UTI diagnostic stewardship can improve antibiotic stewardship
It could be easy to imagine that the gold standard for diagnosis of urinary tract infection (UTI) is urine culturing with antimicrobial susceptibility testing to ensure prescription of the correct antibiotic. However, there are many aspects to the diagnosis of a UTI and an overreliance on urine culture alone can result in asymptomatic bacteriuria being misdiagnosed as a UTI, with an accompanying unnecessary overuse of antibiotics. CLI chatted to Dr Kimberly Claeys (University of Maryland School of Pharmacy) to learn more about how careful diagnostic stewardship of UTIs versus asymptomatic bacteriuria can reduce the use of antibiotics as well as reducing the number of urine culture requests.
Why are urinary tract infections (UTIs) more common in acute care settings and also in older people?
UTIs are very common across all settings, outpatient, and acute care. However, they are one of the most common types of infections we see in acute care settings. In terms of overall incidence, UTIs tend to be more common in older adults (particularly older women) secondary to increased prevalence of specific risk factors, such as urinary catheter, chronic comorbidities, and hormonal changes in women.
However, as clinicians, we’re beginning to realize that a lot of what we categorize as UTIs are misdiagnosed and in reality, are asymptomatic bacteriuria: patients who have bacteria in their urine but don’t have
a true infection. We’ve seen in recent literature that the number of patients, especially in older adults who get treated for UTI, often don’t really have genitourinary symptoms seen with true infection and likely did not benefit from antibiotic therapy.
Asymptomatic bacteriuria is quite common in older adults, as are non-specific signs and symptoms that drive misdiagnosis. Often in older adults, we think about delirium, weakness, falls, or if they just look or sound a little ‘off’ that day, the healthcare professional will often say, “Oh, this might be a UTI”, but that’s often-misattributing UTI to these symptoms and that drives overdiagnosis and overtreatment.
Urine culture is often done on a combination of blood agar and MacConkey agar (Adobe Stock)
What is the gold standard for UTI diagnosis and what are the challenges with diagnosis and treatment?
The idea of a gold standard UTI diagnosis is challenging because, although we often think of the urine culture as being the gold standard diagnostic test, UTI is a clinical diagnosis. For a proper UTI diagnosis, signs and symptoms of infection within the genitourinary tract are needed to accompany that positive urine culture. An over-reliance on urine culture without any associated symptoms is a major driver for inappropriate antibiotic treatment of asymptomatic bacteriuria. Then we try to think about how we can use diagnostic tests in a way to help really rule out infection because it’s hard to rule in when you have asymptomatic bacteriuria (positive urine culture) with non-specific signs and symptoms of infection. And again, this is especially prevalent in older adults. So, we think about combinations of urinalysis (UA) and urine culture results. We have UA components like pyuria (which is the presence of white blood cells or pus in the urine), which when negative in a patient without genitourinary symptoms, we can be more confident that there is no infection. It’s harder to use pyuria to rule in UTI because pyuria can also be present in non-UTI situations. It could be caused by STIs or other non-infectious causes such as certain medications. Older adults in general, have higher incidence of pyuria without infection. Hopefully, therefore, we’re able to use our diagnostic tests to help rule out infection when the UA and urine culture results are negative. Ruling in infection is very challenging because it must be a constellation of diagnostic tests and clinical signs and symptoms.
What is the difference between antibiotic stewardship and diagnostic stewardship?
That’s a great question because antimicrobial stewardship has been around for a while and people tend to be familiar with the concept of antimicrobial stewardship. In the USA, antimicrobial stewardship and the use of antimicrobials is highly regulated. Diagnostic stewardship, however, is a more recent concept that clinicians are becoming more familiar with, but is still not routinely integrated into each health system.
Antimicrobial stewardship is a commitment to promoting and monitoring the use of antimicrobials to preserve their future effectiveness. As the definition suggests, you focus on the antimicrobials: you think about how to optimize their use, whether it’s selection, dosing, duration, to improve patient outcomes and minimize unnecessary use and the evolution of antimicrobial resistance.
Diagnostic stewardship really works upstream of antimicrobial stewardship, and it focuses on how to optimize the use of the diagnostic tests throughout the whole diagnostic process – whether it’s when you’re ordering the test, the test processing or the test reporting – how can we improve that whole process to have decreased risk of diagnostic error and again improve patient outcomes.
Both processes work to improve patient outcomes, but at slightly different phases of the diagnostic process. They are very complementary, and, in fact, sometimes you see overlap, not only of principles but practices, between the two.
Thinking about diagnostic stewardship as a continuum of antimicrobial stewardship has been helpful for me as someone who does antimicrobial stewardship in order to get more involved in diagnostic stewardship.
What does a UTI diagnostic stewardship protocol look like and how can it affect numbers of urine cultures performed and antibiotic days of therapy?
