Hospitals are working to reduce urinary tract infections (UTIs), which account for more than 30% of all healthcare-associated infections. However, hospitals also face a growing problem of multidrug-resistant bacteria, due in part to overprescribing of antibiotics. This issue of Strategies explores research that evaluated a higher cutoff for urine culture colony counts to predict clinically significant UTIs for hospitalized patients, thereby potentially reducing the amount of antibiotics prescribed in this population.

By using more appropriate cutoffs for urine culture colony counts hospitals could reduce the total number of reported UTIs by more than 30% and slash the use of unnecessary antibiotics, according to the findings of a new study (Am J Clin Pathol 2012;137:778–84). Researchers at NorthShore University HealthSystem (NorthShore) in Evanston, Ill., reviewed 185 UTI cases in hospitalized patients and found that those with colony counts ≥100,000 CFU/mL were 73.86 times more likely to have clinically significant UTIs than patients with lower counts.

The authors cited a 2011 survey of clinical microbiology lab directors in which 100% of participants indicated that they report results for nosocomial UTIs the same way as all other UTIs, i.e., by reporting colony counts using two thresholds, ≥1,000 or ≥10,000 CFU/mL, regardless of the patient’s location. However, reporting the lower threshold is a concern because it can lead to overprescribing of antibiotics for UTIs that are not clinically significant, contributing to “development of multidrug-resistant bacteria and antibiotic-related side effects,” according to the authors.

In the investigators’ study, two physicians each reviewed all 185 UTI cases. The first reviewer established which patients had been diagnosed with a UTI by their physician. The second reviewer determined whether these diagnoses had been based on standard evidence-based criteria for nosocomial UTI, using criteria from the National Healthcare Safety Network (NHSN), such as fever and abdominal pain. The researchers then compared cases based on urine culture colony counts, focusing on a 100,000 CFU/mL threshold.

Following NHSN guidelines, the second reviewer classified 5.7% of patients with colony counts < 100,000 CFU/mL as having UTI, compared to 24.3% by the primary review. When classified by the secondary reviewer, patients with < 100,000 CFU/mL had a 98.6% reduced risk of being classified as having UTI. If positive results had been reported only for patients with ≥100,000 CFU/mL, the number of positive cultures could have been reduced by 38%.

“I think the most important finding was that if you use colony count cutoffs that are really appropriate for inpatients, it has the potential to reduce your reported rate of urinary tract infections and lower the amount of antibiotics that are used,” said Lance Peterson, MD, an author of the study. “Microbiologists have been saying for a long time now that we should not be screening urine looking for asymptomatic infection except in certain populations, such as pregnant women.” Peterson is the director of microbiology and infectious disease research at NorthShore.

Labs have not always reported cultures < 100,000 CFU/mL as positive, but over time the cutoff slid down as concern grew over UTIs in young women, according to Peterson. “For many years, the standard was 100,000 CFU/mL. Then several papers came out which found that sexually active, healthy women could have a symptomatic UTI with colony counts down as low as 1,000 CFU/mL. At that point, labs started reporting colony counts down that low so that they didn’t miss that part of the population. Now it has spilled over so that that is the standard for reporting everything—not just that one part of the population,” he said. “There is also the fact that medicine has become very, very complex, and often clinicians do not even realize that we have the NHSN guideline, so these practices can creep in and we end up over-calling and over-reporting these infections.”

Seeing a lab report of an organism in a patient’s urine is a powerful trigger for physicians to prescribe antibiotics, noted Barbara Trautner, MD, PhD, an associate professor of medicine and an infectious disease researcher at Baylor College of Medicine in Houston. “We have found in our research that once healthcare providers see an organism reported in the urine, they become very anxious to treat it, and they’re not very concerned about the colony count,” she said. “Clinicians need to understand that there is not a lab test that proves a patient has a symptomatic UTI, rather than asymptomatic bacteriuria. This truly comes down to clinical judgment: does the patient have symptoms or not?” Trautner was not associated with the study.

Peterson and his colleagues are working to implement some changes at NorthShore based upon their study findings. For inpatients, the cutoff will move to 100,000 CFU/mL. “If the physician really feels that there is a urinary tract infection, he or she can still call the laboratory and ask us to process it more, but we aren’t going to put any more in the report other than negative for UTI and that the cutoff is less than 100,000 CFU/mL,” he said. A clinical follow-up study is planned 1 year after the switch to the higher cutoff.

Trautner has experience taking a different approach—changing physician behavior to decrease the number of urine cultures they order. “This approach has been very effective for us,” she said. “These orders are down nearly 50 percent on the intervention wards at our institution.”

Peterson emphasized that UTIs are a publicly reported patient safety measure, adding extra incentive for proper reporting. “I think labs will be interested in appropriately reporting this kind of infection, otherwise your hospital will look worse and may appear that it’s not a good place to receive care if your rates are high,” he said. “And there are two reasons for rates to be high: one, that the medical staff is not doing a good job of taking care of Foley catheters, or, more likely, that the lab is over-reporting.”