Pharmacological interventions for smoking cessation among people with schizophrenia spectrum disorders : a systematic review, meta-analysis, and network meta-analysis

Background: People with schizophrenia have higher rates of smoking than the general population, and lower quit rates. Several randomised controlled trials have investigated the effectiveness of pharmacological interventions for smoking cessation over the past 20 years. We did a systematic review and pairwise and network meta-analysis of smoking abstinence to guide decision making in offering pharmacological interventions for smoking cessation for people with schizophrenia spectrum disorders. Methods: We systematically reviewed PubMed, Embase, Cochrane Central Register of Controlled Trials, PsycINFO, and China National Knowledge Infrastructure from inception to Sept 30, 2019, for randomised controlled trials of varenicline, bupropion, and nicotine replacement therapy for smoking cessation for people with schizophrenia spectrum disorders or psychotic disorders who were smokers at the time of study recruitment. Data were extracted from published studies on smoking abstinence outcomes and psychotic symptoms. We did pairwise and network meta-analyses for the primary outcome of smoking abstinence. Sensitivity analyses were done on study inclusion criteria, duration, quality, and location. This study is registered with the international prospective register of systematic reviews PROSPERO, CRD42018102343. Findings: A total of 15 111 records were identified by the database searches, and 163 full-text articles were assessed for eligibility. 145 articles were then excluded for several reasons including insufficient data, or abstracts published in later studies, and 18 studies were included in the meta-analysis. In the pairwise meta-analyses, four studies with 394 participants assessed varenicline (RR 3·75, 95% CI 1·96–7·19, p<0·0001; I2=0%), four studies with bupropion and 292 participants (RR 3·40, 95% CI 1·58–7·34, p=0·0002; I2=0%), and three studies with 561 participants assessed nicotine replacement therapy (RR 4·27, 95% CI 1·71–10·65, p=0·0002; I2=0%). All three treatments were deemed superior to placebo. In the network meta-analysis, varenicline was superior to bupropion (RR 2·02, 95% CI 1·04–3·93; p=0·038) but no significant difference was found between varenicline and nicotine replacement therapy, or bupropion and nicotine replacement therapy. No agents were associated with changes in psychiatric symptoms, but varenicline was associated with higher rates of nausea than was placebo. Interpretation: We found evidence to support use of pharmacological agents for smoking cessation for people with psychosis. Varenicline might be superior to bupropion; however, additional direct testing and combination trials of pharmacological agents for smoking cessation are required to inform clinical decision making for people with psychosis. Funding: None