Effective approaches to persuading pregnant women to quit smoking: a meta-analysis of intervention evaluation studies.

OBJECTIVES The study aimed to: (1) assess the effectiveness of prenatal smoking cessation interventions, (2) clarify whether the psychological changes targeted by interventions are related to their effectiveness, (3) identify specific intervention components associated with greater effectiveness, and (4) establish whether aspects of evaluation methodology are associated with a greater effectiveness. DESIGN Differences in proportions of women quitting and odds ratios were calculated for the intervention and control groups. Interventions were categorized in relation to the main intervention target (i.e. cognitive preparation versus increased threat perception), use of follow-up contact, use of individual cessation counselling and other characteristics. Methodological approaches to evaluation were also categorized. METHODS A systematic literature review generated 36 controlled evaluations, including one unpublished study. A meta-analysis was used to relate study classifications to effectiveness. This involved univariate analyses and a multivariate model of the relationship between observed univariate effects. RESULTS A weighted mean odds ratio of 1.93 indicated a good overall effectiveness. Cognitive preparation interventions achieved higher quit rates (6.5%) compared to interventions focusing on threat perception (2.2%). However, this effect was not maintained in the multivariate analysis. CONCLUSION Interventions should employ follow-up, but further research is required to assess the impact of one-to-one counselling. Clarification of the psychological change processes underlying the observed effectiveness of these interventions is required. Future research should seek to identify the active ingredients and cognitive mediators of successful interventions.