What Works

Financial incentives encourage pregnant women to quit smoking

The caveats: They may need to be part of larger programs to be effective. And smoking rates have fallen but remain stubbornly high among people without privilege.

Sarah Kwon

Health care is littered with examples of interventions hyped as breakthroughs that bring about marginal improvements at best. But claims about the benefits of smoking cessation live up to their billing—and the evidence behind them is solid.

Researchers have also compiled a thick dossier on how to encourage smoking cessation. The U.S. Preventive Services Task Force and other groups say convincing evidence supports nicotine replacement therapy (NRT), medications (bupropion and varenicline), and counseling as approaches that improve quit rates in nonpregnant adults who smoke. Public policies like cigarette taxes also work to reduce smoking rates, according to numerous studies.

Overall, cigarette smoking has declined. A CDC report found that the proportion of American adults who reported smoking cigarettes declined from 20.9% in 2005 to 15.5% in 2016. But the battle is far from over. Smoking rates remain high among the poor and people with limited education. And there are populations for whom evidence for NRT or medications is limited but stakes for continuing to smoke are especially high, chief among them, pregnant women. A 2014 Surgeon General’s report found that smoking during pregnancy can lead to low birth weight, premature birth, and sudden infant death. Long term, it can impair a child’s physical growth, intellectual development, and behavioral health. According to government health statistics, in 2016, 7.2% of women smoked cigarettes during pregnancy, with a 12-fold difference between those with higher education and those without (1% for women with a bachelor’s degree or higher versus just over 12% for women with a high school diploma or GED).

Financial incentives and health behaviors—including smoking cessation—have not suffered from lack of research. Even so, the evidence that monetary rewards of various kinds improve quit rates among nonpregnant adults is patchy.

It looks a little more promising for pregnant women. A 2017 Cochrane meta-analysis found that interventions with financial incentives—typically small vouchers ($50 monthly or less) for retail items like groceries—increased smoking cessation rates during pregnancy.

One increasingly popular smoking cessation program for pregnant women is called the Baby & Me-Tobacco Free (BMTF) program. It includes four in-person prenatal counseling sessions, carbon monoxide breath testing at each of those sessions to verify smoking status, and a $25 voucher for diapers given at the third and fourth sessions. To prevent relapse after birth, the program also includes 12 monthly postpartum visits to continue tobacco use testing and provides vouchers for negative test results.

Laurie Adams, BMTF’s founder and executive director, believes that diaper vouchers work as a powerful incentive but isn’t enough by itself. “It’s what brings them in the door. If you have a newborn baby, you want diapers,” she said. “But all three elements need to stay together.”

Adams founded the program in 2001, and the next year, started piloting it in western New York before securing funding from the New York State Department of Health in 2006 to study the program’s impact on quit rates in Upstate New York. A 2011 study published in Maternal and Child Health Journal by Bassett Research Institute, a research group associated with the Bassett Healthcare Network in the region, found that 60% of participants in the intent-to-treat group quit smoking by their fourth prenatal counseling session. The program has been used in 21 states to date. The Tennessee Department of Health adopted it in 2014, then published a study in Maternal and Child Health Journal last year looking at the program’s impact on birth outcomes. The study included nearly 900 pregnant smokers, most of them with Medicaid coverage, and compared them with 11,000 women who were eligible for the program but did not enroll.

Participants who completed at least three counseling sessions and thus received vouchers had a lower rate of low-birth-weight infants than nonparticipants (4.9% vs. 11.6%). After adjusting for potential confounding variables, the study found that participants who completed at least three counseling sessions had 49% lower odds of delivering a low-birth-weight baby than nonparticipants. The study found that the program didn’t have an effect if the women attended fewer than three sessions. Participants completing at least three counseling sessions also had lower rates of preterm birth than nonparticipants, but this result didn’t reach statistical significance.

Odds ratios (ORs) and 95% confidence intervals (CIs) of low birth weight in pregnant smokers according to BMTF prenatal session attendance and smoking cessation
Variables Crude OR (95% CI) Multivariable-adjusted ORa (95% CI)
Program participation
Nonparticipants Referent Referent
Participants with low attendance 0.94 (0.72–1.23) 1.02 (0.77–1.36)
Participants with high attendance 0.39 (0.23–0.66) 0.51 (0.30–0.88)
Smoking cessation
Nonparticipants Referent Referent
Participants without evidence of quitting 0.90 (0.70–1.16) 0.99 (0.76–1.30)
Participants with evidence of quitting 0.26 (0.12–0.56) 0.37 (0.17–0.79)
BMTF=Baby & Me-Tobacco Free, WIC=Women, Infants, and Children supplemental nutrition program.
aMultivariable-adjusted OR: stratified by county of residence and adjusted for age, marital status, race/ethnicity, education, household income, insurance, number of prenatal care visits, WIC participation status, prepregnancy body mass index, history of pregestational diabetes or hypertension, and gestational diabetes or hypertension

Approximately a third of participants completed at least three sessions, of whom 68% quit smoking, confirmed by CO testing at three or four prenatal visits. The study didn’t compare biochemically verified quit rates between participants and nonparticipants.

As with any uncontrolled observational study, differences between the groups may have biased the results. The researchers noted the possibility that nonparticipants may have received smoking cessation interventions, which presumably would have narrowed the difference in results between the participants and nonparticipants.

In a separate analysis, the Tennessee Department of Health projected $2.25 million in savings from the program in 2014. That figure was based on estimates that the program prevented 25 low-birth-weight deliveries and that each of those deliveries would have resulted in $90,000 in hospital costs.

Adams argues that because low-birth-weight deliveries can lead to adverse outcomes beyond the initial hospital stay, savings figures based on hospital costs alone are conservative.