Automatic urine cultures from catheter-obtained samples—time to implement a practice change to improve value

ABSTRACT

Objectives

Urine cultures are often automatically obtained after urinary catheterization in young children, even in the absence of pyuria, to test for urinary tract infections (UTIs). Although this practice conflicts with some newer guidelines, it is frequently followed in busy emergency departments (EDs) to minimize the need for a repeat invasive procedure. To assess the implications of this longstanding practice, we aimed to describe the frequency and characteristics of children with normal urinalysis (UA) and positive urine culture obtained via catheterization, and to describe their clinical course.

Methods

A single center, retrospective cohort study was performed for otherwise healthy children aged 6 to 24 months, presenting to a Pediatric ED between January and June 2019 who underwent UTI testing via a urine catheterization. The point-of-care (POC) UA and urine culture results along with any follow-up phone call documentation were reviewed and analyzed using descriptive statistics.

Results

Of the 818 urine cultures obtained via catheterization during the 6-month study period, 131 (16%) cultures were reported as positive. Of these positive cultures, 14 (10.7%) of the patients meeting inclusion criteria had a normal POC UA. In follow-up phone calls after the ED visit, the majority of these 14 patients were asymptomatic without any antibiotic treatment and 3 (2.3%) patients were still febrile and classified as a potentially missed UTI.

Conclusions

The routine practice of sending urine cultures from all catheterized urine samples in children 6 to 24 months, regardless of POC UA results, rarely detect missed UTIs. In alignment with more recent practice guidelines, this practice should be reconsidered in low-risk children seen in EDs to improve overall care quality and resource utilization.

Urinary tract infection (UTI) is among the most common source of serious bacterial infections in young children presenting to the emergency department (ED) (1–3). Most guidelines define a UTI diagnosis as the presence of both pyuria on a urinalysis (UA) and a positive urine culture (4–6). In practice, children who can provide a clean catch urine sample generally only have a urine culture ordered after an abnormal UA. However, in infants and young children who are not yet toilet trained, obtaining a urine sample for UTI testing can be challenging and is often collected via invasive catheterization. At many institutions, these catheterized samples are simultaneously sent for both UA and urine culture to avoid the need for repeat invasive testing. Ordering both tests together from catharized samples can sometimes lead to a urine culture becoming positive despite a normal UA and thereby leads to a clinical conundrum of whether to diagnose a child with a true UTI and the subsequent downstream impacts (7,8).

Prompt and accurate diagnoses of UTIs in young children are challenging for many reasons. As the turnaround time for a positive urine culture is generally 1 to 2 days after a healthcare visit, an empiric UTI diagnosis and treatment is frequently based on the UA results alone and only later confirmed once the urine culture returns. Second, for children less than 6 months of age, the literature is conflicting surrounding UA reliability, and there is not enough literature regarding dipstick UA versus urine microscopy for this age group. Despite the relative sensitivity of the UA (6,9,10), older studies have shown that UA is potentially less reliable in children less than 6 months of age due to frequent bladder emptying not allowing enough time for conversion of dietary nitrates to nitrites by bladder bacteria (11). Additionally, while noninvasive UTI screening techniques can be beneficial, reliability of noninvasive samples are variable and still require further study in younger infants. For instance, the false-positive culture results of bag collection are too high to be clinically acceptable (4,5). Once obtained, a positive urine culture can not only indicate a UTI diagnosis (1), but can also represent asymptomatic bacteriuria or a contamination, rather than a true UTI (9).While a UA can help distinguish these, the sensitivity of leukocyte esterase for UTI is not 100%, and the rate of positive urine culture without an abnormal UA is roughly the same as rate of asymptomatic bacteriuria and contamination (2–3%) (12).

