QUALITY OF REPORTING OF ABSTRACTS OF PEDIATRIC RANDOMIZED CONTROLLED TRIALS

Purpose: To assess the quality of reporting of abstracts describing randomized controlled trials (RCTs) in four major pediatric journals.

Methods: Pediatric RCTs from four journals - The Journal of Pediatrics, Pediatrics, Pediatric Research, and European Journal of Pediatrics- were selected using a MEDLINE search. The resulting 299 abstracts were sorted for eligibility and coded based on a standardized abstraction manual constructed from the criteria of CONSORT for Reporting Randomized Controlled Trials.  Eligible trials were coded independently by members of the research team.  Verification checks of each coded element were performed by a second team member, and disagreements were handled by consensus.   Two hundred twenty abstracts comprised the final sample.

Results: Of the 220 analyzed abstracts, one hundred eleven (50.45%) had the word “randomized” in the title.  Nearly a quarter of the abstracts (54/220; 24.55%) did not provide contact information for any of the authors.  Most articles (177/220; 80.45%) failed to describe the specific trial design (139/220; 63.18%).  Twelve RCTs (5.45%) did not describe the participant eligibility, and one hundred sixty-six (166/220; 75.45%) did not describe the setting where the data were collected.  Only 91 abstracts (41.36%) clearly defined the primary outcome with an even smaller percentage (32/220; 14.55%) defining secondary outcomes.  For those abstracts reporting a primary outcome, few presented an effect size (41/220; 18.64%) or a confidence interval (35/220; 15.91%).  Three (3/220; 1.36%) described the method of randomization, and none described the method for concealing the allocation sequence.  Blinding was mentioned in 61 abstracts (61/220; 27.73%) with most of those not reporting who was blinded (45/61; 73.77%).  Randomized sample size was not reported in 134 abstracts (60.91%); analyzed sample size was not reported in 55 abstracts (25.00%).  Many abstracts did not report the status of the trial (210/220; 95.45%).  Adverse events and side effects were reported in 29 abstracts (13.18%), not applicable in 64 abstracts (29.09%), or simply not reported (127/220; 57.73%).  Trial registration numbers and trial registries were mentioned in only one abstract (1/220; 0.45%), and no abstracts mentioned a funding source.

Conclusion: The quality of reporting of RCT abstracts published in pediatric journals shows room for improvement despite the publication of CONSORT guidelines in 2008.  Pediatrics journals should consider the adoption of CONSORT guidelines for abstracts to improve reporting quality.