National Children’s Study Advisory Committee debates sampling strategies

The committee was given seven possible sampling strategies to consider as alternatives, including two probability designs, two convenience designs and three hybrid designs

For the first time since the President’s Budget came out proposing a $28 million decrease for the National Children’s Study (NCS) in Fiscal Year 2013, the Advisory Committee met to discuss the potential design of the study moving forward. NICHD Director Alan Guttmacher explained to the Advisory Committee members that the Vanguard Centers had been piloting several recruitment strategies after determining the cost and length of time of the initial two-stage geographic-based probability sample and door-to-door household recruitment strategy might compromise the NCS’ feasibility. After piloting a provider-based model and a model relying on direct outreach to the public, the NCS Program Office is considering options using prenatal care providers. In a recent white paper (PDF, 1.5MB), the Advisory Committee was given seven possible sampling strategies to consider as alternatives, including two that were probability designs, two convenience designs and three hybrid designs. Dr. Guttmacher invited the Committee members to weigh in on the advantages and disadvantages of the various sampling strategies.

The Committee discussed the need for simulation models and a possible independent review of the data collected so far, as well as further exploration into what kind of information the NCS would be able to provide scientists or policymakers if it is not a random design. The Committee also debated the need for 100,000 participants in the study, as this number was only required for a probability sample. Bob Kaplan, Director of NIH’s Office of Behavioral and Social Sciences Research, added that it has been 12 years since the study was authorized by Congress, so there is some urgency for the full study to move forward with recruitment. Several principal investigators that are currently involved with the study recommended that the NCS retain a probability sample and continue the NCS in the current 105 sites that were randomly selected and work with providers in those areas to help recruit.  As one Committee Member stated, “the devil is in the details.” NCS Director Steve Hirschfeld indicated that NIH wanted input from the Advisory Committee and would be consulting with federal partners, statisticians and NIH Director Francis Collins before making any final decisions on the design.