A descriptive analysis of pediatric trials published in 2007

Tags: Poster
Hamm M, Hartling L, Milne A, Tjosvold L, Curtis S, Vandermeer B

Background: Randomized controlled trials (RCTs) are important in informing high quality clinical care. Information is needed to improve the design, conduct and reporting of these studies for child health. Objectives: To describe a representative sample of paediatric RCTs published in 2007, with a focus on methodology, risk of bias and consistency with protocols. Methods: The Cochrane Central Register of Controlled Trials was searched and 300 paediatric RCTs were randomly selected from the results. Data extracted included publication and trial characteristics; outcomes and conclusions; methodological quality (Jadad scale, Cochrane risk of bias tool) and reporting (CONSORT); and registration and protocol characteristics. Trial registration and protocol availability were determined for each study based on the publication, an internet search and author follow-up. Results: Most studies were efficacy trials (82%) with parallel designs (90%), 40% evaluated drugs and 29% were placebo-controlled. The most common diagnostic categories were neonatal (9%), oral health (7%) and developmental, psychosocial and learning (7%). Primary outcomes were explicitly reported in 39% of trials. Statistically significant results were reported in 75% of trials, 65% of which favoured the treatment group. Of trials with a declared source of funding, 35% were supported by industry. Less than half (42%) reported adverse events. Trial registration was rare, with 12% declaring registration in the publication and 23% found online. Specification of primary outcomes in the registry was high at 88%. The overall risk of bias was high in 65% of trials and low in 6%. Analyses are ongoing to compare methodological quality with a sample of published trials from earlier years, and to measure the association between risk of bias and magnitude of effect estimates. Conclusions: This study provides a description of recently published trials in child health. The results will provide directions for continued improvement in the design, conduct and reporting of child health research.