A qualitative evidence synthesis of employees’ views of workplace smoking reduction or cessation interventions

The need to reduce smoking has become a major policy and public health issue in recent years, due to the well-documented burden of smoking-related diseases on individual health and society as a whole. The workplace is seen as an appropriate context for assisting smoking cessation because it enables access to a large and fairly stable population, facilitating higher participation rates for interventions, and with the potential to use peer group support and positive peer pressure [1]. The United Kingdom has introduced national workplace, smoke-free policies in the last 10 years and the National Institute of Health and Care Excellence has produced guidelines on how to implement these policies and how best to assist employees with smoking cessation [2]. Previous systematic reviews have found that smoke-free workplaces encourage quitting and reduction in smoking rates [3] and that proven stop-smoking methods (i.e. from outside the workplace), including group therapy, individual counselling and nicotine replacement therapy, are equally effective when offered in the workplace [1]. Such interventions have also been found to be cost-effective [4]. The evidence has been found to be more equivocal for other approaches, such as self-help methods, social and environmental support, and incentives and competitions [1].

There is currently no synthesis of qualitative evidence related to workplace smoking cessation programmes, but the value of such a synthesis is twofold. First, a consideration of the qualitative evidence of employees’ views regarding workplace interventions or programmes might help to explain outcomes, that is, why some interventions might work and others not. This is one function of qualitative evidence synthesis [5, 6]. Second, many of the non-pharmaceutical interventions for smoking cessation or reduction are so-called “complex interventions”, that is, “interventions with several interacting components” [7]. It is recommended that complex interventions be developed based on theory [7], but qualitative evidence synthesis offers an alternative approach from theory for identifying appropriate and acceptable elements of such interventions [8].

There are many recognised theories that seek to explain health behaviours, such as the Transtheoretical model of behaviour change [9], so it makes sense to utilise such models and theories in evidence syntheses that address questions relating to health behaviour or decision-making [10]. Synthesis methods that adopt this approach include realist and framework synthesis [11‐13] and “best fit” framework synthesis [8, 14]. The “best-fit” method was chosen as it has been found by the authors to be suitable for questions relating to individuals’ decision-making regarding health behaviours [8, 14]. It explores relevant theory within a specific context, and can generate a refined, context-specific, conceptual model that can be used to understand the reported effectiveness or otherwise of interventions, as well as being used itself to develop interventions. The method is described in full elsewhere [8, 14]. In essence, it involves the identification of relevant theories or conceptual models to create an a priori framework. Data from primary research studies identified for the synthesis are coded against this a priori framework; data not captured by this framework are then analysed using secondary thematic analysis to generate new themes. A new framework and conceptual model is thus created using all themes, pre-existing and newly specified, found by reviewers within the data.

The search generated 747 unique citations. Sixty-five full papers were retrieved as potentially relevant, of which 14 studies were found to satisfy the inclusion criteria. One additional relevant study was identified from the references of an included study [26]. See Figure 1.

For the characteristics of the included studies, see Table 2. Six studies examined people’s views about employers’ decisions to restrict smoking within or at a workplace. Five studies explored views relating to multi-faceted interventions, i.e. involving a combination of at least two or more of the following: self-help or educational materials, smoking cessation resources or “props” such as nicotine patches, nicotine replacement therapy (NRT) or pencil cigarettes, support groups, peer support, telephone counselling, and competitions or incentives. One employed telephone counselling only and one incentives only. Two studies did not specify the intervention, but rather elicited people’s views on the principle of a workplace smoking cessation intervention. Where reported, there was considerable variety in terms of organisation sector and size. Eight studies employed satisfaction surveys with closed questions, and seven studies employed one-to-one interviews, focus groups, discussion group posts or open-ended survey questions.

