A systematic review of smoking, smoking cessation and the homeless: there is a will but is there a way?

Background: Smoking continues to be a lead risk factor of morbidity and early mortality, which is particularly unevenly distributed amongst marginalised groups such as the homeless. This research aims to provide the first systematic review of the evidence relating to smoking prevalence, smoking cessation interventions and barriers and facilitators to smoking cessation amongst homeless adults.

Method: A systematic review (registration number: PROSPERO 2017 CRD42017081843) of peer-reviewed research literature was conducted from inception until February 2018. Data sources included several scientific databases and online libraries (e.g. MEDLINE (PubMed), EMBASE, CINAHL (via Ebsco), AMED (via Ovid/Ebsco), BNI, PsycINFO, Cochrane Library). Search terms: 'smoking' AND 'homeless' AND 'tobacco'. Studies were eligible for inclusion if they were i) written in English; ii) documented smoking rates in adult (18 years and older) homeless populations and/or data on the nature and success of quit attempts and smoking related behaviours (e.g. craving, withdrawal), and/or measured the efficacy and effectiveness of smoking cessation interventions (e.g., NRT), and/or documented barriers and facilitators to quit; iii) participants were users of smokeless tobacco.

Results: 54 studies were included (48 from the US). Due to large heterogeneity in study designs and reporting a narrative and descriptive synthesis was conducted. 14 studies reported smoking prevalence rates which ranged from 57-82% (mean 73%). 20 studies reported some form of smoking intervention and outcomes; only one RCT. Heterogeneity in intervention designs and follow up impeded reporting of aggregated cessation rates. Follow up rates varied from 1-26 weeks, dropout rates ranged from 10-77%. 35 studies identified barriers and facilitators to smoking cessation programmes. Three categories of barriers were identified personal (e.g. co-morbidities, own awareness and knowledge), social (e.g. pressure from other smokers, ability to socialise) and structured and practical barriers (e.g. access, staff not prioritising client’s needs) were identified. Facilitators to smoking cessation included offering financial incentives, social support and trained staff.

Conclusions: Smoking prevalence is disproportionally high amongst homeless adults, even when compared to other marginalised groups. Efforts to reduce smoking rates and subsequent health inequalities in the homeless represent a key group in need of support. Due to the low number of RCTs and wide differences in intervention studies there is no evidence to support one type of intervention targeted at this group. The reviewed evidence suggests smoking cessation interventions are accepted and taken up by homeless adults but cessation is low. Furthermore, the evidence suggests multi-targeted and holistic approaches are needed, placing the person and the situation in the centre of the care, including interventions which offer staff support and training and incentives for follow-up.