Step one: Literature review
Literature was first reviewed to identify theoretical determinants useful in predicting and explaining physical activity. The Theory of Planned Behaviour was chosen as a particularly useful model upon which to base the intervention, having been very widely applied in the domain of exercise and physical activity e.g. [28
]. Briefly, the Theory of Planned Behaviour (TPB) states that the proximal determinant of behaviour is intention, which encapsulates the motivational force that spurs an individual to action [25
]. Intention in turn is held to be determined by attitudes toward the behaviour, subjective norms (perceptions of pressure to engage in the behaviour) and perceived behavioural control (the extent to which an individual believes they are capable of performing the behaviour). To the extent that perceived behavioural control reflects actual control over behaviour, then this factor can also directly predict behaviour. Reviews in this area have indicated that the TPB can typically account for between 42–45% of the variance in physical activity intentions and 27–36% of the variance of physical activity behaviour [28
There is now increasing consensus that each of the constructs within the TPB may be better split into two components [32
]. For example, attitude may be better conceptualised as including both affective (extent to which behaviour is seen as likeable and enjoyable) and instrumental (extent to which behaviour is seen as beneficial and useful) components. Subjective norms may be split into injunctive norms (what important others think) and descriptive norms (what important others do). Finally perceived behavioural control may be split into self-efficacy (an individual's confidence in their ability to perform the behaviour) and perceived control (control within the environment). Thus the current intervention contained strategies to change each of these subcomponent variables of the TPB.
The final determinant to be targeted in the current intervention was knowledge. This was added in recognition of the fact that a baseline level of knowledge about what physical activity is and how one might engage in physical activity was necessary if participants would be required to try to increase their own level of activity. To summarise, the key determinants which the current intervention targeted were: knowledge, affective and instrumental attitudes, subjective norms, self-efficacy and perceived control and intention.
Second the literature was reviewed to determine which types of strategies are most effective at changing physical activity behaviour. This was completed in November 2006. A large systematic review [35
] and a review of reviews [36
] of the effectiveness of interventions to increase physical activity were identified. Both reviews evaluated the effectiveness of interventions found within a range of settings (e.g. community and workplace) that were deemed relevant to the existing study.
In their systematic review, Kahn et al [35
] identified a total of 95 interventions which they split into general categories according to the strategies used in each intervention. Not all of their intervention groupings were relevant to the current project (e.g. school based interventions), however, they found evidence to strongly recommend the use of the following types of interventions which may be of use in workplace settings.
1. Point of decision prompts. Having signs by lifts and escalators to encourage people to use the stairs.
2. Community wide campaigns, encompassing a wide range of intervention studies. Common elements of these types of studies were that they were 'multi-component' (e.g. containing many different activities), and included elements such as support and self-help groups, counselling, screening and education, community events and walking trails.
3. Social support in community settings. Again, a wide range of interventions were contained within this grouping. Common elements were the focus on building, strengthening and maintaining social networks through the use of strategies such as buddy systems, behavioural contracts, walking groups and discussion groups.
4. Individually adapted interventions. The authors found strong evidence to suggest that interventions tailored to an individual's specific interests where personal goals were set also appeared to be effective. Other common elements of these interventions included behavioural self-monitoring, building social support and reinforcement through self-reward.
5. Creation/enhanced access to places for physical activity with information and outreach. These interventions focused on environmental changes such as provision of gym/fitness equipment and walking trails. Other common elements in this theme included provision of screening, support or buddy programmes, seminar and counselling.
In order to ascertain more specific information about which types of intervention strategies showed the most success the 26 studies in the Kahn review that showed significant positive effects on physical activity were collated and examined further. Abraham and Michie [37
] have developed a taxonomy of behaviour change techniques which have been used in interventions to change behaviour. This was used to code the specific strategies that made up the intervention described within each paper. The description of the intervention reported in each paper was read by the first author and coded using the taxonomy. The table of results recording the techniques used in each successful
intervention can be found in additional file 2
. Those techniques that were found most frequently in these successful interventions were 1) planning for social support/social change, 2) prompting intention formation, 3) providing instruction, 4) providing opportunities for social comparison, 5) prompting self-monitoring, and 6) prompting barrier identification. It is important to note that because most interventions in the review were multi-faceted interventions it is not possible to isolate those components within the intervention that have caused the change in behaviour. Thus it is not possible to conclude which of the components identified most frequently have the strongest influence on behaviour. For example, social support/change occurs very frequently but always alongside at least one other intervention component.
