Improving standards for paediatric clinical trials

We may say that “children are not small adults”, but do we take this sentiment to heart when prescribing medicines? Publicity about the increased risk of suicide in youth with mood disorders treated with selective serotonin reuptake inhibitors (1), and the risk of serious side effects in young children administered over-the-counter cough and cold medications (2) highlights attention that is being devoted worldwide to the lack of evidence on which to base therapeutic decisions for the care of children. Physicians who treat children often must decide between withholding treatment proven effective in older patients or using medications ‘off-label’, with insufficient information about drug dose, metabolism, known side effects or appropriate formulations. However, there is strong pressure for change, and to improve the number, quality and reporting of paediatric clinical trials. The WHO has been advocating for ‘better medicines for children’, and legislative incentives in the United States and Europe encourage the conduct of paediatric clinical trials. There is an urgent need for Canada to act with similar legislative incentives in this area. Clinical practice guidelines are based on meta-analyses and systematic reviews, which until recently, have been largely dependent on published reports of clinical trials. For years, publication bias has been known to have a major influence on the array of studies available to review, and investigators/authors have been as much at fault as journal editors in delaying or not publishing studies with negative or inconclusive results (3,4). However, there has been a major shift among journal editors to encourage the submission of well-designed clinical trials regardless of findings. Also, since 2005, registration of clinical trials has been promoted by the WHO, required by journals of the International Committee of Medical Journal Editors and increasingly demanded by research ethics boards. Registration makes publicly available the fact that a study is/has been conducted, and records the features of study design and outcome measures. It is, thus, possible to examine the difference between what was proposed at the start (registration) of a study and what was ultimately presented for publication, which guards against the tendency for authors to submit, and of journals to accept, only the statistically significant findings of a study. Medical journal editors have also been influential in standardizing the elements of clinical trial reports. The Consolidated Standards of Reporting Trials (CONSORT) statement improves the ability to assess results and to make meaningful comparisons among studies (5). There is room for continued improvement. Adult trial publications dominate the literature and are increasing at a faster rate than paediatric trials in almost all specialties (6); plenty of work needs to be done to improve methodological issues surrounding the conduct and reporting of paediatric studies (7). Not all journals stipulate adherence to reporting guidelines. Meerpohl et al (8) found that many paediatric journals do not include recommendations that aim to improve publication practice in their author instructions. Moreover, they found that approximately one-fifth of journals did not require authors to disclose conflicts of interest at the time of manuscript submission, and more than three-quarters did not require/recommend trial registration. In 2009, Mathieu et al (9) compared the registered and published primary outcomes of 323 randomized controlled trials conducted in adults, and revealed that fewer than one-half of the studies were adequately registered, and that there was selective reporting of outcomes favouring statistically significant results. A review of 300 paediatric trials published in 2007 (10) found that only 23% had been registered and only 8% were at a low risk of bias (the likelihood of distortion of treatment effects in clinical trials). Several reviews indicate that the risk of bias is prevalent in paediatric clinical trials, although lower in studies that are registered (11). There is also concern that the CONSORT elements do not fully capture study considerations that are important to children such as the nature of a drug formulation (tablet or bubblegumflavoured liquid?), length of follow-up to assess long-term impacts such as effects on growth and development, and whether parents were involved in planning the study (can the child still go to school or play with his or her friends during the study?) (12). National and international organizations are forming to enhance the design, conduct and reporting of trials involving children. Standards for Research in Child Health (StaR Child Health; www.ifsrc.org) is a global initiative that brings together an international group of leading methodologists, clinicians, regulators, funders and decision makers who aim “to enhance the quality, ethics and reliability of paediatric clinical research by promoting the use of evidence-based standards or guidance for clinical studies with children” (13). The first international StaR Child Health Summit was held in Amsterdam, the Netherlands, in October 2009, and a second meeting was held in Vancouver, British Columbia, in September 2010 (see Summit reports at www.ifsrc.org). International StaR Child Health teams have formed Standards Development Groups that are addressing topics such as adequate sample sizes, the need for data safety and monitoring committees, risk of bias, and recruitment and consent. StaR Child Health and others are working on adaptations to the CONSORT statement to include child-relevant elements. Global Research in Paediatrics (GRIP) is an international research consortium, launched in February 2011, with a focus on the development of a paediatric clinical pharmacology training program. Closer to home, Canada’s Maternal Infant Child www.micyrn.ca) brings together 17 Canadian child/maternal-child health research organizations into a national entity committed to fostering and enhancing national and international collaborations by removing barriers and building capacity for the conduct of high-quality health research. A MICYRN review of registered clinical trials – viewable through the WHO International Clinical Trials Registry Portal (www.who.int/ictrp/child/search/en) – indicates that more than 580 paediatric clinical trials are actively recruiting at sites in Canada, and more than 1000 paediatric clinical trials have been conducted in the past five years, making Canada second only to the United States in world paediatric trials activity. The MICYRN organizations are committed to building an integrated platform to support perinatal and paediatric clinical trials in Canada and beyond, and a first meeting to this end was held in May 2011. MICYRN is partnered with StaR Child Health, and is a member of the European Medicines Agency’s Network of Paediatric Research Coordinating Group, along with other national and international subspecialty paediatric research networks striving to facilitate research into medicines for children. Why should and how can paediatricians contribute? First, better trials mean better care and better outcomes for kids. Therefore, paediatricians should engage in clinical trials research, either through investigation, collaboration or encouragement of children and families. Second, we need to support journals in the promotion of guidelines reporting clinical trials. Authors’ conflicts of interest must be declared; there should be documentation of clinical trial registration before commencement of study enrollment; and the CONSORT report format should be followed. Third, when we are involved in the development or adoption of clinical practice guidelines, we should expect scrutiny of trial registration information to determine whether unreported studies have been conducted and whether there has been selective reporting of outcomes favouring statistically significant results.