Will Smokers be Left out in the Cold By an ‘Un'-Affordable Care Act?

Background: The 2010 Affordable Care Act (ACA) and accompanying federal regulations require health insurance plans cover two smoking cessation attempts per year without copayments. The ACA also allows a tobacco user surcharge of up to 50% of the base monthly insurance premiums that cannot be reduced by government subsidies, potentially making insurance unaffordable for smokers, particularly for those who live in the 21 states that have not expanded Medicaid. In addition, because prevalence of smoking and lack of health insurance is higher among similar population subgroups, the impact of this policy on smokers could be severe. To assess the potential impact, we tested whether: 1) individuals lacking health insurance smoke at higher rates compared to insured individuals; 2) smokers in general, and subgroups with higher prevalence of smoking would be more likely to lack insurance; and 3) uninsured individuals would be more likely to smoke when controlling for known high prevalence groups.

Method: Using data from the 2013 National Health Interview Survey collected before the first insurance exchange plans were implemented, estimates of smoking prevalence were obtained among high rate subgroups by insurance status, adjusting for non-response bias and weighted to be nationally representative. Differences among subgroups were evaluated using the Wald test for weighted data. Model 1: The first multivariable logistic regression model evaluated odds of being uninsured by smokers and subgroups. Model 2: The second model evaluated the odds of current smoking for uninsured populations, controlling for age, race, education, poverty, U.S. census region, limited abilities, and sexual orientation.

Results: Significantly higher rates of smoking were found among uninsured compared to insured populations (29.3% vs. 15.5%, p<.0001). Smoking rates were also significantly higher among uninsured compared to insured subgroups (each p≤.0001, highest rates reported) by age (45-64: 32% vs. 18%), race (White: 42% vs. 16%), education (high school or equivalent: 37% vs. 21%), poverty status (below poverty: 33% vs. 27%), U.S. census region (Midwest: 35% vs. 18%), physical or mental limitation (any limitation: 41% vs. 23%), and sexual orientation (LGB: 41% vs. 24%). Model 1: Current smokers had significantly increased odds of being uninsured (OR: 1.90, 95% CI: 1.72-2.10). With the exception of sexual orientation and living in the Midwest census region (reference Northeast), all other subgroups had significantly higher odds of lacking health insurance. Model 2: After controlling for subgroup status, uninsured individuals had significantly increased odds of being a current smoker (OR: 1.89, 95% CI: 1.71-2.08).

Conclusion: Uninsured individuals are more likely to be smokers, and the potential for unaffordable health care insurance premiums for high smoking prevalence subgroups is an important policy concern. The unequal implementation of the law in expanding Medicaid services in some states further reduces the impact that the ACA has as tobacco control policy. Ensuring that individuals have access to affordable health insurance and free preventive care that includes tobacco cessation and prevention is an important step for tobacco control. States can mitigate these concerns by expanding Medicaid and reducing the tobacco surcharge, particularly for low-income smokers.