Current guidelines recommend for all children with a suspected UTI, based on UA results and clinical suspicion to have a urine culture sent (13–16). However most guidelines, do not clearly address interpretation of a negative UA obtained via catheterization when a urine culture is collected automatically, except for the AAP (American Academy of Pediatrics) febrile infant less than 60 days (12) that does not recommend automatic urine cultures. To better understand the clinical implications and to address these conflicting recommendations from newer guidelines, our study aimed to describe the clinical course of children with normal UA and positive urine culture sent from catheterization, and to better ascertain whether their clinical features are suitable for a true UTI diagnosis or alternatively contamination or asymptomatic bacteriuria.

METHODS

Study design and setting

A single-center, retrospective cohort study was performed in the pediatric ED over a 6-month period between January 1, 2019, and June 30, 2019, at the Hospital for Sick Children (HSC). HSC is an urban, tertiary care children’s hospital serving Southern Ontario, Canada, with approximately 80,000 ED visits per year. During this time period, most children not yet toilet trained, who had suspicion for UTI in our ED had a urinary catheterization performed and the urine sample automatically had POC UA completed by a nurse and sent for culture to the laboratory, regardless of the UA results.

Selection of participants

Inclusion criteria included patients who had a documented point-of-care (POC) UA (Clinitek Status, Siemens Healthcare, Munich, Germany) in the ED and urine culture sent to the laboratory on a catheterized urine specimen. These tests were performed at the discretion of the treating physician during the ED visit. Due to the theoretical physiology differences in children less than 6 months of age (11) that might change the sensitivity and specificity of POC UA (17), we used an age cutoff of 6 to 24 months. Patients with the following criteria were also excluded: Urine samples for culture were collected by any means other the urine catheterization; known genitourinary (GU) abnormalities, immunodeficiency, medical complexity or unstable, suspected complicated UTI (4), or possible sepsis concerns needing immediate antibiotic treatment, documented treatment with an antimicrobial agent within 1 week prior to obtaining the urine culture.

Definitions

A positive dipstick was defined as positive POC UA for either any leukocyte esterase (trace to large) and/or positive nitrite (10). A positive culture for a specimen obtained via catheterization was defined as potential urinary pathogen with CFU count ≥50,000/ml, according to the AAP guidelines (7,18).

Follow-up urine culture process

In our ED, follow-up phone calls are made for every child with a positive urine culture, and negative cultures (if a child was misdiagnosed with a UTI and antibiotic treatment incorrectly started), either by a nurse practitioner, physician assistant, or trained nurse with staff physician oversight. For positive cultures, the follow-up is usually completed once sensitivities return, unless the child was not initially diagnosed with a UTI at the ED visit, and then the patient is contacted at the time of a positive culture result, often prior to sensitivities. The management of these patients with a positive culture (including the need for antibiotic treatment) is left to the individual discretion of the clinician contacting the family after symptom reassessment.

Data collection and analysis

Charts were reviewed retrospectively by pediatric emergency medicine fellow (I.K), and data were extracted from the hospital’s electronic medical record, Epic (Epic Systems Corporation, Verona, WI), onto a standardized data collection form (Excel 2016 version 2203) on a secure web-based database. The following parameters were extracted: age, sex, medical history, antimicrobial treatment during the week prior to ED visit, chief complaint, signs and symptoms, reason for urinary sample collection, method of urine collection, POC UA results, disposition, antibiotics prescribed during ED visit or during admission, urine culture pathogens and colony count, and follow-up call documentation.

RESULTS

A total of 818 urine cultures were collected via catheterization during the study period for children ages 6 to 24 months. One hundred and thirty-three POC UA were positive (16%) and 685 were negative (84%). One hundred and thirty-one (16%) of the urine cultures met criteria as positive. Within this group of 131 positive cultures, 26 (19.8%) patients had a negative POC UA. After reviewing the medical charts of these 26 patients, 12 patients met exclusion criteria (Figure 1) leaving a total of 14 (10.6%) patients with a negative POC UA and a positive culture (Table 1). Of these, 13 presents to the ED due to a febrile illness and 1 for a complaint of diarrhea.

