Pediatric UTIs: Short-Course vs. Standard-Course Antibiotics — Is It Time for a Change?

Background: There is a shifting paradigm towards shorter durations of antibiotics in pediatric infections. Conflicting international guidelines recommend treatment of urinary tract infection (UTI) with antibiotic courses ranging from just 3 days to 7–14 days.^1–4 Antimicrobial resistance is a global health crisis, underscoring the importance of antibiotic stewardship. Investigators in the SCOUT Trial examine the impact of short-course (5 day) antibiotic therapy in UTI, with potentially far reaching implications.

Click here for Direct Download of the Podcast

Paper: Zaoutis T, Shaikh N, Fisher BT, et al. Short-Course Therapy for Urinary Tract Infections in Children: The SCOUT Randomized Clinical Trial. JAMA Pediatr. 2023;177(8):782-789. PMID: 37358858

Clinical Question: In children aged 2 months to 10 years who exhibit clinical improvement after 5 days of antimicrobial therapy for UTIs, is short-course (5 days) oral antimicrobial therapy as effective as standard-course (10 days) therapy in preventing treatment failure and related outcomes?

What They Did:

• Study Design: Multicenter, randomized, double-blind, placebo-controlled, non inferiority clinical trial.
• Enrollment: Recruited patients from primary care offices, emergency departments, and inpatient wards at Children’s Hospital of Philadelphia and UPMC Children’s Hospital of Pittsburgh.
• Summary: Potentially eligible children being treated for UTI who were afebrile and had no clinical signs or symptoms of UTI were assessed in person on days 2–5. After consent, children were randomized 1:1 to receive 5 additional days of antibiotics or 5 days of a matching placebo. Investigators assessed clinical outcomes via 2 in-person visits on days 11–14 and days 24–30.

• Definitions: 
  • UTI: Presences of all 3 of the following criteria:
    1. One or more of the following: fever, pain in the suprapubic, abdominal, or flank area; urinary frequency, urgency, or hesitancy; dysuria in children ≥2 years old; in children < 2 years old, poor feeding and vomiting.
    2. Pyuria: ≥10 WBC/mm3 or ≥5 WBC/HPF (centrifuged specimen) or leukocyte esterase ≥ trace on dipstick urinalysis.
    3. Positive urine culture—growth of a single uropathogen at counts ≥5 x 104 or higher per CFU/ml (suprapubic aspiration or catheterized s specimen) or 105 or higher CFU/ml (clean voided specimen)

• • Asymptomatic Bacteriuria: positive urine culture with the same colony counts as above and without regard to pyuria, but in the absence of symptoms.
• Funding: The National Institute of Allergy and Infectious Diseases, Division of Microbiology and Infectious Diseases funded the study.
• Trial Registration: ClinicalTrials.gov: NCT01595529

Population:

Inclusion Criteria:

• Children aged 2 months to 10 years.
• Diagnosed with UTI.
• Prescribed one of five frequently used antimicrobials:
  • Amoxicillin-clavulanate
  • Cefdinir
  • Cefixime
  • Cephalexin
  • Trimethoprim-sulfamethoxazole
• Recruited from primary care offices, emergency departments, and inpatient wards.

Exclusion Criteria:

• Recovery of a second uropathogen (>104 colony forming units [CFU] per mL for samples collected by suprapubic aspiration or catheterization, or >5 × 104 CFU per mL for samples collected by clean catch) at the time of diagnosis.
• Hospitalization for bacteremia.
• Admission to the intensive care unit.
• Urine culture yielding a pathogen resistant to the initially prescribed antimicrobial.
• Catheter-associated UTI.
• History of UTI within 30 days.
• Phenylketonuria.
• Congenital or anatomic abnormalities of the genitourinary tract 
• Previous genitourinary tract surgery (except circumcision).
• Inability to tolerate oral medications.
• Presence of an immunocompromising condition.
• Type I hypersensitivity or anaphylactic reaction to study products.
• Previous enrollment in the study.
• Enrollment in another therapeutic drug study (excluding vaccine studies).
• Gestational age of <36 weeks (only for children younger than 2 years of age at enrollment).
• Not available for follow-up visits.

