Improving Child Health Research: The Role of Randomized Clinical Trials

The importance of randomized clinical trials (RCTs) in improving health outcomes cannot be disputed. Unlike virtually all other study designs, well-conducted RCTs minimize and potentially eliminate confounding by unknown factors. Causality can be concluded from the results of RCTs. As stand-alone articles, they influence practice. For example, the recent publication of the RCT of mRNA vaccine for severe acute respiratory syndrome coronavirus 2 in adolescents led to approval by the US Food and Drug Administration, endorsement by the Centers for Disease Control and Prevention, and saved millions of adolescents from infection and the consequences of coronavirus disease 2019.1Frenck Jr., R.W. Klein N.P. Kitchin N. Gurtman A. Absalon J. Lockhard S. et al.C4591001 Clinical Trial GroupSafety, immunogenicity, and efficacy of the BNT162b2 covid-19 vaccine in adolescents.N Engl J Med. 2021; 385: 239-250https://doi.org/10.1056/NEJMoa2107456Crossref PubMed Scopus (584) Google Scholar An RCT published in 2011 assessing the clinical effect of ivacaftor in the treatment of children with cystic fibrosis fundamentally altered the treatment of children with this disease.2Ramsey B.W. Davies J. McElvaney G.N. Tulllis E. Bell S.C. Drevinek P. et al.A CFTR potentiator in patients with cystic fibrosis and the G551D mutation.N Engl J Med. 2011; 365: 1663-1672https://doi.org/10.1056/NEJMoa1105185Crossref PubMed Scopus (1737) Google Scholar As part of meta-analyses, RCTs produce high-quality evidence. Virtually every professional society (including the American Academy of Pediatrics) and governmental agency, which create guidelines that also influence clinical care, state that RCTs are at the top of the evidence pyramid, with most also acknowledging the importance of meta-analyses of RCTs. For example, the American Academy of Pediatrics included the use of semaglutides in the treatment of obesity in adolescents in their recent guideline largely based on a single RCT.3Hampl S.E. Hassink S.G. Skinner A.C. Armstrong S.C. Barlow S.E. Bolling C.F. et al.Clinical practice guideline for the evaluation and treatment of children and adolescents with obesity.Pediatrics. 2023; 151e2022060640https://doi.org/10.1542/peds.2022-060640Crossref Scopus (150) Google Scholar The US Preventive Services Task Force, which relies principally on RCTs to make recommendations, is another important example; A and B Recommendations must be supported by insurance companies. A recent report sheds new light on issues relating to RCTs and child health. Rees et al detailed the failure of a substantial number of National Institutes of Health (NIH)-sponsored RCTs involving children to meet their study objectives and of investigators to meet their ethical obligations. In a cross-sectional analysis, they reported that, of 5062 NIH pediatric grants with completion dates between 2017 and 2019, 413 (8.1%) were RCTs.4Rees C.A. Narang C. Westbrook A. Bourgeois F.T. Dissemination of the results of pediatric clinical trials funded by the US national institutes of health.JAMA. 2023; 329: 590-592https://doi.org/10.1001/jama.2022.24025Crossref PubMed Scopus (4) Google Scholar Only 64% were prospectively registered, 13% had submitted results to ClinicalTrials.gov within 12 months of completing the trial, 54% had published results by 48 months after grant completion, and 41% results were neither published nor posted on ClinicalTrials.gov. Similar issues have been noted in adult trials. Although this study reviewed a limited number of journals over a short period, those are the most recent relevant and important data available. Because patients enroll in clinical trials not knowing whether they personally will benefit, some of these investigators failed in their ethical obligation to report the results of these trials, and to ensure that the data were in the public domain. The findings by Rees et al are also reflected in the far smaller increase in the number of published pediatric RCTs compared with adult RCTs.4Rees C.A. Narang C. Westbrook A. Bourgeois F.T. Dissemination of the results of pediatric clinical trials funded by the US national institutes of health.JAMA. 2023; 329: 590-592https://doi.org/10.1001/jama.2022.24025Crossref PubMed Scopus (4) Google Scholar A study of RCTs published in 174 journals over the 30 years between 1985 and 2005 found that the number of RCTs that included adults increased by 90.5 per year, whereas the increase in pediatric RCTs was 16.9 per year.5Cohen E. Goldman R.D. Ragone A. Uleryk E. Atenafu E.G. Siddiqui U. et al.Child vs adult randomized controlled trials in specialist journals: a citation analysis of trends, 1985-2005.Arch Pediatr Adolesc Med. 2010; 164: 283-288https://doi.org/10.1001/ARCHPEDIATRICS.2009.291Crossref PubMed Scopus (0) Google Scholar A more recent analysis for the period 2005-2018 of 7 general medical journals with the highest 2017 impact factors found that publication of adult RCTs increased by 5.1 per year, although there was no increase in pediatric RCT publications.6Groff M.L. Offringa M. Emdin A. Mahood Q. Parkin P.C. Cohen E. Publication trends of pediatric and adult randomized controlled trials in general medical journals, 2005-2018: a citation analysis.Children. 