Basic POL microbiology: diagnosing UTI.

The diagnosis of urinary tract infection (UTI) is based on a quantitative estimate of the concentration of bacteria in the urine specimen. We assume each bacterium multiplies to become a colony, hence we call it a "colony-forming unit" (CFU). In an asymptomatic individual, more than 100,000 CFU/mL in an adequately collected specimen indicates UTI. Many believe this is ordinarily the diagnostic threshold (see p. 36, "What is the significance of finding greater or lesser numbers of CFUs?"). In symptomatic individuals or those susceptible to recurrent infection, the threshold can be as low as 1,000 CFU/mL.

Pathogen isolation/identification

Isolating and identifying pathogenic bacteria generally require that two types of media be inoculated:

Selective, for example, MacConkey or EMB, for isolation and identification of enteric gram-negative rods such as Escherichia coli.

Sheep blood agar medium suitable for estimating the number of CFUs and for isolating and identifying group D streptococci.

Inoculate the blood agar and selective medium using a special platinum loop calibrated to deliver either 0.01 or 0.001 mL of urine to the surface of the medium. Although most labs use the 0.01 mL loop, some use both (inoculated on separate plates) as a QC check.

Flame and cool the loop, then immerse it in a well-mixed, uncentrifuged urine sample. Streak the loopful of urine in one line across the center of the plate, then spread it evenly over the surface of the medium at right angles to the original streak, using the same loop. Don't flame the loop between streakings.

Continue spreading the inoculum so that an even lawn of bacterial growth will be produced. Tech tip: Hold the colony-count loop flat against the surface of the medium and let the pressure from the weight of the loop do all the work. Let it glide across the surface; do not apply extra pressure. This can lead to tearing of the medium surface and erroneous colony counts.

After inoculation, incubate all plates with the medium sides up in an incubator, with room air, at 35 [degrees] C. Examine the plates for bacterial colonies after 18-24 hours.

Interpreting colony count

In your examination, count the number of colonies growing on each plate. Multiply the number of colonies by the dilution factor to estimate the bacterial load in the original urine specimen For example, if you counted 50 colonies while using the 0.01 mL inoculating loop, the calculation is 50 x 100 = 5,000 CFU/mL, with 100 being the dilution factor. If you used the 0.001 mL loop to do the streaking, the colony count is calculated as: 50 (colonies) x 1,000 (the dilution factor) = 50,000 CFU/mL.

After you have gained experience, you can usually estimate 100 or greater colonies by cursory inspection of the plate without actually performing an accurate count. Confluent growth generally is considered to be in the range of 1,000 colonies, a number that is then multiplied by the dilution factor. Traditionally, bacterial colony counts have been reported in the following categories:

* [less than or equal to] 10,000 (or [10.sup.4]) CFU/mL

* About 50,000 (5 x [10.sup.4]) CFU/mL

* [greater than or equal to] 100,000 ([10.sup.5]) CFU/mL

A colony count of greater than 100,000 CFU/mL generally indicates a UTI; counts of less than 10,000 CFU/mL usually represent contamination. The significance of counts in the range of 50,000 CFU/mL must be determined by the physician, based on clinical and other laboratory information.

For example, is the patient having symptoms? Are there pus cells in the urine specimen? Is the leukocyte esterase test and/or the nitrite test parts of the urine chemistry panel positive (as determined with a urine dipstick)? Has the patient been on antibiotic therapy, in which case the lower colony count may represent the beginning of bacterial resistance?

The clinician needs to know if there are more than 100,000 ([10.sup.5]) CFU/mL of urine because the likelihood of a significant UTI increases at this threshold for asymptomatic individuals. The reliability of a culture of a single, clean-catch specimen yielding [greater than] [10.sup.5] CFU/mL is approximately 80% in women and nearly 100% in men. Two consecutive specimens from a woman increases the yield to 95% of the tested group that has an infection.

Women with a symptomatic UTI often have urine bacterial colony counts ranging from 100 to 1,000 CFU/mL and therefore may not be diagnosed by the standard urine culture. Performing this semi-quantitative assessment with a 0.001 mL loop cannot detect [less than] 1,000 bacteria/mL. If over 100 colonies per mL of urine is considered significant in symptomatic patients, it is prudent to inoculate a second plate with the 0.01 mL loop. Significant bacteriuria can then be identified as [greater than] 100,000 CFU/mL for asymptomatic women and [greater than] 100-1000 CFU/mL for symptomatic women

RELATED ARTICLE: WHAT IS THE SIGNIFICANCE OF FINDING GREATER OR LESSER NUMBERS OF CFUS?

Each colony on a culture plate generally started with a single bacterium in the original specimen. Although uninfected bladder urine is sterile, samples collected by the spontaneous passage of urine are never sterile because the cleansing process is not completely efficient. Therefore, bacterial colonies generally will be found in the urine of uninfected as well as infected patients.

Studies have shown that less than 100,000 bacteria/mL of urine can be found in uninfected individuals. If [greater than] 100,000 bacteria/mL of urine are found in a "clean catch" specimen, it was found that the patient was very likely to have a urinary tract infection even if he was experiencing no symptoms.

In symptomatic patients, as few as 1,000 bacteria (colony forming units)/mL can indicate a urinary tract infection. Potential pathogens should be identified if indicated for patient management decisions.

Richard Belsey is professor emeritus of pathology at Oregon Health Sciences University. Daniel M. Baer is professor of pathology there and works at the Veterans Affairs Medical Center, also in Portland. He is a member of MLO's Editorial Advisory Board.