In the absence of clinical correlation, a positive urinalysis does not necessarily predict a positive urine culture, and in turn a positive urine culture is inadequate to diagnose CAUTI.

A prospective observational study in an urban New York public teaching hospital¹ assessed whether positive urinalysis could predict positive urine cultures in intensive care unit patients with indwelling catheters. For 4 months, in order to compare urinalysis and subsequent quantitative culture results, investigators collected urine samples from 106 randomly selected ICU patients with urethral catheters in place for >12 hours. Patients with genitourinary trauma or instrumentation prior to admission, urinary tract infection (UTI) diagnosed prior to admission, or anuria with urine production <15 mL/day were excluded.
Positive urine cultures were defined as ≥10⁵ organisms/mL urine with ≤2 species present. The authors found 44 positive cultures out of 300 samples. Test characteristics for urinalysis findings were as follows:

Nitrites in urinalysis were 91.8% specific but only 29.5% sensitive for a positive culture, giving +LR (likelihood ratio) 3.52 (95% CI 2.3-5.3) and −LR 0.56 (95% CI 0.6-0.9). The most specific finding was the combination of nitrites and leukocyte esterase, which was 95.7% specific and 20.5% sensitive, with +LR 4.76 and −LR 0.83, respectively. This combination’s only moderately diagnostic LRs, however, do not obviate the need for culture. An important study caveat is the high prevalence of candida and enterococci species without nitrite reductase activity, which may limit the applicability of findings about the utility of nitrites.

Another study² performed in Texas sought to determine whether bacteriuria identified based on positive urine cultures correlated with symptomatic episodes of CAUTI in patients with long-term urinary catheters over a year-long period. Patients were included if they were enrolled in a hospital-based home care program and had indwelling catheters for >14 days. No exclusion criteria were specified. Nurses collected urinary samples from 14 patients weekly for 5 to 6 weeks and then monthly
thereafter. Outcomes included the number and colony counts of different bacterial species in specimens to assess the relationship of these findings to CAUTI incidence. Number and colony counts were also performed sequentially before, during, and after antibiotic courses. Clinical or microbiologic criteria for CAUTI were not clearly defined, and the diagnosis of CAUTI was instead at the discretion of attending physicians.

Results showed that 111/177 (63%) urine cultures grew ≥10⁵ cfu/mL of two to four different microbial species. Only three patients experienced fevers during the trial, each of which was diagnosed as CAUTI; one patient was diagnosed twice. In these patients, 3/4 urine cultures were sterilized during antibiotic treatment, but microorganisms returned to their same concentration 2 to 12 days after therapy, further supporting the idea that colonization is common and does not correlate with febrile episodes.

Study caveats included small size and failure to clearly delineate the symptoms defining a CAUTI nor the required workup to rule out other causes of incident fevers. Given that CAUTI symptoms are particularly challenging to interpret since typical UTI symptoms (e.g., dysuria, urgency) may be absent with an indwelling catheter in place, these issues limit the applicability of this study.