Effectiveness and cost-effectiveness of computer and other electronic aids for smoking cessation: A systematic review and network meta-analysis

摘要

Background: Smoking is harmful to health. On average, lifelong smokers lose 10 years of life, and about half of all lifelong smokers have their lives shortened by smoking. Stopping smoking reverses or prevents many of these harms. However, cessation services in the NHS achieve variable success rates with smokers who want to quit. Approaches to behaviour change can be supplemented with electronic aids, and this may significantly increase quit rates and prevent a proportion of cases that relapse. Objective: The primary research question we sought to answer was: What is the effectiveness and cost-effectiveness of internet, PC and other electronic aids to help people stop smoking? We addressed the following three questions: (1) What is the effectiveness of internet sites, computer programs, mobile telephone text messages and other electronic aids for smoking cessation and/or reducing relapse? (2) What is the costeffectiveness of incorporating internet sites, computer programs, mobile telephone text messages and other electronic aids into current NHS smoking cessation programmes? and (3) What are the current gaps in research into the effectiveness of internet sites, computer programs, mobile telephone text messages and other electronic aids to help people stop smoking? Data sources: For the effectiveness review, relevant primary studies were sought from The Cochrane Library [Cochrane Central Register of Controlled Trials (CENTRAL)] 2009, Issue 4, and MEDLINE (Ovid), EMBASE (Ovid), PsycINFO (Ovid), Health Management Information Consortium (HMIC) (Ovid) and Cumulative Index to Nursing and Allied Health Literature (CINAHL) (EBSCOhost) from 1980 to December 2009. In addition, NHS Economic Evaluation Database (NHS EED) and Database of Abstracts of Reviews of Effects (DARE) were searched for information on cost-effectiveness and modelling for the same period. Reference lists of included studies and of relevant systematic reviews were examined to identify further potentially relevant studies. Research registries of ongoing studies including National Institute for Health Research (NIHR) Clinical Research Network Portfolio Database, Current Controlled Trials and ClinicalTrials.gov were also searched, and further information was sought from contacts with experts. Review methods: Randomised controlled trials (RCTs) and quasi-RCTs evaluating smoking cessation programmes that utilise computer, internet, mobile telephone or other electronic aids in adult smokers were included in the effectiveness review. Relevant studies of other design were included in the cost-effectiveness review and supplementary review. Pair-wise meta-analyses using both random- and fixed-effects models were carried out. Bayesian mixed-treatment comparisons (MTCs) were also performed. A de novo decision-analytical model was constructed for estimating the cost-effectiveness of interventions. Expected value of perfect information (EVPI) was calculated. Narrative synthesis of key themes and issues that may influence the acceptability and usability of electronic aids was provided in the supplementary review. Results: This effectiveness review included 60 RCTs/quasi-RCTs reported in 77 publications. Pooled estimate for prolonged abstinence [relative risk (RR) = 1.32, 95% confidence interval (CI) 1.21 to 1.45] and point prevalence abstinence (RR = 1.14, 95% CI 1.07 to 1.22) suggested that computer and other electronic aids increase the likelihood of cessation compared with no intervention or generic self-help materials. There was no significant difference in effect sizes between aid to cessation studies (which provide support to smokers who are ready to quit) and cessation induction studies (which attempt to encourage a cessation attempt in smokers who are not yet ready to quit). Results from MTC also showed small but significant intervention effect (time to relapse, mean hazard ratio 0.87, 95% credible interval 0.83 to 0.92). Cost-threshold analyses indicated some form of