Hospital Based Sampling: Lessons Learned from the SAFE Study

• Activity Number: 406
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Brenda Cox and Nicole Santomauro*+ and Jane Sheehan and Dena Margolis
• Companies: Social & Scientific Systems and Boston University and Boston University and Boston University
• Keywords: hospital-based sampling ; infant care practices ; oversampling ; composite size measure ; replacements ; two-stage sampling

Abstract

The Study of Attitudes and Factors Affecting Infant Care Practices (SAFE Study) collected data on sleep position, bed sharing, pacifier use, and other infant care practices from a national probability sample of about 3,000 mothers of 2 to 10 month old infants. SAFE employed a two-stage sample with 32 hospitals sampled using a composite size measure to oversample minorities. Hospitals unable to participate were replaced with a randomly selected hospital from that stratum. In each of 3 cycles, hospitals were assigned targets for sampling and enrollment of Hispanic, Black, and all other race mothers. Enrolled mothers completed a short baseline interview, and then were asked to complete a follow-up survey by web (66%) or telephone (34%) to achieve each cycle's goal of at least 1,000 completed surveys, 250 each from Hispanic and Black mothers. In this paper, we present solutions to the operational challenges this design posed for frame building and stratification; sample allocation and minority oversampling; hospital recruitment and maintenance over the three year study; mother selection and enrollment; and maximizing response rates for follow-up surveys.