Views of policy makers and health promotion professionals on factors facilitating implementation and maintenance of interventions and policies promoting physical activity and healthy eating: results of the DEDIPAC project

A healthy diet, regular physical activity (PA), and the reduction of sedentary behavior (SB) are lifestyle factors known to help prevent or delay the onset of non-communicable diseases (NCDs), including cardiovascular diseases and cancers, and to improve overall physical and mental health [1, 2]. The World Health Organization (WHO) launched a global action plan in 2013 comprising six steps to increase the levels of these protective behaviors at the population-level and to further reduce additional behaviors known to increase NCD risk [3]. One of these steps addresses the underlying social determinants of modifiable risk factors for NCDs by suggesting the creation of health promoting environments. Actions to create health promoting environments may include the implementation of effective interventions and policies aimed at promoting a healthy diet and PA and at reducing SB. Such public health and policy interventions may target behavior change by operating at multiple levels, such as the social level by addressing social norms related to health behaviors in interventions, or by changing aspects of the built environment to impact behavior [4‐7].

In the ‘Determinants of Diet and Physical Activity’ (DEDIPAC) Knowledge Hub [8], one of the aims was to gain a better insight into factors that facilitate adoption, implementation, and maintenance of multi-level interventions and policies to promote healthy eating, PA (and reduce SB). In a first step, two umbrella reviews were conducted [9, 10]. The first umbrella review was aimed at identifying good practice characteristics of interventions and policies promoting a healthy diet, PA, and the reduction of SB [10]. Using the WHO framework [11], 53 good practice characteristics were identified and subsequently categorized into three domains: (1) main intervention/ policy characteristics, referring to the design, targets, and participants, (2) monitoring and evaluation processes, and (3) implementation issues. The aim of the second umbrella review was to identify evidence-based conditions for successfully implementing interventions and policies promoting a healthy diet, PA, and a reduction in SB [9]. Using the RE-AIM (Reach, Efficacy, Adoption, Implementation and Maintenance) framework [12], 83 potential conditions that were identified in 112 publications (50 systematic reviews, 17 stakeholders’ documents, and 45 position review papers) were grouped into five domains: (1) reach in the target population (n = 8; e.g., strategies facilitating recruitment processes, cultural and social issues relevant for reaching target populations), (2) efficacy (n = 8; e.g., satisfaction with implementation, feasibility and acceptability), (3) adoption by the target staff, setting, or institutions (n = 24; e.g., community support for implementation, governmental and legislative involvement), (4) consistency, costs, and adaptations made during the implementation process (n = 43; e.g., accessibility for participants, cultural context of implementation), and (5) maintenance of the effects in individuals and settings over time (n = 3; e.g., sustainability). The vast majority of implementation conditions (73 of 83) were supported by documents referring to both interventions and policies. Only seven policy-specific implementation conditions were identified. For example, increasing complexities of coexisting policies/legal instruments and a lack of politicians’ collaboration in implementation were stated as barriers to policy implementation. Merely three implementation conditions were specified for interventions only: the degree to which an intervention is delivered as intended (compared to the protocol), assessment of fidelity of delivery, and involving available human resources in the delivery. To conclude, results of these two umbrella reviews suggest that the number of research papers on implementation conditions has grown rapidly in recent years.

However, there is still lack of research providing more details on implementation processes from the perspective of health promotion professionals, policy makers, and practitioners working in the field. Glasgow and colleagues [13] noted that many interventions are developed and successfully evaluated but most of them get “lost in translation” because they are not sustainably implemented in the different “real life” settings. To facilitate long-term improvements in PA, sedentary and dietary behaviors, stakeholders in these settings ought to play a major role in the implementation and maintenance of interventions/ policies after completion of research projects. However, little is known about factors facilitating or impeding policy or intervention implementation from the stakeholders’ point of view. Therefore, information collected directly from intervention and policy stakeholders, using qualitative approaches, is important. Hence, the aim of the present article is to explore what health promotion professionals, policy makers, and various other stakeholders think facilitates or impedes the adoption, implementation, and maintenance of interventions/ policies promoting healthy eating, PA, and a reduction in SB.