Intervention

• Short-course therapy: Patients received 5 days of antibiotics and 5 additional days of a matching placebo.

Comparator

• Standard-course therapy: Patients received 10 days of antibiotics.

Outcome

Primary Outcome:

• Treatment Failure: Defined as the occurrence of UTI between day 6 and the day 11–14 visit.

Secondary Outcomes:

• UTI at any time after the day 11–14 visit.
• Asymptomatic bacteriuria at the day 11–14 visit.
• Gastrointestinal colonization with antimicrobial-resistant Escherichia coli or Klebsiella pneumoniae.
• Clinical symptoms of UTI between day 6 and the day 11–14 visit.
• Positive urine culture between day 6 and the day 11–14 visit.
• Proportion of children with reported adverse events between days 6 and 38–44.

Results:

Study Population: 

• 32,271 children were screened.
  • 20,391 children were excluded due to negative cultures.
  • 4,391 children were excluded for “other” reasons.
  • 1175 parents could not be reached.
  • 933 children did not receive 1 of the accepted antibiotics.
  • 103 children were excluded because their symptoms had not resolved at day 5.
  • 986 parents declined participation.
• 693 children were randomized.
  • 348 allocated to Standard-course therapy.
    • 20 excluded after randomization.
  • 345 allocated to Short-course therapy.
    • 9 excluded after randomization.
• Median age of children enrolled was 4 years (Range 2mo to 10 years)

Primary Outcome: 

• Standard-course (10 day) therapy had a lower rate of treatment failure (0.6%) compared to short-course therapy (4.2%).
• The upper limit of the CI for the absolute difference was 5.5%, which exceeded the prespecified 5% margin.
• Treatment failure was not related to age, fever, prescribed antimicrobial therapy, or study site.
• Short-course therapy DID NOT prove to be statistically noninferior to Standard-course (10 day) therapy.

Secondary Outcomes: 

• Short-course therapy was associated with higher rates of asymptomatic bacteriuria and positive urine cultures at the day 11–14 visit. 
• No significant differences were observed in other secondary outcomes or adverse events.

Strengths:

• Investigators asked a patient-oriented research question.
• The double-blind, randomized, placebo-controlled trial study design is the gold standard for medical literature and helps to limit bias.
• Investigators appropriately employed a noninferiority trial design to assess whether a new shorter course of an existing therapy, with the potential for fewer side effects, cost savings, and lessening the global burden of antimicrobial resistance, is not worse than the standard length treatment. 
• Researchers chose a narrow noninferiority threshold of 5%, making it more difficult to prove noninferiority.
• They used strict criteria to diagnose UTI, ensuring that the population being studied was likely to have the disease in question.
• Researchers enrolled patients in multiple clinical environments, including the ED, hospital wards, and primary care offices, which increases generalizability.
• Investigators enrolled a large sample size, increasing external validity.
• There were no major differences in baseline demographics, indicating that randomization was sound and patients had a similar prognosis at the start of the trial.
• The placebo had the same volume, appearance, flavor, consistency, and bottle as the study drug, decreasing the possibility of unblinding.
• Trial participants, their parents, investigators, and trial personnel were blinded to the treatment assignment.
• Investigators analyzed patients with an intention-to-treat and per-protocol analysis, allowing for a comprehensive assessment of the results.
• Loss to follow-up was very low and similar in each cohort.

