Socioeconomic variations in access to smoking cessation interventions in UK primary care: insights using the Mosaic classification in a large dataset of primary care records

BACKGROUND Smoking prevalence is particularly high amongst more deprived social groups. This cross-sectional study uses the Mosaic classification to explore socioeconomic variations in the delivery and/or uptake of cessation interventions in UK primary care. METHODS Data from 460,938 smokers registered in The Health Improvement Network between 2008 and 2010 were analysed. Logistic regression was used to calculate odds ratios for smokers having a record of receiving cessation advice or a prescription for a cessation medication during the study period by Townsend quintile and for each of the 11 Mosaic groups and 61 Mosaic types. Both of these measures are area-level indicators of deprivation. Profiles of Mosaic categories were used to suggest ways to target specific groups to increase the provision of cessation support. RESULTS Odds ratios for smokers having a record of advice or a prescription increased with increasing Townsend deprivation quintile. Similarly, smokers in more deprived Mosaic groups and types were more likely to have a documented cessation intervention. The odds of smokers receiving cessation advice if they have uncertain employment and live in social housing in deprived areas were 35% higher than the odds for successful professionals living in desirable areas (odds ratio (OR) 1.35, 95% confidence interval (CI) 1.20-1.52; absolute risks 57.2% and 50.1% respectively), and those in low-income families living in estate-based social housing were 50% more likely to receive a prescription than these successful professionals (OR 1.50, 95% CI 1.31-1.73; absolute risks 19.5% and 13% respectively). Smokers who did not receive interventions were generally well educated, financially successful, married with no children, read broadsheet newspapers and had broadband internet access. CONCLUSIONS Wide socioeconomic variations exist in the delivery and/or uptake of smoking cessation interventions in UK primary care, though encouragingly the direction of this variation may help to reduce smoking prevalence-related socioeconomic inequalities in health. Groups with particularly low intervention rates may be best targeted through broadsheet media, the internet and perhaps workplace-based interventions in order to increase the delivery and uptake of effective quit support.