The results section is structured as follows: Firstly, results from the interviews conducted for the intervention cases are reported followed by those gathered in the policy cases. For each category of the respective interview guides, a brief summary of results across countries is provided. Exemplary quotes of stakeholders’ comments in these countries can be found in Tables 3 and 4. Furthermore, results of the case studies are organized to reflect the three phases of implementation described above.

The results of 40 interviews conducted with stakeholders of 12 intervention and policy cases in five countries suggest five main factors facilitating intervention/ policy implementation and maintenance: a.) the development and/or existence of stakeholder networks supporting and working towards an implementation of an intervention/ policy, b.) newly created or existing structures in settings to support intervention/ policy implementation, c.) continued funding and political support in the respective country or state, d.) standardized training for staff following detailed intervention/ policy implementation protocols, and e.) socio-cultural adaptations or tailoring of intervention/ policy content to fit the needs and context of the targeted population. The majority of factors contributing to a successful implementation were deemed relevant for both intervention and policy cases across countries. Also, whether an intervention or policy was targeted towards dietary changes or changes in PA made little difference in the issues raised by interviewees.

The results obtained in our study resemble findings reported by Mikkelsen and colleagues [32] who outlined commonalities vs. differences in challenges faced during the implementation of three multi-level multi-component interventions conducted in Europe, North America, and the South Pacific promoting healthy eating and PA to reduce childhood obesity. Similar to our findings, the main challenges identified by the authors included the development of stakeholder networks before and during implementation to continuously support the implementation process, the planning of the dose and intensity needed to reach the target behaviors in various populations (including training and education of staff in the settings), and capacity building to warrant sustainability of interventions after research examining the effectiveness of the intervention is completed. Another important issue raised by Mikkelsen and colleagues [32] which was not mentioned by interviewees in our case studies was the harmonization of delivery mechanisms of the different intervention components across intervention levels or settings (e.g., optimization of the timing of different intervention activities, building activities in later stages of the intervention onto those in earlier stages).

In our study, we also looked at factors influencing policy implementation and found that conditions impacting policy uptake, implementation, and maintenance appeared to be similar to those affecting implementation processes of multi-level interventions to promote healthy eating and PA across participating countries (for an overview of commonalities vs. differences, see Table 5). A recent study, taking a multiple case study approach similar to the one we employed in our study, investigated factors impacting uptake and implementation of a U.S. Department of Agriculture snack policy in schools across several U.S. states [33]. In this study, interviews were conducted with stakeholders from nine high schools across eight states to examine perceptions of school snack implementation. Similar to our study, factors facilitating policy implementation included strong internal and external partnerships with other stakeholders and the incorporation of activities mandated by the policy into existing nutrition education programs.

The synthesis of results across countries generated in the DEDIPAC project is currently made available at a designated website (https://​www.​dedipac.​eu/​toolbox/​) which combines new theoretical frameworks to use as a basis for the development of future interventions/ policies [34], current empirical evidence regarding the effectiveness of multi-level interventions and policies in promoting changes in diet, PA, and SB (i.e., findings of studies and systematic literature reviews), and knowledge of conditions necessary to successfully implement interventions/ policies from a stakeholder’s point of view (i.e., findings of systematic literature reviews and qualitative research). The aim of the website is to make this evidence available to researchers and practitioners in the field of health promotion implementing current interventions or planning new interventions and to assist policy makers in shaping future health policies in the respective countries. Because of the heterogeneity in funding situations, systems (e.g., education systems), and political situations in the five countries, country-specific results will be reported by researchers involved in this study which will allow for interpretations of the data and specific recommendations for future intervention/ policy development in light of the socio-cultural context of each country.

A few examples of unique barriers/ facilitators which were reported in single countries are the following. For example, in Ireland, poor road and transport infrastructure (e.g., a lack of pathways and cycle lanes outside of schools) impeded the implementation of a program promoting active travel of students from and to schools. In Germany, stakeholders who had been involved in the implementation of an intervention to promote healthy eating and PA stated that the research institution in charge of implementation had not sufficiently integrated the program activities into existing school curricula. As a result, stakeholders at times felt that day-to-day activities at schools were interrupted or disturbed by the required intervention activities. In Norway, positive consequences of the implementation of the “Keyhole” food label to promote healthy food options were described by stakeholders. As a result of almost 10 years of policy implementation, there has been an increase in the number of products with the label, as well as items purchased by consumers. Reasons for this development highlighted by interviewees were that the “Keyhole” focuses on basic products that are bought by many; that it is now a common Nordic labelling which means that it has a more secure footing; and thirdly that there is a regular dialogue between the government and the industry with a common interest in sustaining the label. Because of that the implementation of the “Keyhole” continues to be well prepared and there is continuous dissemination.

