Regulatory Science Virtual Symposium: “Make Informed Decisions: Key Statistical Principles to Clinical Trial Design” Session 4: Pediatric Trials (2022)

1. Pediatric Trials are Hard, so Why Do Them?
   1. Pediatric Research Equity Act (PREA) requires mandatory studies initiated by application for new dosing, route, API, or indication
      1. Waiver/Partial Waiver can be submitted with appropriate justification
   2. Under Best Pharmaceuticals for Children Act (BPCA), FDA can grant a product with additional product exclusivity if voluntary pediatric studies are conducted
2. Who are “Children”?
   1. According to CFR, children are those who cannot consent to treatments/procedures involved in research
   2. Recruitment
      1. Different levels of protections apply to different groups of children: neonates, children, older children
      2. All subject to extra protections
   3. Neonates (day of birth/day 0 to 28^th day post-birth)
      1. Viable
      2. Nonviable – expected to expire
      3. Uncertain Viability
   4. Must the child consent?
      1. Normally, they cannot. Might be able to assent
   5. Can the parents consent?
      1. No, they cannot either.
      2. It is actually called parental permission where parents/guardians allow child or ward to participate in research
      3. Both parents might have permit if research is greater than minimal risk with a prospect of direct benefit
   6. Must the child assent?
      1. Depends on the study
      2. Yes, unless they are too limited to assent i.e., neonate, cognitive ability, and parent can override child if there is a benefit avalible when informed consent for an adult would be waived
      3. Assent is a child’s affirmative agreement to participate in research
3. How do logistics and recruitment bias data?
   1. Do I recruit across IRB-imposed age thresholds?
      1. Consenting vs Assenting vs Permission yields different cohorts
   2. Will I have greater data loss at different time points and is it a confound?
      1. Children can age out of your study
      2. Children can have progressive intellectual disability
   3. How are the dates and times of study activities biasing data?
      1. Parents might be aware during the day versus the end of a workday
      2. Children might behave differently while at school vs. out of school i.e., vacation
   4. Is an adult leaving work a different burden than their child missing school?
      1. Yes i.e., homework, note to school, extracurricular activities
      2. Biases from adults differ from biases from children
   5. Do parents accept different benefits/risks for themselves vs their children?
      1. Parents are less inclined to allow their children to participate in research because stigma and fears of endangering children/unknown
4. Caffeine
   1. Lowers apnea, hypoxemia, mechanical ventilation, and heart defects
   2. Post-natal age (days from birth) as well as Breast fed vs. formula fed affects caffeine metabolism differently
   3. Takeaway: Aspects of development can affect study design!
5. Post Menstrual Age (PMA)
   1. Not everyone has an accurate PMA, which causes more variance in the data
   2. Homoscedasticity: uneven variance among research participants i.e., race, ethnicity, age of the mothers, which can cause bias in the study
   3. Digit preference can cause age data to not be distributed continuously, affecting statistics
6. Demographics of Mothers i.e., race, education, and age affect breastfeeding behaviors
7. Best predictor becomes women’s confidence in continuing to try breastfeeding
   1. This was a hidden (latent) factor!
8. Studying Caffeine for Newborns
   1. Age Range i.e., marketing, PK/PD, variable accuracy, variance, chronological vs. developmental
      1. For device trials, it is important to consider when to remove implants since implants do not grow with children
   2. Moderating Effects i.e., demographics
   3. Mediating Effects i.e., physiology
   4. Latent Factor i.e., maternal confidence
      1. Latent factors can be difficult to address because often, they are unknown
   5. Recruitment and Allocation Strategy (what attributes to control?)
9. Are my outcomes measurable in this child?
   1. Some measurements can only occur when children are young vs. older due to physiological and behavioral changes
   2. Lab work i.e., limited blood volume for younger children
   3. Questionnaires can be separated into the following categories: cognition, behavior, and pain
10. What Else is Changing?
    1. Cardiopulmonary
    2. Skeletal
    3. Hormones
    4. Environmental i.e., changing schools and developmental milestones such as puberty
11. Takeaway: Your data are complex, and your intuition is wrong.

Accompanying text created by Annie Ly | RKS Project Administrator, SC CTSI  lyannie@usc.edu