Interpretation of Urine Culture Colony Counts

Unintential title alliteration. :)

In many cases, a CFU (Colony Forming Unit) count of < 100x10^6 (or 1 with 8 zeroes after) CFU/L means that you can safely assume the culture represents colonization, not an actual pathogen. This is especially true when you have multiple bugs growing in the same culture, and only one of them has >100x10^6 CFU/L.

I’ve had a few cases lately of obviously symptomatic patients with sub-threshold colony counts where the infection is taken less seriously due to the “colonization-only” culture. Intuitively I knew we should not be ignoring a positive culture in a symptomatic patient (and they did receive proper treatment) but I would like to know more about why these cultures would appear, on first glance, as colonization.

Starting with UpToDate (“Sampling and evaluation of voided urine in the diagnosis of urinary tract infection in adults”):

However, in symptomatic women with pyuria, lower midstream urine counts (ie, ≥10^2/mL) have been associated with the presence of bladder bacteriuria. Thus, in such instances, the findings of a colony count <10^5 but ≥10^2/mL may still be indicative of a UTI.

Note that you need to multiply 10^5 CFU/mL by 1000 (to get it in L) so we are comparing like units.

Best time for actually sending a urine culture is with the first morning pee (aka the one that is exceedingly urgent at 5am when you awake to a cat lying directly on your bladder), with the idea that not only will that pee be the most concentrated (since you likely haven’t drank anything for many hours) and the bacteria will have had the longest time period to replicate. Given that most scientific studies will strive to follow ideal practice, this >100 x 10^6 established threshold could have been determined from studies where they consistently drew this optimal morning culture. Unfortunately, what appears to be the originating study - from 1956! - on colony counts, doesn’t specify what time of day the cultures were drawn.

In practice, the most we can usually hope for is that the culture is drawn prior to the first dose of antibiotics.

UpToDate lists several more modern studies that try to clarify whether or not <10^5/mL should truly be considered contamination. A great one is available for free on NEJM from 2013: Voided Midstream Urine Culture and Acute Cystitis in Premenopausal Women. Unfortunately, they too don’t list the time of day the cultures were drawn. They didn’t care though because they were comparing voided midstream urine with catheter drawn urine, and found with E.Coli you could go as low as 10^2/mL (10^5/L) and still have it represent significant bladder bacteria. I also liked that they started with patients symptomatic of UTI before culture.

Other factors that may skew the relevence of colony count:

• Cultures from men (rarely colonized)
• Cultures from patients with initial exposure to antibiotics (but not yet in concentration enough to wipe out the bacteria entirely prior to culture)
• Cultures from pregnant women (treat whether it’s colonization or not)

In summary: Treat symptomatic infections, using the culture to direct antibiotic choice… regardless of colony count. Ignore asymptomatic bacteruria…. regardless of colony count. If symptoms cannot be assessed, colony count may be of use.