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

Overview

Smoking among Latino males living in the U.S. is a significant public health problem, one that can contribute to disparities in life expectancy and increase mortality. Latinos smoke at the same rate as White males but are less likely than Whites to quit. Interventions do not reach Latino smokers because many speak only Spanish and previous interventions have had notable limitations. First, most have recruited volunteers, so hard-to-reach Latino smokers likely did not participate. Second, cessation effects were short-term only. Third, no program has attempted to boost Latino cessation rates by capitalizing on a "teachable moment", a time when quitting may seem especially relevant. To address these deficits, we propose to conduct a teachable moment intervention trial for Latino smokers. We will attempt to capitalize on the potential teachable moment of Latinas' pregnancy as an impetus for Latinos' cessation. We will include couples, rather than just men, to sustain intervention effects. We will partner with community leaders to develop an intervention based primarily on Social Cognitive Theory, the Teachable Moment Model, and the Cognitive-Behavioral Couple Therapy Model. Some elements will be at the individual level to help Latinos quit smoking and others will be couple-based to improve communication and reduce stress in the postpartum relationship. The program will be culturally sensitive to Latino values, such as familismo, valuing of and duty to the family and personalismo, valuing warm personal relationships. We will recruit Latino couples (n=366) into a Guia (control) arm in which men receive a culturally appropriate smoking cessation guide, and a couple-based smoking cessation counseling arm. Hypothesis 1: Latino expectant fathers who receive couple-based counseling to quit smoking will be more likely to be abstinent from smoking at 28 weeks in pregnancy and 12 months post-randomization than Latino expectant fathers who receive a self-help smoking cessation guide. Hypothesis 2: Couple-based counseling will improve mediators, such as couple communication about smoking, self-efficacy, outcome expectancies, stress levels, risk perceptions, emotion, and self-image, which in turn, will increase cessation rates among Latino expectant fathers. Hypothesis 3: Couples in the counseling arm will have a greater increase in cessation during pregnancy and a lower decrease in cessation at 6 and 12 months post-randomization than couples in the Guia arm.

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.