Incentive interventions for smoking cessation in pregnancy: a mixed methods evidence synthesis

Abstract

Background:
Incentives for smoking cessation in pregnancy are attractive to policy makers because evidence of effectiveness for other interventions is scarce. Our aims were to establish the effectiveness of incentive interventions delivered within or outside the NHS to individuals, families, or organisations that are designed to increase and sustain smoking cessation in pregnancy; investigate how incentive delivery processes work, their acceptability, and how they fit with existing barriers, facilitators, and intrinsic and extrinsic motivators to behaviour change; and work in partnership with mother-and-baby groups to inform the design of incentive trials.

Methods:
This study combines a mixed methods evidence synthesis with primary qualitative and survey research to investigate the perspectives of service users, care providers, the general public, experts, and policy makers. Uniquely, two mother-and-baby groups in areas serving disadvantaged populations in Aberdeen and Blackpool, UK, are study coapplicants, and are providing broad, dynamic, and longitudinal service-user contributions. Researchers attend mother-and-baby groups every 4—6 weeks to feedback findings from evidence synthesis, and record and transcribe discussions of included studies that will iteratively inform future qualitative data collection and analysis, guided by a grounded theory approach. Evidence synthesis follows Cochrane guidance. Detailed searches were done in Medline, Medline-in-Process, Embase, CINAHL, PsycINFO, Web of Science, CENTRAL, Cochrane Database of Systematic Reviews, DARE, HTA, MIDIRS, Applied Social Sciences Index and Abstracts, and the Trials Register of Promoting Health Interventions. 1469 abstracts were identified and 215 full-text reports were screened by two researchers. 21 studies—20 incentivising individuals, one incentivising an organisation—were included in quantitative data analysis and quality assessed with instruments such as Cochrane Risk of Bias and guidance from the Centre for Reviews and Dissemination (dependent on whether the population was randomised). All studies and one survey of attitudes to incentives were included in the delivery processes evidence synthesis, with quality assessment with a Mixed Methods Appraisal Tool. Therefore, one mixed methods coding and data extraction form (MMF) was designed iteratively through piloting with four information-rich studies with form modification after discussion between quantitative and qualitative team members. Quantitative data were extracted from full texts and crosschecked by a second reviewer. Two qualitative researchers independently identified themes that were coded by one with the MMF, with a sample crosschecked by the second. Interpretive themes emerged through discussion and a final thematic framework incorporating continuing service user perspectives is under construction, assisted by NVivo 9 data management software.

Findings:
Incentive interventions identified were multifaceted. 14 (70%) of the 20 patient-level studies verified smoking cessation biochemically rather than relying on self-report. Incentives ranged from four packets of gum to larger incentives—eg, US$50 per month of abstinence. Incentives were often combined with additional smoking cessation components, with varying intensity: 13 (65%) included counselling or behavioural support, 13 (65%) included self-help guides or educational materials, six (30%) included advice to quit, and six (30%) involved social support (eg, including partners or peers) through education materials or encouragement. Few studies were directly comparable, with only four suitable for inclusion in a formal meta-analysis. The relative risk of cessation was 2·77 (95% CI 1·69—4·24), indicating that incentives were effective. Several key themes emerge for delivery processes and acceptability: individual or relational focus; certain or lottery incentives; hedonic or utilitarian incentives; programme bureaucracy meeting chaotic lifestyles; health professional or independent providers; continuity of care; and targeted interventions.
I
nterpretation:
Incentives show promise for smoking cessation in pregnancy and contextual factors probably moderate effectiveness.