Further to this review we also drew on a review of reviews of public health physical activity interventions [36
]. Although this review does include two reviews of workplace physical activity interventions [38
] covering 49 quasi-experimental studies and experimental studies, the findings were inconclusive. The characteristics of studies showing an increase in physical activity at six months were:
• Health screening and counselling
• Follow-up and re-assessment of progress
• Encouragement to self-select moderate physical activities
• Opportunities to participate in supervised and unsupervised programmes of physical activities including aerobics, walking and cycling.
Although it is difficult to directly compare these elements with those identified above in the Kahn review, it seems there is a clear need to provide information and support, and opportunities for monitoring performance. Moreover, across the various settings and population groups in this review, there is convincing evidence that promoting moderate intensity activity such as walking and not requiring attendance at a facility are features of successful interventions. Based on these two reviews it was possible to identify important components for our intervention and to better understand the target behaviour of the intervention – the promotion of moderate intensity activities that are selected by the individuals themselves. Finally, it was clear from the literature review that there is, as yet, no strong evidence base for the design of workplace physical activity interventions.
Step one: Focus group results
On completion of the literature review, the next step was to conduct focus groups amongst a selection of the participating organisations. Table shows the most frequently cited barriers (left panel) and levers (right panel) to engaging in physical activity as identified in the focus group discussions. The figures to the left of each panel indicate the number of focus groups in which each of the barriers or levers was mentioned at least once. The most frequently cited barriers were a lack of time or competing demands on time. A lack of motivation (e.g. can't be bothered, I just want to relax when I'm not working) was another problem highlighted by many. For others, the barriers represented a lack of control (e.g. a lack of facilities) and many people referred to the negative consequences associated with exercising (e.g. being hot and sweaty, being embarrassed or the manager frowning on you taking breaks).
Barrier and facilitators to engaging in physical activity identified in the focus groups
Participants referred to a different set of factors when discussing the factors that facilitate engagement in physical activity. Social support for physical activity was considered important (e.g. being inspired by and engaging in activity with other people). Doing things that were enjoyable was also perceived to support activity. Other factors focused on having a goal, planning and monitoring activity levels and making it part of a routine. Having access to local facilities was also mentioned frequently.
Step two: identification of outcomes, performance objectives and change objectives
As mentioned above the overall target outcome of the current intervention was to increase physical activity. Currently it is recommended that adults should engage in moderate intensity physical activity for at least 30 minutes on at least 5 days of the week [13
]. However, based on focus group discussions it was recognised that this may not be a realistic goal for many individuals who are very sedentary and may be regarded as somewhat unachievable given the 3 month timescale for the actual intervention. It was therefore decided that the objective of the intervention would be to 'increase levels of moderate intensity activity' with a view to achieving the recommended levels. An increase in moderate intensity activity was defined as any increase lasting at least 10 minutes, as this 10 minute period has been posited to be the minimum length required for health benefits [18
]. Focusing on increasing levels of physical activity also meant that the intervention was inclusive to all employees within the worksite, particularly those sedentary employees for whom achieving the recommended levels of physical activity in a short period might be an unrealistic goal. Participants were encouraged to set graded goals for themselves and show cumulative increases in activity throughout the course of the intervention.
Due to differing work patterns and commitments within different organisations it was clear that it was not feasible for all employees to increase the amount of physical activity during their working day. For example, an office worker may be able to utilise lunch breaks or flexible working policies to engage in activity during the day, but a bus driver confined to their cab for 4 hours at a time with a limited break would not find this easy. Therefore the behavioural outcomes focused on increasing moderate intensity physical activity in three areas: a) during work, b) in leisure time, and c) during commute to work.
