Smoking Cessation and Relapse Prevention in Women Postpartum

5.1.1 Sample The 526,761-inhabitant region of West Pomerania comprises 28 pediatric practices. Every mother presenting her newborn for the first time in the pediatric practices will be screened. Of these, all women smoking at the date the female received verification of her pregnancy are eligible for the study. A power analysis on grounds of data from Johnson et al. (2000) and using the software nquery (Elashoff, 2000) revealed that a sample size of 259 women per group will provide a power of 80 % using the five percent one-tailed significance level to compare the experimental and the control group with regard to the prevalence of sustained smoking abstinence 6 months after intervention by a two groups continuity corrected Chi2-test of equal proportions. Assuming an attrition of approximately 25 %, it results in a sample size of 330 women per group. According to the statistical regional authorities of the mentioned region, 3,527 babies were born alive in the year 1998. Therefore 1,234 mothers smoking at the beginning of pregnancy (35 %) per year are expected. Assuming a pediatric practice participation rate of 70 % and an 80 % participation rate of women eligible for the study, about one year of data gathering will be necessary to reach the estimated total sample size of 660 mothers.

5.1.2 Procedure Procedure of the study This study is a randomized controlled trial following a repeated measures factorial design and enabling the comparison of one intervention group with a control group. The work plan of the study incorporates 6 months of preparatory work including the adaptation of instruments for the assessment of core constructs of the Transtheoretical Model and a pre-test, 12 months of screening by receptionists in pediatric practice and the recruitment of eligible women for the study by the pediatrician after giving written informed consent (figure 1). Random assignment to control and intervention group will be done after forwarding the women's addresses to the liaisonal service provided by research staff. The intervention phase of the study will be completed 6 months after the last recruitment. Follow-ups will be carried out 6, 12, 18 and 24 months after baseline assessment. Data analyses and publication will be started in the third year. Analysis of the 18 and 24 month data will exceed the three year period of support.

Figure 1: Time schedule postpartum tobacco smoking relapse prevention intervention

Procedure for participants Figure 2 details the succession of steps for participants. Every woman at her first postpartum visit to a pediatric practice (usually 4 to 6 weeks after giving birth) will be screened by receptionists concerning her smoking status, considering the time at the beginning of the pregnancy, and informed about the study by the pediatrician. All participants will receive treatment as usual at the time of preventive examination of newborns by the pediatrician. After baseline assessment by phone, the intervention group will receive an initial face-to-face counseling at home one week after the pediatric practice visit at home. Two brush-up counseling sessions will be provided by phone 3 and 6 months later. Follow-up assessment will be conducted by phone.

Figure 2: Procedure for participants

5.1.3 Intervention In the intervention group, motivational interviewing will be conducted by psychologists qualified by a comprehensive training curriculum offered at the University of Greifswald according to Miller \& Rollnick (1991). Face-to-face and phone counseling will be tailored by data from baseline assessment and are conceived for a duration of up to 45 minutes. A tailored information and self-help manual will be given at the initial counseling. If the allowance of the participant is received, a tape recording of a random selection of counseling sessions will be done for intervention control. The control group will receive advice as usual from the pediatrician.

5.1.4 Measures To ensure cross-study comparison, all measures will be operationalized by internationally standardized and validated state of the art instruments, if available.

Pre-intervention measures

• pregnancy history
• smoking behavior and history, severity of nicotine dependence
• environmental smoking
• health behavior
• child health
• Adapted Transtheoretical Model core constructs (Stage allocation, processes of change, decisional balance, and temptation/self-efficacy) based on previous work of Ruggiero et al. (2000) and Bane et al. (1999).

Post-intervention measure

• Current and sustained smoking abstinence prevalence, amount of cigarettes smoked per day
• Adapted Transtheoretical Model core constructs

Statistical analysis for the main outcome measure will be done by Chi²-test, T-test and repeated measure MANOVA. Outcome predictors will be analyzed by logistic regression analysis.

5.2 Collaboration 5.2.1 Within EARLINT There will be a close cooperation within EARLINT with all studies comprising motivational enhancement intervention by personal counseling. Quality assurance and training curricula will be matched for all projects.

5.2.2 External This study will be conducted in cooperation with pediatric offices in West Pomerania. Contact has been established with the regional professional organization of pediatricians and Prof. Fusch (Department for Pediatrics at the University of Greifswald). Further collaboration exists with the only research group with comparable work in Germany (Dr. Lang, Prof. Greiser, University of Bremen).

5.3 Ethical and legal correctness All ethical and legal safeguards have been met. The study is currently being refereed by the local ethical committee. The study involves no risk to participants. To the contrary, health benefits are expected. Individuals will be informed of their right to refuse to participate in any or all aspects of the study and will be given opportunities to terminate enrollment. Informed consent forms and an introductory letter will assure individuals that their participation is voluntary and that data protection laws have been met.