Journal of Pediatrics
May 2017 Volume 184, p1-246
The Editors’ Perspectives
Pediatric clinical trials—number needed to recruit
Denise M. Goodman
Published in issue: May 2017
As pediatricians we often extrapolate findings from adult clinical trials to comparable pediatric populations, acknowledging that this is an imperfect approach while bemoaning the lack of adequately powered pediatric trials. There are many reasons for the difficulties in completing clinical trials in children. For instance, the prevalence of certain conditions may be lower, and outcomes different, than for adults. By way of example, mortality is not infrequently used as an end point for adult studies of intensive care unit patients, but mortality in the pediatric intensive care unit is low and thus, an insensitive marker by which to assess the success of many interventions. In addition, there are special considerations in advancing studies in the vulnerable population of children, including a more complex informed consent process and potential reluctance on the part of parents as surrogate decision makers.
In this context, any empiric data regarding the execution of clinical trials in children is helpful. In this volume of The Journal, Schandelmaier et al report on premature discontinuation of clinical trials in children. They drew from studies approved by 6 research ethics committees in 3 countries. By using this approach they minimized issues with incomplete registration and reporting bias that might be present using trial registries or publications as the basis for examining trials. They also waited for as long as 10 years or more to ensure that recruitment and publication could be accomplished. These investigators found that 40% of pediatric trials, compared with 29% ofadult clinical trials, are prematurely discontinued, with slow recruitment the most common reason across the board. After controlling for other trial characteristics, such as source of funding, they found, however, that being a pediatric trial was not in andof itself an independent risk factor for discontinuation. This suggests that other features of trial implementation, such as planned recruitment targets and adequate funding, may be more important.
These findings underscore the importance of robust clinical trial design, including a realistic recruitment strategy, an adequately sized pool of potential enrollees, and sufficient support for patient screening and consenting. The evidence needed to support good clinical decision making rests on rigorous science and a disciplined approach to trial implementation. We owe both the patients we treat and those who generously participate in clinical trials no less.