Diagnostic stewardship protocols will look different based on the different sites they’re implemented in, as discussed in our paper “Implementing diagnostic stewardship to improve diagnosis of urinary tract infections across three medical centers: a qualitative assessment” [Claeys et al., 2023]. When we think about protocols, we want to think about those phases of interventions: ordering, processing, and reporting. Previously, we published an expert consensus document about ascertaining the best diagnostic stewardship principles around urine culturing [Claeys et al., 2022], and we used that expert consensus to build the protocols within this current study, focusing on those phases of intervention. The elements of best practice that an institution will adopt from those recommendations will be heavily dependent on that institution and the other interventions that are already in place. Interventions targeting at least one or two of those stages will be helpful for improving urine culture diagnosis and antibiotic use, particularly when focused on ordering or processing.
Urine culture ordering
With urine culture ordering, there’s a series of practices that were considered to be appropriate and practices that are considered to be inappropriate. For instance, having urine cultures embedded and automatically ordered with order-sets is an inappropriate practice. This leads to urine cultures being ordered without any real critical thought around that diagnostic test. Instead, something more like the clinical decision support that we detail in “Implementation of diagnostic stewardship to improve urinary tract infection antibiotic use across three medical centers” was deemed appropriate [Claeys et al., 2025].
It is useful to have rules in place to help guide the use of urine testing, because, not least, urine testing can mean a lot of different things.
It could be UA by itself, or UA with urine culture, or even urine culture by itself. Not every patient who is seen needs both a UA and a urine culture. Some patients just need a UA. There might be a non-infectious work-up in place, where patients might need a urine culture to rule out infection, for example, in very specific conditions such as pregnancy or other urological conditions.
So, can we first create clinical decision support around which tests should be utilized? With clinical decision support for ordering, the goal is to help inform clinicians about appropriate indications for when to order a UA and a urine culture and when it might be inappropriate or unnecessary. A lot of this is around symptomatology: do patients have symptoms, genitourinary symptoms, refined to the urinary tract to guide that treatment decision, because the non-specific signs and symptoms, such as those mentioned above for older adults (altered mental status, falls, weakness) could be attributed to a lot of different non-infectious etiologies. Clinical decision support will guide the prescriber through those decisions about whether there are true indications of infection: is there a clinical indication based on symptomatology? This is important as it has been found that the ordering phase is extremely impactful for mitigating not only urine culture ordering but also antibiotic use.
Urine processing
With the processing phase, reflex urine culturing can be two disparate different methods. A lot of people, when they see the term ‘reflex’ urine culturing, assume that if you order a UA and certain criteria are met, the urine culture is automatically performed. That practice was deemed as inappropriate by our expert panel, as it’s not a critical evaluation of the need for urine culture.
What we’ve studied is ‘conditional reflex’ urine culture to try to differentiate it. It’s a term that we’re hoping people adopt to differentiate the practices. Conditional reflex urine culturing is the opposite. The clinician intended to order a UA with culture, so the intent was to order the culture. However, the culture will not performed unless certain criteria are met on UA. The urine sample is held for up to three days, so the clinician can still call and order it if they still have true suspicion of infection. So, they’re opposing practices, and the conditional reflex urine culture was determined to be an appropriate practice.
Also, the conditional reflex criteria of cut-off of 10 urine white blood cells per high-power field (WBCs/hpf) on UA was deemed to be likely the minimum appropriate criteria. Meaning that if the count is less than 10 WBCs/hpf on UA, the urine culture will not be performed. However, there are different UA criteria that can be used, the combinations of WBC count and biochemical assays, such as leukocyte esterase and nitrites, but from review of the literature experts determined that a cut-off of 10 WBCs/hpf had sufficient negative predictive value without leading excess of ‘false positive’ cultures.
Changing urine culture processing is relatively easy because it is something that happens in the background – clinicians are not always aware of if, so it doesn’t involve as much behavioural change as the ordering and recording phases do. Up front, it can be a challenge to implement these changes in practice because, operationally, two labs are often involved: the UA is done in the chemistry lab and the culture is done in the microbiology lab. Hence, implementation involves coordination and policies that involve more cross-communication than would have been there previously, and so an effort has to be made upfront in the logistics of the labs working together, but once it’s implemented, it’s easy to maintain and has consistently shown immediate decreases in urine cultures processed by upwards of 40%.
Result reporting
The reporting phase also has the potential to further decrease antibiotic prescription. Being more selective or restrictive of your reporting could be helpful, but that’s sometimes harder to operationalize. For instance, not including certain antibiotics in the report, those that have a high risk of resistance or adverse drug events, especially if there are other options available. There’s also cascade reporting, which entails broader spectrum antibiotics not being released unless there’s resistance to the narrower spectrum, preferred, antibiotics. These are interventions that are commonly used within antimicrobial stewardship, which is where overlap of the diagnostic stewardship and antibiotic stewardship occurs and have been shown to be effective. Another potential, intervention is withholding of results of the culture entirely. In Canada, there have been several studies showing that in non-catheterized adult patients having a report that says, essentially, “this urine culture doesn’t meet criteria, we’re not going to release the results unless you call for it” has been shown to decrease treatment. However, the complete withholding of results is difficult to operationalize, and whether it’s even possible it might depend on the governance of the institution.
