Montini G, Tessitore A, Console K, et al. Short Oral Antibiotic Therapy for Pediatric Febrile Urinary Tract Infections: A Randomized Trial. Pediatrics. 2024 Jan 1;153(1):e2023062598. doi: 10.1542/peds.2023-062598. (Original study)

BACKGROUND AND OBJECTIVES: Febrile urinary tract infection (fUTI) in well-appearing children is conventionally treated with a standard 10-day course of oral antibiotic. The objective of this study is to determine the noninferiority (5% threshold) of a 5-day amoxicillin-clavulanate course compared with a 10-day regimen to treat fUTIs.

METHODS: This is a multicenter, investigator-initiated, parallel-group, randomized, controlled trial. We randomly assigned children aged 3 months to 5 years with a noncomplicated fUTI to receive amoxicillin-clavulanate 50 + 7.12 mg/kg/day orally in 3 divided doses for 5 or 10 days. The primary end point was the recurrence of a urinary tract infection within 30 days after the completion of therapy. Secondary end points were the difference in prevalence of clinical recovery, adverse drug-related events, and resistance to amoxicillin-clavulanic acid and/or to other antibiotics when a recurrent infection occurred.

RESULTS: From May 2020 through September 2022, 175 children were assessed for eligibility and 142 underwent randomization. The recurrence rate within 30 days of the end of therapy was 2.8% (2/72) in the short group and 14.3% (10/70) in the standard group. The difference between the 2 groups was -11.51% (95% confidence interval, -20.54 to -2.47). The recurrence rate of fUTI within 30 days from the end of therapy was 1.4% (1/72) in the short group and 5.7% (4/70) in the standard group (95% confidence interval, -10.4 to 1.75).

CONCLUSIONS: This study demonstrates that a 5-day course is noninferior to a 10-day course of oral amoxicillin-clavulanate.

Discipline Area Score
Family Medicine (FM)/General Practice (GP) 6 / 7
Pediatrics (General) 6 / 7
Comments from EvidenceAlerts subscribers

Dr. shmuel davidson (1/14/2024 11:08 AM)

In a very "promising" RCT with a non-inferiority design reported by Montini G et al, a multicenter trial was conducted in Italy from May 2020 through Sep 2022 to compare the efficacy of a standard ( long duration, 10 days) treatment of UTI with amoxicillin-clavulanate 50 + 7.12 mg/kg/day with a control group treated for a short period of time (5 days) with the same treatment protocol in 3-month to 5-year-old children. The primary end point was the recurrence rate of urinary tract infections within 30 days after therapy completion. The treatment duration was unblinded. At the end of the study, the recurrence rate within 30 days of the end of therapy was 2.8% (2/72) in the short group and 14.3% (10/70) in the standard group, with a difference between the groups of -11.51% ( 95% CI -20.54 to -2.47 , P<0.005). Undoubtedly, the abstract section of the publication did demonstrate a fantastic benefit of the short duration treatment of 5 days vs the long/standard (10 days) treatment. Should we immediately accept the treatment of acute pyelonephritis with the short duration (5-day) protocol? Dear practitioners in Pediatrics, please not yet.
The study's summary is not lying, of course , but it does not mention 2 facts that might turn these results from facts to artifacts:
1. the planned sample size for this noninferiority study using a clinically acceptable noninferiority threshold of 5% and a power of 80% was estimated to be 520 patient (260 per arm); 2. An interim analysis was planned by study protocol 2 years after the patients were enrolled. By means of the criteria defined by O'Brien-Fleming to stop the study early for benefit (presence in the interim analysis of a difference between the 2 groups with a P value <0.005), the study was stopped for benefit after enrolling 142 patients, 72 in the short arm and 70 in the standard 10-day arm, with a total of 12 events of recurrence in the 2 groups. I stopped reading the article at this point (page 39, upper paragraph). The possible devastating effect of over estimation of the treatment effect (random high or random low phenomena; in this case, random low) was well described by the STOPIT 2 study group led by Gordon Guyatt and was very well demonstrated in a master paper published in JAMA 2010.
I would suggest all the pediatricians taking care of children with acute pyelonephritis in this age group to stick with the results of a new randomized placebo-controlled trial published very recently in JAMA by Zaoutis T et al (JAMA Pediat. 2023) that compared 5 versus 10 days of therapy for UTI in 664 children aged 2 months to 10 years at 2 US hospitals from 2012 through 2023. The results of this study were published in Pediatics in a commentary to the current paper of Montini G et al (Charles R Woods. Pediatrics. January 24) . In the JAMA study , treatment failure by day 11 to 14 occurred in 4.2% (14/336 ) of the children assigned to the short course versus 0.6% (2/328) children assigned to a standard longer therapy with a risk difference of 3.6% (P< 0.01). It's too bad that the authors of this commentary article related the difference in the results between the 2 papers (JAMA. December 2023, Pediatrics January 2024) to variation in definitions, in population, and in procedures and not to misleading results of a study that was stopped early for benefit because of a wrong decision taken by the statisticians and investigators. Stopping early for benefit due to a random high or a random low effect was well described and documented in clinical trials in the past. The end results of this are distorted results, deviation from the truth, and artifacts instead of facts.