To our knowledge, this is the first publication to evaluate the scaling-up of a PA plan to population level using the RE-AIM framework. PAFES was successful in increasing population access to “healthy routes” and in attaining high levels of WPAD participation (Reach dimension). At population level, three out of 10 people receiving PA advice from their health professional followed the recommendation, and the proportion of patients with at least one cardiovascular risk factor who were “sufficiently active” (moderate or high IPAQ-SF score) increased from 2006 to 2010-2013 (Effectiveness dimension). By 2015, the Plan was applied by all PHC teams, all larger municipalities, and in many cases included WPAD celebration (Adoption dimension). Implementation was accomplished with good penetration in all health regions by 2013, with a relatively low use of resources and estimated cost, and by 2013 the Plan was embedded within the health system (Maintenance dimension). Nonetheless, the coverage of PA advice by health professionals was modest after 10 years of PAFES implementation: only four out of 10 inactive adults with at least one cardiovascular risk factor who visited the PHC centre received PA advice (Reach dimension) and almost 90% of advice given was to reinforce the behaviour of active adults (Implementation dimension).
The RE-AIM framework, designed to evaluate the internal and external validity of public health programs and to address important dissemination and generalization aspects [
31], was useful to present the results of a complex, multilevel intervention like PAFES. Data about each dimension of RE-AIM provided valuable information concerning the translation of research to practice [
19]. Incorporating information on RE-AIM dimensions into scaling-up of promising programs improves their uptake and expansion into practice [
32]. Nevertheless, a global literature review on PA promotion programs implemented through PHC centres found no results for implementation studies fully embedded into the health system. An implementation research study from Finland [
33] presents results following RE-AIM dimensions but, after 4 years of implementation, the program was never embedded into the system. A population analysis of the Exercise Referral Schemes in England [
34] presents data only on the Reach and Implementation dimensions, and the process evaluation of London PA Pathway [
35] was based on data from only 6 general practices. Other studies have evaluated the Swedish PA referral scheme [
36‐
40], Welsh Exercise Referral Scheme [
41,
42], Green Prescription in New Zealand [
43,
44], and Let’s Get Moving in Kent, England [
45]; all of these studies analysed effectiveness in a smaller sample of centres and patients.
Reach
Even though coverage results for screening and advice could be considered modest, screening rates showed a 55.2% increase from 2008 to 2015, while advice to inactive adults with at least one cardiovascular risk factor went from one in 10 during the implementation to four in 10 in the maintenance phase. We are unable to compare results with other interventions, as data are not available for similar studies. We do know that a PHC smoking cessation intervention scaled up in Catalonia from 2002 to 2016 achieved 82.7% coverage of screening and 46.4% of advice [
46]
On the other hand, almost half of the Health Survey respondents recalled having received a PA recommendation from their health professional in the previous year, a higher proportion than the 32.8% of people in a German study who recalled being advised [
47] or the 24.2% in Australia [
48].
PAFES is implemented in actual PHC settings, in which PA advice competes with other preventive services and health problems that might be perceived as more important by a health professional with only 10 minutes for each visit [
49‐
51]. On the other hand, screening and advice coverage was based on the EMR, so there could be a degree of underreporting [
52].
Effectiveness
Our proxy for the effectiveness of PA advice (28.3%) was similar to findings of a study in Spanish PHC centres, with 18.8% effectiveness in the intervention group (14.0% for adults younger than 50 years and 23.6% for older adults) [
53]. Numerous other studies have provided evidence that PA advice from PHC professionals has a significant impact in increasing adults’ PA levels [
54,
55], with long-term effect [
56,
57]. In Catalonia, with 72% of people visiting their PHC professional in the preceding year [
58], it is convenient to promote PA through PHC. Data on the effect of PAFES advice on PA levels are being gathered and will be published in the near future. Although it may be early to observe an impact on PA at the population level, the Health Survey shows a general increase of PA between 2006 and 2015. However, it is worth noting that PA increased particularly among women and people with at least one cardiovascular risk factor and between 2010 and 2011, coinciding with the years of higher investment in the Plan.