Our findings
From our current paper [Claeys et al., 2025], we found that the conditional reflex urine culturing was the most impactful, decreasing urine culturing by 30–50%, and we also saw a relative decrease of almost 40% in UTI antibiotic days of therapy. The decrease in urine culturing that we saw with the implementation of the conditional reflex urine culture protocol is consistent with previous literature, which is usually quoted at around 30 or 40%. Few papers have been able to show a decrease in UTI antibiotic use associated with these protocols, but there are some that have included this. I think one of the other papers that I’d like to highlight showed that a combination of ordering and processing changes led to a decrease in antibiotic use [Watson et al., 2020].
So, to reiterate, the protocols adopted locally are going to be highly dependent on doing an assessment of your own institution and what will work. We found that the conditional urine culturing is probably the easiest way to decrease urine culturing immediately. However, to have a long-term impact on antibiotic use, there probably needs to be a combination of interventions along the diagnostic stewardship and antimicrobial stewardship spectrum.
What comments would you have for anyone wanting to introduce or further develop diagnostic stewardship for potential UTIs?
A major consideration determining how to implement these practices at your institution. There’s no one-size-fits-all approach and making sure you meet with key stakeholders, front-line clinicians, and leadership to determine what will be the most impactful at your institution is key. I hope these series of papers can serve as a framework to help further conversation about urine culture diagnostic stewardship locally.
The idea of diagnostic stewardship and urine culture diagnostic stewardship is still relatively novel, and even just increasing aware-ness of these concepts is a major positive impact of this paper. That is, hospitals and institutions have more discussion around how they can improve diagnostic test utilization and be more critical about how they’re using their current diagnostic tests, especially in an infection as common and challenging as UTIs where we don’t really have a gold standard diagnostic. I think wider-spread implementation of diagnostic stewardship interventions will help decrease inappropriate treatment of asymptomatic bacteriuria. We need to be more critical about how we use those diagnostic tests, and I hope that our paper [Claeys et al. 2025] helps demonstrate ways that we could potentially do that.
We still have important work to do in the coming years to refine the conditional reflex criteria for urine culture. In our study, and as reported in previous literature, one commonly used criterion is that a urine culture is not performed when the UA shows a WBC count of <10 cells/hpf, unless specifically requested. This cut-off is based on prior studies; however, much of that evidence comes from retrospective data may not fully account for differences across patient populations. Moreover, these retrospective analyses are vulnerable to misclassification, as cases of asymptomatic bacteriuria may have been incorrectly classified as UTI. Therefore, further research is needed to determine the optimal UA WBC cut-off, and to assess whether that threshold should be uniform across all patients or tailored to specific populations. And if trailered by patient population, how does that become operationalized within a medical centre or health-system.
Looking ahead, one of the biggest challenges will be operationalizing some of these nuances into a stronger, more integrated workflow. Clinical decision support I think will increasingly be helpful, as a lot of our clinicians are very busy, have a lot of cognitive overload and high cognitive burden (with the sheer volume of tests they have to review, and numbers patients they’re seeing), and our workflow and clinical decision support tools sometimes don’t improve that workflow, they actually hinder it. So how do we make these smarter, more patient specific and more streamlined to help clinicians, as opposed to having them be hindered and feel that they are being burdened by them. These are the main things I would hope to see in the future, building from where we are now.
Bibliography
1. Claeys KC, Trautner BW, Leekha S et al. Optimal urine culture diagnostic stewardship practice – results from an expert modified-Delphi procedure. Clin Infect Dis 2022;75(3):382–389 (https://doi.org/10.1093/cid/ciab987).
2. Claeys KC, Weston LE, Pineles L et al. Implementing diagnostic stewardship to improve diagnosis of urinary tract infections across three medical centers: a qualitative assessment. Infect Control Hosp Epidemiol 2023;44(12):1932–1941 (https://doi.org/10.1017/ice.2023.106).
3. Claeys KC, Brown CH, Pineles L et al. Implementation of diagnostic stewardship to improve urinary tract infection antibiotic use across three medical centers. Clin Infect Dis 2025:ciaf411 (https://doi.org/10.1093/cid/ciaf411).
4. Watson KJ, Trautner B, Russo H et al. Using clinical decision support to improve urine culture diagnostic stewardship, antimicrobial stewardship, and financial cost: a multicenter experience. Infect Control Hosp Epidemiol 2020;41(5):564–570 (https://doi.org/10.1017/ice.2020.37).
The interviewee
Kimberly Claeys PharmD, PhD, Associate Professor
Department of Practice, Sciences, and Health Outcomes Research, University of Maryland School of Pharmacy, Baltimore, Maryland, USA, and Fellow of the Center for Innovation in Diagnosis at the University of Maryland School of Medicine, Baltimore, Maryland, USA
Email address: kclaeys@rx.umaryland.edu