Limitations:

• A single-country study limits generalizability to patients outside the United States.
• The definition of UTI used by investigators included some clinical signs typically attributed to pyelonephritis, which most clinicians would treat with a longer duration of antibiotics.
• Enrollment required treatment with 1 of 5 predefined antibiotics, limiting generalizability to patients prescribed other antibiotics.
  • >900 patients were excluded due to treatment with other antibiotics.
• Follow up required an in-person visit (at child’s home or at an affiliated clinic) on days 2–5, and two additional in-person visits on days 11–14 and days 24–30, potentially limiting the generalizability to an ED population outside of a clinical trial.
• There were some important demographic considerations which may further limit generalizability: 
  • 96% of patients were female
  • 64% were white
  • 91% were non-Hispanic
  • 58% were aged 2–6 years old
  • Very few children in the 2 mo – 23 mo age range
  • 87% of patients were treated with cephalosporins
  • E. coli was the isolated pathogen in 90%
• 32,000 patients were initially assessed, but only 693 were enrolled. More than 20,000 patients were excluded due to negative urine cultures, while an additional 4,300 were excluded for unspecified “other” reasons, raising suspicion of potential selection bias.
• The trial protocol called for assessment of the primary outcome between day 6 and the 11–14 day visit while some patients in the standard therapy group were still receiving antibiotic treatment, which introduces detection bias.
• Investigators achieved much better failure rates than they initially assumed. In a noninferiority trial, inaccurate assumption rates may make it easier for the new treatment to meet the noninferiority limit.
• Researchers excluded patients with missing data from the primary outcome. Patients lost to follow-up are known to have worse outcomes than those who completed the study. While seemingly a relatively small number, the 9 patients lost to follow-up in the standard-course therapy arm would be 4.5x larger than those who failed treatment.
• Researchers had limited information on adherence to originally prescribed antibiotics on days 1–5.

Discussion:

Inside The Numbers:

In the intention-to-treat (ITT) analysis, the upper bound of the 95% CI for treatment failure (5.5%) exceeded the 5% noninferiority threshold. ITT analysis includes all patients based on their original randomization, regardless of adherence to the trial protocol, generally providing a more cautious estimate of treatment benefit. However, hypothetical inclusion of many non-adherent patients in a control group could potentially lead to a false interpretation of statistical noninferiority. In the SCOUT Trial, overall adherence was good after randomization during days 6–10, but we know little about adherence during the initial 5 day treatment period.

In a noninferiority trial the per-protocol analysis holds equal importance. Eliminating non adherent patients introduces some bias, but can provide reassurance that patients received the appropriate treatment. In the per-protocol analysis of patients who adhered to at least 80% of their assigned study therapy, the upper bound of the 95% CI for treatment failure was 3.9%—below the preset 5% noninferiority margin.

The Relative Risk of 6.8 indicates that patients who received 5 days of antibiotics were nearly 7 times more likely to have a treatment failure compared to 10 days of therapy. However, cautious interpretation is warranted as relative risk can exaggerate the magnitude and impact of the results. Moreover, the Absolute Risk Difference of 3.6 indicates that approximately 4 more patients out of 100 would experience a treatment failure with short-course (5 day) therapy compared to standard-course (10 day) therapy.

Hidden In Plain Site:

The SCOUT trial included only patients who demonstrated clinical improvement by day 5 of antibiotic treatment. While this cautious approach prioritizes safety, information regarding response to therapy would be unavailable to most ED clinicians outside of a clinical trial. Consequently, implementing this strategy in a real-world clinical setting could be challenging, as clinicians making decisions about antibiotic treatment may lack a reliable means of assessing whether a patient would improve by day 5.

Utilizing strict exclusion criteria, investigators eliminated patients with negative urine cultures. More than 20,000 patients prescribed antibiotics DID NOT have a UTI,  highlighting the sheer magnitude of the problem and difficulty associated with diagnosing UTI in a pediatric population. Patients with nonspecific findings such as fever and trace leukocyte esterase on a urine dipstick would meet 2 of 3 inclusion criteria. The third criterion, a positive urine culture, would not be available immediately to inform real-time clinical decisions. In a clinical trial, ensuring the cohort under investigation has the disease in question is pivotal to obtain accurate data and guide clinical recommendations. Yet, in practice, clinicians ubiquitously treat patients based on probability and risk assessment when diagnostics are unavailable. A pragmatic trial, with less stringent inclusion criteria, investigating all patients treated for UTI would presumably yield more favorable results, albeit with the inclusion of many patients without infection. Nonetheless, such a trial would be more reflective of actual clinical practice and the uncertainty embedded in point-of-care clinical-decision making.

