Claire Tatton, G J Melendez-Torres
Our analysis provides a framework to understand what options could be used to tailor interventions to underserved groups. These include offering adapted communication, habit building approaches and social network support, improving relationships with providers, and reducing practical barriers to engagement. Further testing is warranted to more accurately assess how approaches may relate to effectiveness, and for who.
OBJECTIVE: Despite reductions in rates of smoking in pregnancy in most high-income countries, disparities persist between those from different socioeconomic backgrounds and ethnic groups. Our recent systematic review that sought to understand the equity impacts of psychosocial interventions indicated effectiveness of tailored interventions for prenatal smoking cessation and postnatal abstinence, but with significant unexplained heterogeneity. We conducted this review to explore how interventions were tailored and why some may work better than others.
METHODS: We used intervention component analysis to characterise the features of tailored interventions. We conducted meta-analyses for prenatal smoking cessation (second or third trimester) and postnatal abstinence (<12 months from childbirth) and ran meta-regressions to explore how components may relate to effectiveness.
RESULTS: We included 70 papers related to 54 randomised controlled trials. We identified 39 components common to both tailored and universal interventions and 26 components unique to tailored interventions. Equity focussed components sought to improve the relevance of the intervention, and to improve accessibility through intervention position. Meta-analyses indicated effectiveness for both prenatal smoking cessation and postnatal abstinence (odds ratio [OR] 1.47, 95% CI 1.32 - 1.63 and OR 1.30, 95% CI 1.13 - 1.49). However, meta-regressions showed no conclusive evidence that any individual tailored intervention component produced differential effectiveness. This may be due to insufficient statistical power resulting from the limited number of trials containing some of the components reported.
CONCLUSION: Our analysis provides a framework to understand what options could be used to tailor interventions to underserved groups. These include offering adapted communication, habit building approaches and social network support, improving relationships with providers, and reducing practical barriers to engagement. Further testing is warranted to more accurately assess how approaches may relate to effectiveness, and for who.