Trial Information

En Pareja: A Latino Couples Intervention to Help Expectant Fathers Quit Smoking

BACKGROUND & SIGNIFICANCE Latino men are just as likely to smoke as non-Hispanic White men;
[CDC, 2004] however, few smoking cessation programs have shown long term effects among
Latinos. Further, many Latinos cannot enter traditional smoking cessation programs because
they do not speak English. To be effective, programs should capitalize on a "teachable
moment". [McBride, 2003] Pregnancy has been shown to be a teachable moment for women to
quit smoking; whether it is a teachable moment for partners has not been explored.
Including Latinas and intervening with the couple might help sustain cessation effects long
term. The aim of this randomized controlled trial is to compare a culturally tailored
intervention to a minimal intervention (self-help smoking cessation guides or Guia) in
promoting and sustaining smoking cessation among expectant Latino fathers during pregnancy
and postpartum.

DESIGN & PROCEDURES

When describing this study to the pregnant women, we will explain that half of the couples
will be randomly assigned to receive face-to-face counseling plus a self-help cessation
guide (Guia) while the other half will receive the self-help smoking cessation guide only.
When a woman agrees to have us meet with her and her husband/partner, we will send her home
with two consent forms, one for her and one for her husband/partner, along with a small gift
for the upcoming baby. We also will set up a time and place to meet the couple for their
initial data assessment. We will not collect information about legal status of patients.

Couples who are randomized to the face-to-face support intervention will be asked to attend
two two-hour in-home sessions, one in pregnancy and one after the baby is born. Each man and
each woman will receive four booster calls, two in pregnancy and two post-partum. These
calls last approximately ten minutes. At the individual level, the intervention focuses on
skills training for behavior change (smoking cessation for men and diet or exercise for
women) and goal setting. At the couple level, the intervention focuses on skills training
for communication (speaking and listening skills and problem solving).

Sample size considerations A total of 366 couples will be randomized to the two arms within
strata defined by whether this is their first pregnancy. The primary objectives are to test
whether there are arm differences in the proportion of fathers who are abstinent from
smoking at 28 weeks of pregnancy (time 1) and at 12 months post-randomization (time 2). We
expect attrition rates among fathers of 20% and 30% at time 1 and time 2, respectively;
dropouts will be imputed to be non-abstinent in an intention-to-treat analysis. The
chi-square test for a difference in two proportions will be used to test the null hypotheses
of no difference in abstinence rates between the two arms. The overall one-sided alpha level
will be controlled at 0.025 by conducting each test at a one-sided alpha of 0.0125. Based on
Nevid and colleagues' previous study that compared counseling to pamphlets,[Nevid, 1997] we
anticipate abstinence rates of 10% and 5% in the Guia only arm at time 1 and time 2,
respectively. The alternative hypotheses are that the arm difference at time 1 will be 15%
(25% versus 10%) and at time 2 will be 10% (15% versus 5%). A sample of size 366 was
selected to have at least 90% power for each alternative hypothesis. For the effect at time
2 to have 90% power, the effect at time 1 is overpowered at 97%. It is assumed that the arm
effects of 25% versus 10% at time 1 and of 15% versus 5% at time 2 consider those
participants who dropout and have their outcomes imputed to be non-abstinent. Thus, under
the alternative hypotheses, the probabilities of response among just those 80% and 70% who
do not drop out are (31% versus 13%) and (21% versus 7%), respectively.