Even though a modest proportion of the target population received advice at the PHC centre and the total investment was low, the impact on the increment of active adults per the Catalan Health Survey might have been influenced by other factors. Many PHC teams and municipalities networked to implement the Plan in a more intense dose (e.g., Granollers, Barcelona, Manlleu), and other regional and local entities also promoted PA over the period studied. In addition, PA levels are affected by social determinants of health [
59] that were not taken into account in the present study. For example, the increase might be related to recent cultural changes that have been observed, including greater interest in PA and sports in the general population [
60], especially among young men of a high socioeconomic level [
58] for whom sport activity has become a trend in our setting. All these factors might have had a synergistic effect towards the desired impact.
Adoption
The train-the-trainer strategy helped PHC professionals to effectively adopt PAFES [
61]. When training stopped in 2011, it had a direct effect on the stagnation of registering PHC teams, which increased again after training was resumed in 2012. Municipality adoption of the Plan was intensively led by Sports Departments until 2012. Increasing identification tools (PA screening, advice, registry) and community resources at the local level has been linked to an increase in active adults [
62]. In PAFES, those two elements were accompanied by an increase in local networking and collaboration for PA promotion, enhancing intervention effectiveness [
63]. Finally, WPAD showed good adoption with involvement of health, sports and education organizations from all around Catalonia, with a very small investment. By 2015, WPAD was a well-established event in Catalonia. Data on reach or adoption of WPAD celebrations in other countries were not found in the literature review. For most PA campaigns, impact is measured by awareness through health survey questions at population level [
64,
65]; for example, Agita Sao Paulo found that 52.9% of people interviewed were familiar with the program. A question about WPAD awareness should be included in the Catalan Health Survey in order to assess its impact.
Implementation
Professionals’ fidelity showed that, while PA recording increased through the years, most screening and advice was given to already active adults, with only 1% of advice recorded for inactive adults. There may be several reasons for this finding: at the advice variable, “reinforcement” was the first option listed; thus, professionals would more easily record it and may have associated advice with reinforcement. In addition, only the PHC PA Champions were likely to be totally familiar with all of the PAFES variables in the record, a limitation of this variable itself and of the information delivery through the train-the-trainers strategy. Moreover, the Champions are motivated to deliver PA advice, while the rest of the team might be less motivated. Last but certainly not least, a small remuneration was incorporated into meeting the PA target. This appears to have been successful in increasing PA advice and recording, but could have had an inverse effect on program fidelity, as recording an inactive person would go against the remunerated PA target. Setting a paid target is a positive way to reinforce a new program, but has potential adverse effects [
66,
67]. The “advice” variable has been redesigned to address these concerns. During the early phases, increased screening coverage would likely have been a better target, as health professionals often feel uncertain about the effectiveness of their PA advice [
51]. To improve that confidence level, continuos training and tools are needed, such as the PAFES web site and newsletter [
68].
The estimated €5 million overall cost of the Plan should be valued as an investment, especially as physical inactivity has a high cost (€992 million annually in Spain), and there are great potential savings from increased PA (5% reduction in inactive people could save €204 million per year in Spain) [
69]. Most of the costs were indirect and were assumed by the Health and Sports Departments. After 2011, investment was cut almost in half, before the Plan was fully embedded into the system or incorporated into daily practice by PHC clinicians. All of this offers a stark contrast to the finding by Levy et al [
70] at the time of PAFES implementation that US$10/person/year is the minimum investment in health promotion programs needed to achieve health gains.
Maintenance
PAFES survived three political changes and an economic crisis that affected political prioritization and funding, including the PHC budget. PHC teams were under pressure, with very high professional turnover that aggravated the usual lack of time per visit and affected professional motivation. Moreover, health professionals confronted with patients’ socioeconomic and health-related problems might perceive health promotion as less of a priority. To counteract those difficulties, the Catalan Department of Health intensified implementation in three ways: 1) continuation of the train-the-trainer strategy for new PHC PA Champions and yearly reinforcements to the ones already trained, 2) reinforcement of communication mechanisms, and 3) involvement of regional public health professionals. Long-term policy strategies are needed to sustain change in systems and environments, and community and organizational infrastructure is needed to carry out those strategies [
71]. After eight years, PAFES has become institutionalized and embedded into the health system and strategies, even though its political priority and funding diminished. In this stage, Plan components and activities should be carried on to maintain the achievements [
72].
