Five RCTs (n=695 participants, range 30 to 450) were included: three double-blind placebo RCTs and two non-placebo RCTs. Risk of bias was stated to be low as most trials satisfied most quality criteria; details were not reported.
No effect of nicotine replacement therapy on smoking cessation in later pregnancy was found (pooled RR 1.63, 95% CI 0.85 to 3.14, I2=45%; five RCTs).
Five of the seven safety outcomes (mean birth weight, preterm births, perinatal mortality, post-randomisation foetal deaths, neonatal intensive care admissions) were more positive among infants born to women who had used nicotine replacement therapy, but none of the differences reached statistical significance.
No differences were found between groups in the risk of miscarriage. No differences were found between groups in rates of adherence and three RCTs that reported details of non-serious side effects found no significant differences between groups.
Efficacy estimates were significantly higher in studies that were at higher risk of bias (non-placebo RCTs RR 7.81, 95% CI 1.51 to 40.35; I2=0%) compared with those at lower risk of bias (placebo RCTs RR 1.17, 95% CI 0.83 to 1.65; I2=0%).