Of the 14 patients with a positive culture, 9 (64%) had no fever in the follow-up call documentation and antibiotic treatment was not prescribed during the initial visit (Figure 2). Of the remaining five patients, two (15%) were admitted, but not directly related to suspected UTI: one was admitted during the initial ED visit for respiratory distress and the urine culture had been obtained for fever. Antibiotic treatment for a UTI was only started as an inpatient when the culture was reported as positive with Escherichia coli, however, the patient was already afebrile before antibiotic was started. The second patient returned to the ED with fever and a complete blood count (CBC) showed neutropenia prompting hospital admission. A urine culture was repeated and repeat result showed no growth (negative). The patient was ultimately diagnosed as viral-induced febrile neutropenia and not a UTI. The other three patients (21%) with a negative POC UA still had a fever on the day of the follow-up call (Table 2). These three patients, representing 2% of the overall positive urine cultures, were prescribed antibiotic treatment for suspected missed UTI after the culture returned positive. None of these three patients returned to the ED or were admitted.

DISCUSSION

In this study, we explored the clinical course and association between negative POC UA results and subsequent laboratory-positive urine cultures among children aged 6 to 24 months who presented to the ED and had a urinary catheter sample obtained as part of their diagnostic process. In those patients, a small percentage had a positive urine culture and the majority of these patients were asymptomatic when the culture returned positive suggesting a more likely diagnosis of asymptomatic bacteriuria or contamination and not a true UTI. No adverse outcomes were documented in these patients despite no empiric antibiotic treatment.

• 1 Of the three patients with a possibly true positive urine culture and negative POC UA, two patients, both 9-month-old males, only had a 1-day history of isolated fever when the urine sample was collected. While a 1-day fever history is generally not an indication for UTI screening according to the AAP and CPS (Canadian Paediatric Society) recommendations (4,6), it is sometimes still completed in busy EDs when outpatient follow-up is uncertain or is felt to potentially mitigate future ED return visits. Both of these patients’ cultures grew E. coli, one of which demonstrated extended-spectrum beta-lactamase (ESBL) resistance. This patient had a history of phimosis and the antibiotic prescribed during the initial follow-up call when sensitivities were not yet reported would not appropriately treat the ESBL resistance. Despite this resistance pattern, the patient’s fever had already dissipated when the resistance profile was resulted. Additionally, although phimosis was not included as one of the GU abnormalities for exclusion criteria in our study, prior studies have shown higher risk of UTIs with phimosis under 1 year of age (19). Finally, the last of these three patients was a 19-month-old female with a history of previous UTI’s who presented with 7 days of fever and the culture was positive for Klebsiella species. Despite a history of recurrent UTI’s, she had no radiologic workup ever done to assess for a GU abnormality that could be a risk factor for UTIs and part of our study’s exclusion criteria. Importantly, all three of these patients did well despite their delayed diagnosis and treatment of UTIs. Evidence supports that short-term delays in antibiotic treatment in low-risk children are generally not associated with increased renal scarring or urosepsis (20). Notably it is felt that the pyuria as opposed to the bacteriuria mediates scarring (21).

If cultures were only sent for children with positive POC UA’s after catheterization, we could decrease the number of cultures sent by 83% (685/818) with a low risk of missing clinically significant, true UTIs. This practice change could lead to a significant impact on patient care, downstream impacts, and healthcare resource utilization, which is more relevant in the post-COVID-19 pandemic than ever. From a laboratory perspective, urine cultures take time and require skilled technicians. Additionally with a turnaround time of 24 to 48 h, there must be additional healthcare resources to follow up the urine culture results with families and their subsequent management. Finally, positive urine cultures that are not true UTIs, can lead to unnecessary antibiotic exposure, side effects and subsequent antimicrobial resistance. In an era of rising healthcare costs and resource constraints, this approach could potentially save important resources while ultimately providing better quality of care.

