From Research Question to Conducting a Randomized Controlled Trial on Continuous Antibiotic Prophylaxis in Prenatal...

Continuous antibiotic prophylaxis (CAP) use to prevent urinary tract infections (UTI) in infants with prenatal hydronephrosis (HN) remains controversial. Lack of consensus guidelines and diverse practice patterns for postnatal management of HN highlight the dire need for higher level of evidence studies. Herein, we aim to describe the steps from developing a well-defined research question to execute a multicentered randomized controlled trial (RCT) to address the issue of CAP use in patients with prenatal HN.

The steps involved were (1) choosing the proper research question, (2) survey of practice patterns and establishing clinical equipoise, (3) systematic review of the literature, (4) reviewing own practice, (5) longitudinal prospective study, (6) pilot study, (7) cost-utility analysis, and (8) definitive RCT (clinical trials registry number: NCT01140516). An update of our previous systematic review was conducted using two electronic databases and gray literature from 2010 to 2015. Eligibility criteria included studies of children <2 years old with postnatally confirmed prenatal HN, receiving CAP or not, and reporting on development of UTIs, capturing information on voiding cystourethrogram result and HN grade. Full-text screening was conducted by two independent reviewers. UTI rates in patients with high-grade HN were compared across different study designs. Finally, blinded comparative analysis of UTI rates between placebo and treatment groups was carried out using chi-square test.

UTI rates in patients with high-grade HN by their respective study design were: 25% for systematic review, 20% for retrospective study, 21% for prospective and pilot studies, and 13% for the definitive RCT thus far. Regardless of the type of study design, patients with hydroureteronephrosis had significantly higher (threefold to sixfold) UTI rates than those with isolated HN. Our updated systematic review yielded 486 citations, of which 9 (n = 1987 infants) observational studies met eligibility criteria.

UTI rates in patients with high-grade HN dropped from 25% in observational studies to 13% in our RCT. This decline in UTI rate demonstrates that study designs lacking strategies to minimize bias are more prone to overestimate treatment effects. These findings highlight the importance of conducting methodologically sound RCTs to answer clinically meaningful questions, such as the one presented here.