Overall, the qualitative research approach appeared well suited to obtain an overview of conditions facilitating or hindering uptake, implementation, and maintenance of interventions and policies promoting a healthy diet and PA. Nonetheless, employing other qualitative methods, such as fieldwork and observations, may have provided additional information. Furthermore, the generalizability of the results is limited because the case studies were conducted in selected countries not representing the whole of Europe. In addition, our findings may not be generalizable across other interventions or policies. Several methodological limitations regarding the selection of policy and intervention cases and of interviewees were noted. Firstly, the classification of cases for the different countries as either intervention or policy was difficult because often no clear distinction between intervention and policy was made in practice. However, investigators in all countries used the definitions from the original grant proposal of the DEDIPAC study to categorize a case as a policy/ intervention. Secondly, it was not possible to realize the original aim of the study which was to include an equal number of successful, as well as unsuccessful or less successful cases, meaning that the final selection of cases consisted of more cases which were rather successful at implementing and maintaining interventions/ policies. Thirdly, the sample of interviewees was a convenience sample. Willingness to participate in the case studies seemed to be somewhat higher when interviewees were convinced of the effectiveness of the intervention/ policy and when they personally had made positive experiences while implementing an intervention/ policy. Similarly, social desirability may have played a role during interviews because interviewees may have felt more comfortable reporting facilitating than hindering implementation conditions. Both, data collection in a convenience sample of cases and interviewees and social desirability, may have caused a bias in the data gathered in the case studies towards reporting of more positive results, rather than an accurate reflection of the reality of the implementation process. Lastly, a recall bias may have affected responses of interviewees because in some cases the intervention/ policy had been implemented five to 10 years prior to the case studies and interviewees had trouble accurately remembering the implementation phase of a given intervention/ policy. For example, in Germany, the “IDEFICS” intervention had been implemented more than 5 years prior to the interviews.

Also, interview prompts derived from the systematic umbrella literature review on conditions for intervention/ policy uptake, implementation, and maintenance/ transferability [9] may have limited the responses of interviewees. These prompts were used to elicit new information from interviewees on important conditions for the implementation of interventions and policies to alter dietary behavior and/or PA in “real-life” settings. However, our results indicate that most of the conditions that were identified as relevant for implementing interventions and polices in the literature review were also raised by the interviewees in the case studies. Beyond the conditions that were prompted in the interviews, factors highlighted as relevant by interviewees included the consideration of daily routines in settings during intervention/ policy implementation, raising awareness regarding policy/ intervention themes (e.g., nutrition guidelines) in a given setting, environmental conditions, such as poor road and infrastructure affecting policy implementation, and the importance of an international exchange among policy stakeholders of different countries to facilitate learning from experiences made in other countries.

To conclude, five main factors (i.e., development and/or existence of stakeholder networks, newly created or existing structures in settings to support intervention/ policy implementation, continued funding and political support, standardized training for staff, and socio-cultural adaptations or tailoring of intervention/ policy content) facilitating intervention/ policy implementation and maintenance were identified in this study which appeared to be similar across the five involved European countries. Recommendations for both policy and practice derived from these findings include building stakeholder networks in favor of and continuously supporting a given policy or intervention and using or adding onto existing structures in a given system or setting to successfully implement and maintain future policies and interventions. To ensure a standardized implementation across settings or intervention levels and possibly across countries, detailed implementation protocols should be drafted and followed. Funding and political situations, as well as socio-cultural context, require adaptations of intervention/policy content to reach target populations in different countries.

This project is one of the first to provide a stakeholder-informed overview of conditions necessary for effectively implementing interventions and policies targeted towards changing diet and PA at the population-level. The results contribute to a better understanding of factors impacting on adoption, implementation, and maintenance of existing interventions and policies in Europe comparing these processes in different countries. They may also inform intervention/policy stakeholders and policy makers’ decisive actions with regard to future implementation of such interventions/ policies.