Evidence was found to support all of the key concepts in the a priori framework, i.e. beliefs about smoking, the pros and cons and perceived norms regarding smoking in the workplace, and the factors moderating the relationship between any intervention and successful quitting: Dependence, priority of quitting and self-efficacy (perceived ability to quit). The new concepts produced by the synthesis, perhaps unsurprisingly, related to the contextual specifics of the setting (the workplace) and the interventions (components, delivery etc.), which were not well represented in the behavioural theories that formed the basis of the a priori framework. The synthesis does not appear to be sensitive to the data collection methods employed by the included studies because all of the concepts in the model were informed by evidence from all included study designs. In the same way, the overall synthesis was not greatly affected by the type of intervention being examined in the included studies, though the concepts of employer support and employee attitudes to smoking had some distinct elements derived exclusively from studies exploring experiences of workplace bans and restrictions.

There is nothing to suggest that the workplace itself offers a particularly special setting for smoking cessation or reduction interventions, as the intervention preferences described and the key concepts of priority and perceived ability to quit can all exist outside the workplace. After all, unassisted smoking cessation is the most frequently successful approach [43]. However, the workplace does provide a context for interventions and this evidence synthesis suggests that employees’ expectations regarding employers’ support for, and enforcement of interventions or restrictions might facilitate smoking cessation. If an intervention is to take place in the workplace, then any benefits of that environment and context must be exploited.

The findings of this evidence synthesis suggest that the following elements of the a priori framework’s foundation theories represent the single most important concepts in smoking cessation theory within the workplace context: Motivation to change [17] and preparation for change [19], from versions of the stages of change model [9]; and Conrad’s cue-to-action addition to the Health Belief Model [21]. These ideas are captured by the broad concept of readiness for change in this paper’s conceptual model and, in particular, in the concepts of priority of quitting and perceived ability to quit. The concept of readiness for change is well-known in theories relating to smoking cessation, but this is the first evidence synthesis to validate that finding with specific reference to the evidence of employees’ views relating to workplace smoking programmes.

The findings of this synthesis are contextual and should be viewed alongside the effectiveness evidence. Indeed, they might help to explain differences in the apparent relative efficacy of interventions. Self-help methods, social and environmental support, and incentives and competitions, were all dismissed as ineffective by a Cochrane review [1], although some of the included trials reported that the priority of quitting variable might either predict cessation rates [44] or predict participation, and participation has been found to correlate with cessation [45, 46]. However, this evidence synthesis suggests that such intervention components are liked by employees, particularly those for whom quitting is a priority and a perceived possibility. Indeed, competitions and incentives are thought to offer a valuable, additional motivation. The trials included in the Cochrane review that evaluated these interventions were aimed at all employees who smoke, regardless of readiness for change (for example, [47‐49]), yet, as noted by this qualitative evidence synthesis, such incentives would be unlikely to have any effect unless the smoker was ready to quit. Interventions that have appeared most effective in terms of overall rates of cessation or reduction, such as group therapy and individual counselling [1], address the issue of health behaviour change more broadly, rather than through a single component, such as an incentive. They focus on altering employed smokers’ beliefs about smoking, to affect a change in opinions about the importance of, or need and ability to stop smoking, as well as seeking to motivate and sustain change. This evidence synthesis suggests therefore that there might be value in testing a multi-faceted intervention that includes components to appeal to all employed smokers, whatever stage they occupy in the process of change. An intervention should involve visible employer support, incentives, competitions, and self-help methods, as well as NRT, and individual and group counselling. Such an intervention would obviously be resource intensive, and its cost-effectiveness would need to be assessed, but it has the benefit of potentially addressing the needs and preferences of all employees who smoke.

Employees’ views about smoking in the workplace, and their employers’ roles and responsibilities, are mostly shaped by their beliefs about smoking in general. Workplace interventions should have multiple components to satisfy the requirements of all employees who smoke, both those who feel ready and able to quit, and those for whom quitting is not yet a priority or a possibility. This is because the needs and preferences of these two groups can be quite different.

CC conceived and designed the case study; CC, JR and JL extracted the data, appraised included studies and analysed and interpreted the data. CC drafted the paper, JR, JL, DF and AB undertook critical revision of important content of the manuscript. All authors approved the final version of the manuscript.