Aims. The purpose of this study was to determine predictors of smoking cess
ation from a sample of pregnant Medicaid recipients. Of special interest wa
s whether patient stage of change, based on the transtheoretical model, was
predictive of smoking behavior change during pregnancy. Participants/setti
ng. The sample was drawn from a cohort of pregnant smokers who were partici
pants in a prospective, randomized clinical trial conducted in four public
health maternity clinics in Birmingham, Alabama, USA. Design/measurements.
The 435 participants entered prenatal care on or before their 24th week of
gestation and had saliva collected for cotinine assays at baseline and foll
ow-up. In this secondary analysis, descriptive statistics defined the sampl
e, cross-tabulation procedures identified a preliminary set of predictor va
riables, and discriminant function analyses predicted group membership-quit
ter or smoker. Findings/conclusions. Discriminant function analyses reveale
d that patient baseline cotinine value, duration of smoking habit, self-eff
icacy, exposure to environmental tobacco smoke, and exposure to patient edu
cation methods were predictive of non-smoking status assessed during the th
ird trimester of pregnancy.