You are here:
RECRUITING FOR A LONGITUDINAL STUDY OF CHILDREN'S HEALTH USING A HOUSEHOLD-BASED PROBABILITY SAMPLING APPROACH
Citation:
MENDOLA, P., K. SCHOENDORF, K. RAPAZZO, S. KEIM, W. GALKE, R. BRENNER, D. B. WALSH, A. H. WILLIAMS, AND L. MESSER. RECRUITING FOR A LONGITUDINAL STUDY OF CHILDREN'S HEALTH USING A HOUSEHOLD-BASED PROBABILITY SAMPLING APPROACH. Presented at International Society for Environmental Epidemiology Annual Meeting, Paris, FRANCE, September 02 - 06, 2006.
Description:
The sampling design for the National Children¿s Study (NCS) calls for a population-based, multi-stage, clustered household sampling approach (visit our website for more information on the NCS : www.nationalchildrensstudy.gov). The full sample is designed to be representative of both urban and rural births in the United States, 2007-2011. While other survey research efforts have used similar designs with success, this would be a novel recruitment strategy for an environmental epidemiologic study of preconception factors and pregnancy outcomes.
A comprehensive pilot study is planned for 2006-2009 in central North Carolina - BEFirstNC. We will screen a total of approximately 10,000 households from four defined geographic areas (2 urban and 2 rural). The areas were chosen to have a broad range of demographic variation with respect to race/ethnicity and income as well as urban/rural housing units. The study will enroll women planning pregnancy and those currently pregnant, with a small number of women enrolled in the hospital after delivery.
BEFirstNC study visits mimic the visit schedule planned for the first two years of the NCS. Frequent home and clinic visits will continue until the child is 18 months of age. Data collection from participants (women, partners and children) will focus on topics that are likely to yield meaningful results in a small sample of infants (~400 children). Participant burden associated with study visits, in-person interviews, mail-in forms, biologic specimen collection, environmental sample collection and clinical tests will be characterized. Frequent qualitative assessments will evaluate the acceptability of the visit schedule and planned data collection methods including: likes and dislikes of study participants regarding each visit, important factors in the participant¿s decision making process to join the study, continue to participate or drop out. Spatial environmental monitoring data and community resource information (parks, etc.) will be combined with individual family data to assess the neighborhood as well as household environment.
This pilot study can provide valuable information on the feasibility of using a household-based sampling approach to identify, enroll and retain women and families for a longitudinal study of child health and development. Demographic differences in participant response and retention will be informative for the national study as well as the qualitative assessment of study acceptability.
Using a population-based sampling frame for a study of pregnancy outcomes can avoid biases associated with recruiting volunteers who may be more health conscious than non-participants and/or may be more concerned about their environment. Measuring aspects of the environment before a couple becomes pregnant allows for an interesting opportunity to assess peri-conceptional factors that may relate to pregnancy outcomes as well as child development.