. In early studies, >95% of patients with bacterial counts ≥105 CFU/mL in voided urine, but only a minority of those with lower counts had concomitant bacteriuria on a catheterized specimen (Hooten & Gupta, 2021; Meyrien, 2021).

Subsequent studies have identified many females with symptoms and pyuria consistent with a UTI, but colony counts <105 CFU/mL in voided urine (Hooten & Gupta, 2021; Meyrien, 2021). This was demonstrated in a study of 202 premenopausal, nonpregnant females who presented with at least two symptoms of acute cystitis, a midstream, clean-catch urine collected, and who subsequently underwent urethral catheterization to collect a bladder urine specimen (Hooten & Gupta, 2021). Of the 121 females who had E. coli grow in the catheterized specimen, 49 (40%) had counts of <105 CFU/mL in their voided urine. A ≥102 CFU/mL threshold in voided specimens had a 93% positive predictive value for bladder bacteriuria with E. coli. Another study estimated that 88% of females with symptoms and a CFU count ≥102/mL on voided urine have a UTI (Hooten & Gupta, 2021; Meyrien, 2021).

It is not well understood why some infected females have low colony counts. One possibility is that low counts reflect insufficient sensitivity of conventional urine cultures.

In a study that included 220 females with symptoms of acute cystitis and 86 asymptomatic females, urine was tested with both standard culture and quantitative polymerase chain reaction (qPCR) for E. coli, S. saprophyticus, and sexually transmitted pathogens (Hooten & Gupta, 2021; Meyrien, 2021). Among symptomatic females, 81% of urine cultures were positive for any uropathogen, while 95% of samples were qPCR positive for E. coli (19 were positive for S. saprophyticus qPCR, 1 for Mycoplasma genitalium, and 1 for Trichomonas vaginalis). In contrast, urine culture and qPCR were positive for E. coli in only 11% and 12%, respectively. These findings suggest that almost all females with typical urinary complaints and a negative culture still have an infection with E. coli. Management of these patients is similar to those with higher colony counts (Hooten & Gupta, 2021; Meyrien, 2021).

There are also several other settings in which a colony count of ≤105/mL represents true infection rather than contamination (Ali & Wesonga, 2021; Hooten & Gupta, 2021; Meyrien, 2021):

• Among patients already being treated with antimicrobials.
• Among males, in whom contamination is a much lesser problem.
• When organisms other than E. coli and Proteus are present (Pseudomonas, Klebsiella, Enterobacter, Serratia, and Moraxella species).