Smoking cessation in severe mental illness: what works?

Abstract

The physical health of people with severe mental illness (SMI) is poor. Smoking-related illnesses are a major contributor to excess mortality and morbidity. An up-to-date review of the evidence for smoking cessation interventions in SMI is needed to inform clinical guidelines. We searched bibliographic databases for relevant studies and independently extracted data. Included studies were randomized controlled trials (RCTs) of smoking cessation or reduction conducted in adult smokers with SMI. Interventions were compared to usual care or placebo. The primary outcome was smoking cessation and secondary outcomes were smoking reduction, change in weight, change in psychiatric symptoms and adverse events. We included eight RCTs of pharmacological and/or psychological interventions. Most cessation interventions showed moderate positive results, some reaching statistical significance. One study compared behavioural support and nicotine replacement therapy (NRT) to usual care and showed a risk ratio (RR) of 2.74 (95% CI 1.10-6.81) for short-term smoking cessation, which was not significant at longer follow-up. We pooled five trials that effectively compared bupropion to placebo giving an RR of 2.77 (95% CI 1.48-5.16), which was comparable to Hughes et al.'s 2009 figures for general population data; RR = 1.69 (95% CI 1.53-1.85). Smoking reduction data were too heterogeneous for meta-analysis, but results were generally positive. Trials suggest few adverse events. All trials recorded psychiatric symptoms and the most significant changes favoured the intervention groups over the control groups. Treating tobacco dependence is effective in patients with SMI. Treatments that work in the general population work for those with severe mental illness and appear approximately equally effective. Treating tobacco dependence in patients with stable psychiatric conditions does not worsen mental state.