In the most recent AAP clinical practice guidelines for febrile infants, it is recommended that a urine sample be obtained via SPA or transurethral catheterization for febrile infants aged 8 to 21 days, and only sent for a urine culture if UA result is positive (12). To our knowledge, there has been limited evidence to support this recommendation in such young infants and questions the longstanding practice of sending cultures on all urine samples obtained via invasive techniques. Notably, our study provides further support for this practice change to stop automatic urine cultures from all catheterized specimens and instead using a risk based, guided approach to sending cultures post catheterization for UTI screening.

Our findings are consistent with the literature that positive catheter-obtained urinary cultures among febrile neonates and infants with negative POC UA’s may mostly represent contamination or asymptomatic bacteriuria, rather than a true UTI (22–24). When assessing UA and POC UA results in patients with true UTI cases, a 2018 PECARN study found that 94% of febrile infants with UTI’s will have UA positive for leukocyte esterase (6). The sensitivity is even higher for UTI associated with bacteremia (97.6%) (9). In young children between 90 days and 2 years of age, the sensitivity of catheter UA samples compared to culture as gold standard was 0.83 (25), but it is unclear how many of these were between 3 and 6 months of age and whether the sensitivity for older infants over 6 months is higher due to perceived physiological differences as in our cohort. In 2016, a two-step process for ED UTI screening approach was suggested by CHOP (Children's Hospital of Philadelphia) (26) and adopted by many centers whereby if a UA is negative from a bag collected specimen, then there is no need for further testing or obtaining a culture in children. These studies, along with a few others, are reflected in the recent AAP well appearing febrile infants (8–60 days) guideline (12) recommending that urine specimens collected via catheterization or SPA (suprapubic aspiration) only be sent for culture if the UA is positive. Given that the UA sensitivity and specificity may vary more than our study’s cohort due to our age limit inclusion criteria (11), our results are certainly supported by these guidelines and should be considered in future UTI guidelines.

A recent study found that pyuria and leukocyte esterase are not sensitive markers to identify non-E. coli UTI in children under 24 months (27). However, unlike our study, the clinical outcome was not addressed and they included children with GU abnormalities or recent antibiotic treatment. The strength of our study lies in knowing the clinical outcomes of the patients, not just the test results.

Our study has several limitations. First, this is a retrospective study based on the medical records and documentations of follow-up calls. While some patients could have returned to other community hospitals or clinics, we are the only tertiary care children’s hospital and pediatric intensive care unit within the area. Also, due to retrospective chart reviews, the exclusion criteria applied only to the group with suspected false-positive urine cultures and not all groups with urine cultures. A second limitation is potential selection bias by including all patients with urine culture and POC UA sent from the ED. This may have led to patients being included that had urine checked for reasons other than suspected UTI. A third limitation is that this is a single-center study, so even though the population presenting to our center is varied, there may be different regional and geographical variations to consider.

In conclusion, previously healthy children aged 6 to 24 months with a negative UA obtained via urinary catheterization rarely have true UTIs. The routine practice of sending automatic urine cultures from catheterized urine samples in the ED regardless of POC UA results should be reconsidered in all centers for this patient population. Noninvasive urine collection methods for UTI screening in young children should be routinely adopted to reduce unnecessary urine cultures and invasive procedures in children while saving important healthcare resources. Further iterations of this work should include patient populations less than 6 months of age and possibly patients with known GU abnormalities to potentially further expand the impact.

DATA AVAILABILITY

All the database is saved and coded with no personal identifying findings with the main researcher of the study I.K.

ETHICAL APPROVAL

The study received approval by our institution as part of a Quality Improvement study by the Quality Risk Management Committee who granted QI approval and met requirements for Research Ethics Board (REB) exemption. This retrospective chart review study was in accordance with the ethical standards of our institution and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

FUNDING

No funding to report.

POTENTIAL CONFLICT OF INTEREST

All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References