Shared Decision Making:

Short-course (5 day) therapy DID NOT prove to be statistically non inferior to standard-course (10 day) therapy. However, exceeding the noninferiority margin by 0.5% would result in one additional treatment failure in 200 patients. Clinicians willing to accept 5 treatment failures per 100 patients might also accept 5.5. 

A 5.5% failure rate in patients diagnosed with UTI may be unpalatable for some clinicians and parents. However, in the SCOUT Trial, thousands of patients prescribed antibiotics had negative urine cultures, underscoring the fact that many patients treated with antibiotics will not have a UTI. Furthermore, very few clinical institutions have the resources to call patients with a negative urine culture and discontinue unnecessary antibiotic therapy.

When making decisions with incomplete information, short-course antibiotics seem appropriate, and shared-decision making is required. If prescribing short-course therapy, it is prudent that parents understand the increased likelihood of failure and the need for close follow-up to determine if their child needs a longer course of antibiotic therapy.

Author’s Conclusion: “In this noninferiority trial, short-course therapy was not statistically noninferior to standard-course therapy. However, given the infrequent occurrence of treatment failure in the short-course group and the results of a post hoc analysis, short-course therapy may be considered a reasonable option for children with clinical improvement after 5 days of antimicrobial treatment.”

Our Conclusion:

The SCOUT trial did not demonstrate statistical noninferiority for short-course therapy compared to the standard-course in pediatric patients with UTI. However, considering the low failure rate and the diagnostic uncertainty embedded in clinical decision making, short-course therapy may be a reasonable option. It is crucial for clinicians and parents to engage in shared decision-making, recognizing the slightly higher risk of treatment failure with short-course therapy and the need for close monitoring to assess the necessity of a longer antibiotic course.

Clinical Bottom Line:

Consider short-course antibiotic therapy for treatment of pediatric patients with UTI. Engage in shared decision-making with parents, and prioritize close follow-up to assess for treatment failure.

References:

1. National Institute for Health and Care Excellence.Urinarytract infection in under 16s: diagnosis and management (NICE guideline 224). 2022. Accessed 24 October, 2023. Available from: https://www.nice.org. uk/guidance/ng224/ resources/urinary-tract-infection-in-under-16s-diagnosis-and-management-pdf- 66143835667141
2. Buettcher M, Trueck J, Niederer-Loher A, et al. Swiss Consensus Recommendations on Urinary Tract Infections in Children. Eur J Pediatr. 2021;180:663-74. PMID: 32621135
3. SubcommitteeonUrinaryTractInfection,SteeringCommittee on Quality Improvement and Management; Roberts KB. Urinary Tract Infection: Clinical Practice Guideline for the Diagnosis and Management of the Initial UTI in Febrile Infants and Children 2 to 24 Months. Pediatrics. 2011;128:595-610. PMID: 21873693
4. Robinson JL, Finlay JC, Lang ME, Bortolussi R; Canadian Paediatric Society, Infectious Diseases and Immunization Committee, Community Paediatrics Committee. Urinary Tract Infections in Infants and Children: Diagnosis and Management. Paediatr Child Health. 2014;19:315-25. PMID: 25332662
5. Zaoutis T, Shaikh N, Fisher BT, et al. Short-Course Therapy for Urinary Tract Infections in Children: The SCOUT Randomized Clinical Trial. JAMA Pediatr. 2023;177(8):782-789. PMID: 37358858

Post By:

Marco Propersi, DO FAAEM
Vice-Chair, Emergency Medicine
Assistant Emergency Medicine Program Director
Vassar Brothers Hospital, Poughkeepsie, New York
Twitter/X: @marco_propersi

Post Peer Reviewed By: Salim R. Rezaie, MD (Twitter/X: @srrezaie)