Your patient is a healthy 32 year old woman with moderate to severe dysuria and urinary frequency. You suspect cystitis, so you ask her for a urine sample, and send it for culture, to confirm the diagnosis. You prescribe antibiotics, pending culture results, to alleviate her symptoms.
This scenario is repeated countless times for the millions of cases of suspected cystitis that occur annually in the United States. But are there good data to guide interpretation of urine culture results, particularly when the culture shows no growth of Escherichia coli, the most common uropathogen?
The answer, perhaps surprisingly, is no…..
In this week’s NEJM, Hooton et al report the results of a study designed to supply these data. When cystitis is suspected, a midstream urine sample is often obtained. Given the potential for contamination of midstream urine samples by periurethral microorganisms, prior research has endeavored to define thresholds for bacterial colony counts that distinguish infection from contamination. A landmark study in 1982 showed that in patients with symptoms of cystitis, urinary coliform colony counts as low as 10² colony-forming units (CFU) per milliliter correlated with bladder bacteriuria. Yet, until now, no study has sought to corroborate this now three-decade-old finding. In addition, the positive predictive value of other organisms that may grow from a midstream urine sample, in particular enterococcus and group B streptococcus, has not been established.
In the current study, eligible patients met criteria for suspected acute uncomplicated cystitis, defined as bladder infection in healthy, premenopausal, non-pregnant women with no known urologic abnormality. Each participant provided a clean-catch midstream urine specimen, and thereafter underwent bladder catheterization to provide an accompanying reference specimen. 202 paired specimens provided the material for the study. Microbial species and colony counts in each pair were compared to determine the positive and negative predictive values of organisms that grew in midstream urine cultures. Organisms were identified and quantified to 10 CFU per milliliter.
The study results confirmed earlier research showing that Escherichia coli in midstream urine is highly predictive of bladder bacteriuria, even at colony counts as low as 10 CFU/ml. The positive predictive value was 93% for growth of at least 10² CFU/ml, and 99% for growth of at least 10⁴ CFU/ml. Midstream urine cultures with Klebsiella pneumonia and Staphylococcus saprophyticus, both of which occurred only in high colony counts in this study, also correlated with bladder bacteriuria. However, enterococcus and group B streptococcus cultured from midstream urine were not predictive of bladder bacteriuria, even at high colony counts, suggesting these organisms rarely cause cystitis. For enterococcus, the positive predictive value was 10% for growth of at least 10² CFU/ml and 33% for growth of at least 10⁴ CFU/ml, while for group B streptococcus, the positive predictive value for similar colony counts was 8% and 14% respectively; note that these two organisms grew in only a small number of cultures, such that definitive conclusions cannot yet be drawn.
Corroboration of the high positive predictive value of E. coli in midstream urine draws attention to the fact that some commercial clinical laboratories only report growth for counts of 10⁴ CFU/ml or higher, a potential source of false negative culture results for E. coli bladder infection. In an accompanying editorial, Dr. Michael Donnenberg, of the University of Maryland School of Medicine, highlights yet another twist to the story of this seemingly straightforward diagnostic test, noting that one quarter of bladder specimens in the study were sterile, despite accompanying midstream urine cultures showing growth of a uropathogen. Could these patients have symptomatic urethritis, he queries, and if so, are antibiotics beneficial in this setting or not?
NEJM Deputy Editor Dr. Lindsey Baden adds, “Data improving our understanding of this common condition are needed. This report shows the importance of microbial cultures in improving our understanding of cystitis. ” The results, he notes, underscore the extent to which we may, in some clinical settings, need to lower our thresholds for what we consider a clinically significant colony count.
There has been debate about the contribution of midstream urine cultures to the treatment of acute uncomplicated cystitis. Nevertheless, urine cultures are frequently obtained, and have a role in the work up of patients with complicated cystitis, and those failing standard treatment. Studies like this one that address a question as basic as “What is a positive culture?” are welcome indeed.