DESIGN NARRATIVE:

The substudy was a population-based experiment with White, Black, and Hispanic pregnant women whose continued smoking made them ineligible for randomization into the parent study. It was unique in focusing on heavier, more addicted pregnant smokers. PANDA research sites and protocols offered a special opportunity for a low cost test of a disseminable intervention which this project team was uniquely qualified to design and implement. The new intervention, One- to-One, used telephone counselors to assess the counselee's stage in the change process and give stage-appropriate messages, using established techniques of motivational interviewing. Between the two counselor calls spaced 10 days apart, counselees received personalized written feedback and suggestions. The primary aim, increasing quitting during pregnancy, was assessed by unobtrusive urine samples taken during prenatal visits in the ninth month and identified only by study group.

A series of postpartum interviews with subsample cotinine validation was used to examine the second important aim, reduction of infant smoke exposure. A combination of messages, peer modeling, and support helped women sustain cessation after delivery and eliminate smoking around the baby. Project PANDA videotapes and newsletters already contained these messages and required only minimal supplementation to be used with the One-to-One experimental group, regardless of their success in quitting in pregnancy. As in Project PANDA, the assessments were separated from the experiment by enrolling subjects in a university-sponsored study of new mothers' health practices and baby care and by presenting the program as usual care by the health care site.