Interpersonal outcomes were also specified given the importance of social support interventions in achieving behaviour change identified in the Kahn review paper [35
]. Interpersonal outcomes were to d) encourage colleagues to engage in physical activity
and e) to encourage friends and family to engage in physical activity
The next stage of the intervention mapping process was to specify the performance objectives for each of the programme outcomes. In a brainstorming session RL and RRCM listed all the steps that would need to be taken in order to achieve the programme outcomes. This process was informed by theoretical knowledge about determinants (e.g. intention) and facilitators of behaviour (e.g. goal setting, implementation intentions, [40
]). This list was then validated by CJ and MC. The final list of performance objectives can be found in additional file 3
Once the performance objectives had been specified the next stage was to cross these with the theoretical determinants to create matrices of 'change objectives'. For each performance objective we first identified which determinant might be appropriate to achieve it. Then we specified what change we would need in the determinant in order to effect the performance objective. For example, to allow individuals to monitor current levels of activity (performance objective 6), it was deemed that intention, self-efficacy and knowledge should be targeted. Table contains selected examples of change objectives for four different performance objectives.
Examples of change objectives for selected performance objectives
Step three: identifying theoretical methods and practical strategies
The third stage of the intervention mapping process involved identifying appropriate theoretical methods which are thought to change theoretical determinants. Bartholomew and colleagues [20
] have summarised which types of theoretical methods are most appropriate for different theoretical determinants (chapter 7). This was used as a guide in the current process.
The intervention strategies identified as being associated with success in the literature review were also kept in mind when strategies for the current intervention were being developed (for example, self-monitoring, barrier identification). However, care was taken to ensure that all strategies included were based on sound theoretical methods. In addition, the results of the focus groups were reviewed to ensure that the practical strategies identified to implement the theoretically changed methods were appropriate and acceptable to the target group. For example, one strategy identified was having group discussions in which barriers to physical activity were discussed. However, it became clear from focus group discussions that it was not possible to get groups of employees together in the different worksites and that this component would fail as a result. Finally, the proposed strategies were presented to key stakeholders from the different types of organisations. Examples of theoretical methods and strategies related to self-efficacy change objectives for performance objective 6 (monitoring current levels of activity) can be found in Table . For example, one change objective was for individuals to express confidence in monitoring current and ongoing activity levels at work. Theoretical methods deemed useful here were guided practice (explaining how to go about monitoring activity), enactment (actually making the individual go through this process) and persuasive communications. In light of these theoretical methods we decided one strategy would be to provide participants with a 'work-book' where they could record how much activity they performed on different days of the weeks and at different times. It was decided that this might be appropriate in a leaflet format.
Examples of strategies for self-efficacy change objectives for performance objective 6 (monitoring currently levels of physical activity in work)
Stage four: creating an organised programme plan
The first step in stage four was to decide the scope and limits of the current intervention. Hillsdon et al. [36
] comment that many physical activity interventions are delivered at the level of the individual and such interventions may not be economically viable or efficacious for achieving changes across a large population. An expert steering group consisting of health and work psychologists, occupational health specialists, and a physician decided on the core principles of the intervention. Contacts within the participating organisations were also consulted. They had three requirements. The intervention should be 1) flexible enough to be delivered across different sizes and types of organisation, 2) sustainable without the direct input of an expert group, 3) problem focused (i.e. able to work within the restraints of different work patterns and environments). The focus group results helped us to understand the different working contexts and cultures of each organisation and to identify appropriate intervention components. Finally, in order to map onto the three characteristics of successful interventions identified by Kahn et al [26
] an informational, behavioural/motivation and environmental component to the intervention was deemed necessary. Thus, the name 'AME for Activity' (Awareness, Motivation and Environment) was coined.
