Rigorously collected longitudinal developmental data from very preterm infants will propel the fields of child
development and childhood rehabilitation forward in 3 ways: 1) Combining data sets to increase rigor of
predictive models and determine individual and cumulative risk for developmental delay or Cerebral Palsy
(CP), 2) Analysis of clinical trial outcomes using consistent outcome measures could replace the need for
some comparative effectiveness studies, 3) Intervention data from multiple studies will provide data for
preliminary modeling of personalized medicine rehabilitation interventions. However, to achieve any of these
goals, sharing of longitudinal developmental data from cohort studies and clinical trials is crucial.1,2
The purpose of this proposal for Funding Opportunity Announcement (FOA) Number
PAR-22-261 Archiving and Documenting Child Health and Human Development Data Sets is to ensure
that data from this three arm multi-site clinical trial of 85 very preterm infants, funded by NICHD prior to NIH’s
guidelines for data sharing, is made available to research teams who could use it to address research
questions beyond the initial study aims.3 Now is the time to fund the transition of this exciting data from a
private dataset into a shared data set allowing the wealth of secondary analysis or harmonization with data set
from other clinical trials which likely already include sharing plans. For example NICU trials including sensory
intervention4, parenting5, and combined6 trials could be compared with the parent project. In addition, research
teams focusing on understanding development are excited to use this data set and videos to answer questions
on topics ranging from language acquisition to fine motor development.
Infants born preterm are at high risk of developmental delays with up to 20 percent of infants born very
preterm diagnosed with cerebral palsy (CP) at 2 years of age. While the development of the Common Data
Elements for CP (CDE_CP)7 have increased the rigor of individual studies, the use of these highly
recommended outcomes also increases the likelihood of combining data sets. The parent project (R01
HD093624, NCT03518736) uses a combination of outcome measures from the CDE_CP which will enable
harmonization with other data sets. In addition, the use of innovative outcomes that can be used to evaluate
early problem-solving in infants and parent infant relationships starting in the NICU will allow for new and
innovative developmental questions. Thus, researchers from a broad number of fields will benefit from sharing
this data. The aims included in this proposal focus on what can be done with this data set alone.
Aim 1) Provide an accessible dataset with longitudinal data on developmental measures including the
recommended common data elements for CP (CDE_CP) and additional measures that are sensitive to
change in children without CP. Since the 2017 publication of the International Guidelines for the Early and
Accurate Detection of CP8, hospitals with NICUs around the world have been implementing the use of the
General Movement Assessment (GMA) into research and practice. Public access to GMA videos from the
NICU and at 10-14 weeks post term age combined with outcome data could drastically increase the ability to
conduct research on the prognostic value of the GMA and the General Movement Observation Scale (GMOS)
which many be more sensitive to change over time and can be scored from available GMA videos. The parent
project includes GMA data at enrollment between 34-42 weeks of gestation, plus 1-2 videos at the most
predictive age (10-14 weeks post term) and developmental outcomes at 12 and/or 24 months.
Aim 2) Provide a shared data set with longitudinal assessment using the Hammersmith Infant
Neurological Exam (HINE) that can be used to determine if the HINE is sensitive to change over time.
Innovative longitudinal intervention studies, such as the parent project, are beginning to evaluate interventions
initiated in the first weeks of life with the goal of reducing the incidence and severity of developmental delays
and cerebral palsy. The parent project uses the HINE as one component of determining if the child has CP.
However, this prospective cohort is large enough to evaluate sensitivity to change over time using the HINE
which has not been previously evaluated. Thus, is it crucial that data from research with young infants with or
at high risk of having CP are publicly accessible to allow for secondary analysis such as tool validation.
Aim 3) Publish high quality fidelity data which could be compared with other interventions started in
the NICU or at 14 weeks of age. Rehabilitation interventions have a long history of NOT monitoring fidelity.
This data set includes videos of 20% of the intervention visits which have been scored to measure adherence
to the principles of the intervention. Other teams will be able to use the guidance included in the code book
along with videos to improve their own fidelity evaluation or initiate comparative effectiveness study.
Impact: PI Dusing is currently participating in the NICHD funded training program Reproresearch and is a
member of the Databrary Launch Team, magnifying her previous interest in making this incredible data set
available for public use. While epidemiolocal data can provide some insight into the needs or predictors of
infant outcomes, primary longitudinally collected data is likely to be more clinically meaningful and more
accurate. This data will contribute to a greater understanding of development and rehab for children.