2020; 7: 293https://doi.org/10.3390/children7120293Crossref PubMed Scopus (16) Google Scholar Although publication bias could not be assessed in either study, the smaller increase in pediatric RCTs is a concern. There are numerous well-known challenges to conducting RCTs in children, including ethical issues, recruitment of children in busy families, access to enough eligible children with a particular disease, and funding by government or industry. One additional issue is the availability of pediatric physician scientists with expertise in training and interest in conducting these studies. With the focus of academic medical centers on generating clinical revenue, the pediatric physician scientist remains an endangered species, especially for those early in their careers. The mindset that an RCT might not be possible must also be overcome; the development of step-wedge and adaptive designs have increased the feasibility of conducting RCTs for many important clinical questions. Yet, there are some pediatric groups that have been hugely successful. Pediatric oncology is particularly instructive. The first multisite cooperative group in childhood cancer was established in 1955, before any other such group existed.7Chidren's Oncology Group - Our History. Accessed April 5, 2023. https://childrensoncologygroup.org/historyGoogle Scholar Indeed, pediatric oncologists were pioneers in establishing large consortium research networks. In 1990, the then 4 cooperative groups emerged into one, recognizing that this step was necessary to enroll sufficient patients in RCTs to improve child health outcomes in pediatric cancer. Other successful multisite cooperative groups now exist, including the National Institute of Child Health and Human Development Neonatal Research Network and the Pediatric Emergency Care Applied Research Network. Major funders of clinical research, including the NIH, industry, and patient advocacy groups, should consider supporting additional cooperative groups conducting child health research. It is important that investigators consider whether any single site clinical trial, or even those with a limited number of sites, will enroll enough patients to meet study objectives. The data and our experience suggest that investigators are often overly optimistic regarding the potential number of patients that can be enrolled in RCTs, and that this challenge persists in conducting RCTs. Given the size of the child population with a particular disorder compared with so many of the diseases in the adult population, research networks are the key to successful pediatric RCTs. There are other study types that change practice. For example, a recent large, well-designed and well-analyzed cohort study comparing the use of methadone vs buprenorphine during pregnancy concluded that buprenorphine was associated with a lower risk of neonatal morbidity, but had no associated effect on maternal outcome.8Suarez E.A. Huybrechts K.F. Straub L. Hernandez-Diaz S. Jones H.E. Connery H.S. et al.Buprenorphine versus methadone for opioid use disorder in pregnancy.N Engl J Med. 2022; 387: 2033-2044https://doi.org/10.1056/NEJMoa2208601Crossref PubMed Scopus (88) Google Scholar The keys in this study include the size, >2 500 000 pregnancies; a comparison of 2 treatments, methadone vs buprenorphine, mimicking an RCT; and a sophisticated analysis, high-density propensity-score algorithm, which allowed >200 covariates to be used in the analysis. Cohort studies can and do influence clinical practice, but they must be well-designed, of sufficient size to allow for sophisticated analysis, and, most important, the hypotheses to be tested should be stated before any analysis. The widespread adoption of the electronic health record is likely to increase the number of well-conducted cohort studies focused on child health. Just as research networks can support RCTs, they can also be used to conduct high-quality cohort studies. Cohort studies are important in child health, because some of the critical issues in child health, such as pervasive adverse influences of social determinants of health (SDH), will be difficult and unethical to study in RCTs. Unfortunately, unlike RCTs, there is no mandatory registration for cohort studies. Child health investigators should always consider the RCT as the optimal methodologic approach, ensure that the study is well-designed and appropriately powered with the likelihood they will recruit the planned sample size, and report the results of the trial. For example, there is substantial interest in screening for SDH, yet a recent review suggests that it is unclear if this improves patient outcomes.9Eder M. Henninger M. Durbin S. Iacocca M.O. Martin A. Gottlieb L.M. et al.Screening and interventions for social risk factors: technical brief to support the US Preventive Services Task Force.JAMA. 2021; 326: 1416-1428https://doi.org/10.1001/jama.2021.12825Crossref PubMed Scopus (61) Google Scholar Future studies evaluating screening for SDH should be RCTs. Finally, and most important, it is the ethical obligation of every investigator of an RCT to ensure that the results of these studies are reported in ClinicalTrials.gov and that they seek publication in a timely fashion. As we continue with efforts to further improve child health and healthcare, rigorous research—led by RCTs—must be a key element of these endeavors. The authors declare no conflicts of interest.