Next it was necessary to translate the strategies into organised programme components or methods for delivering the strategies. The defining characteristic of the intervention was that it would be delivered 'in-house' by nominated local facilitators and would take no more than five hours each month (for a three month period). The delivery of materials and intervention components via a local facilitator was entirely pragmatic and based on the requirement that this intervention could be delivered in a variety of workplaces without expert input. This also ensured the sustainability of the intervention beyond the involvement of the research team. In addition, as the intervention was to be delivered over a three month period, a drip-feed approach was used in which different messages or activities would be the focus in different weeks. This was to ensure that employees did not habituate to, or cease to register, any of the intervention materials. The theme for the first month focused on health benefits of physical activity, the second month focused on mental health benefits and the final month focused on benefits of physical activity for leading a happy and fulfilling life. These themes were selected to reflect the different types of benefits that physical activity can have on health and life and acknowledging the influence of affective processes on behaviour [e.g. [41
Due to the large number of change objectives and resulting strategies it was necessary to refine strategies into a manageable number of key modes of delivery which could be implemented by the local facilitators. Key considerations in the selection of delivery modes were budget, whether the components could be delivered in all the different worksites, and the amount of time/skill needed by each facilitator to implement them. Four members of the research team (RRCM, RJL, CJ and MC) each reviewed the list of change objectives and strategies and identified the key components that identified the majority of the strategies. Ten components were identified which, although not mutually exclusive, fitted into three themes of 'awareness', 'motivation' and 'environment', (see Table ).
Key components of AME for ACTIVITY intervention
The intervention components differed according to whether they were delivered on an individual level (e.g. leaflets, quiz, plans and targets, self monitoring, newsletters, reminders) or whether they were delivered at a group level (e.g. team challenges). The remaining components drew on existing communication strategies within our collaborating organisations (e.g. posters, management support letters). Emphasis was placed on engaging with group activities and team challenges on offer within the workplace, but if employees did not choose to do this then at the minimum they would receive materials which could be completed on their own. These materials included 3 leaflets (each distributed four weeks apart and covering a different theme) and a monitoring tool which allowed them to keep track of how much activity they achieved each month. These, along with the posters were designed by a professional graphic design agency. A logo was designed to provide a unifying corporate theme to the remainder of the materials which were developed by the research team.
During the process of developing and designing each of these components of the intervention the value of the intervention mapping process was realised. The content of these materials was driven by strategies previously identified. For example, role model stories were integrated into leaflets and newsletters. Participants were encouraged in each leaflet to identify barriers to physical activity and ways of overcoming them.
The final intervention consisted of these 10 components designed to be delivered in a systematic fashion over a 12 week period. The intervention pack for each worksite contained copies of three different interactive leaflets (which encouraged participants to set targets, make plans and provided feedback on their progress) and a 'keeping track' monitoring tool (in the form of an A5 magnet with erasable pen for recording activity levels). These materials were distributed to all participating employees. In addition the intervention pack contained copies of 8 different A3 coloured posters, and electronic templates for newsletters, reminders, letters of management support, quizzes and instructions on how to run team challenges. The suggested timetable for the delivery of each of these intervention components can be found in additional file 4
Stage five: creation of an adoption and implementation plan
Once the intervention strategies had been finalised and their feasibility assessed the next stage was to create a plan for the adoption and implementation of the intervention amongst the target group. A crucial element of this was to ensure that facilitators received the correct training and instruction in order to implement the intervention. Thus, steps 2 and 3 of the intervention mapping process were repeated to focus on the behaviours required from the facilitator to implement the intervention. The outcome of this was a comprehensive facilitator manual including step by step instructions about how to implement the intervention and a suggested timetable for implementing each of the key components. An implementation plan was also developed to provide facilitators with ongoing support throughout the implementation of the intervention. This is currently underway. The research team first contacts the facilitator to arrange a convenient time for the intervention to launch. Then a second member of the research team (CJ) contacts facilitators at monthly intervals to assess progress. As part of the process facilitators are asked to complete an 'intervention' log detailing which components of the intervention are delivered each week. This log will be used as part of a process evaluation on conclusion of the evaluation to determine which elements of the intervention are consistently implemented and which are not.