In 2013, PA advice increased at PHC centres due to several factors: staff at most PHC centres (92.8%) had received training and the centres had adopted the Plan; EMR access to PA items had been improved; and above all, a small remuneration was provided to PHC professionals for meeting the PA target. The role of regional public health professionals in local implementation and networking helped sustain the Plan and motivate adoption by municipalities. Beginning in 2015, an online training course was especially designed and offered for free to all PHC professionals.
The establishment of alliances with different stakeholders was a key element that helped maintain the Plan through challenging political and economic times. In contrast, referral to a local PA program proved to be a complex and inefficient task that had not enough evidence of benefit over individual advice or counselling [
54]. Shifting focus to PHC screening and advice at the individual level became a more sustainable and efficient intervention, following Huijg et al. recommendations [
64] that interventions should not be complex and should have a standard protocol and provide intervention materials. The shift in focus reverberated in greater acceptance by health professionals and better local adaptability of the Plan.
The economic crisis may have had various effects on the Plan. On one hand, the cut in resources was detrimental, having an impact on implementation intensity since 2011, which may have implications for public health [
73]. On the other hand, municipalities that initially had been reluctant to adhere to PAFES became interested once the recession started, since PAFES required a very low investment and had high political visibility. In addition, especially when there is a scarcity of resources, networking becomes even more necessary, thus impelling collaborative programs like this one. At population level, a context of high unemployment and economic shortage may be associated with a decrease in overall mortality and an increase in some healthy behaviours that do not require economic resources, especially in countries with a strong social safety net, which is the case in Spain [
74].
A particular strength of this study was the use of a variety of methods to evaluate the scaling-up of the intervention, applying the RE-AIM framework at both the individual and organizational levels. Being able to evaluate plans that have been implemented in real-world settings provides valuable information. Despite these strengths, the study had several limitations. First, the RE-AIM framework was not incorporated into the initial PAFES evaluation design; therefore, the variables chosen for some dimensions may not have been the most appropriate but were the best available in our data. Study data did not allow an analysis of potential inequality patterns in the different RE-AIM dimensions, which is a recommendation for future studies and an aspect to include in the RE-AIM model. Second, adoption and implementation data were gathered at the PHC level and based on total registry, which may not reflect actual screening and advice due to underreporting. Third, we do not have data on the PHC professional doing the recording (physician/nurse, level of adoption), so all registries in a given PHC centre could have been done by a single motivated professional and not by the whole team. Data on individual health professionals’ effectiveness in recording would help to adapt implementation, guidelines, train-the-trainer, and the overall strategy to better meet their needs. Finally, the interventions had different components depending on location (intensity of implementation). We are currently evaluating the health impact of PAFES, analysing PHC registry data on increased PA and taking into consideration the different intensities of implementation by municipalities and PHC teams.
Although the PHC PA Champions survey had a low (24%) RR, respondents were a representative sample of the Catalan PHC PA Champions by profession (general practitioners, nurses and public health professionals) and territorial distribution. Although online surveys are now the usual choice of researchers, for the speed and low cost of data collection [
75], the RR tends to be low compared with mail surveys and has declined in recent years [
76]. In addition, there is a cultural factor to consider [
77]: a 2012 Spanish study comparing online survey RR between various health professions found that PHC has the lowest rate (<33% compared with 63% response by hospital staff) [
78]. Nevertheless, the RR obtained might be interpreted as suggesting our respondents were the most motivated health professionals in our sample. Monitoring population PA levels provides the opportunity to evaluate public health policies and strategies [
79]. Nevertheless, assessing PA through questionnaires has well-known limitations, such as recall and social desirability bias [
80,
81], but can be adequately applied as an activity-ranking instrument [
82]. Until 2016, assessment of PA by the Catalan Health Survey was done with an adapted version of IPAQ. In 2016 the standard IPAQ-SF questionnaire was included, even though it has only been validated for adults younger than 70 years. In 2017, the question “In the past year, did your health professional advise you to walk 30 minutes a day?” was added. Finally, WPAD evaluation was limited to the available data on participation and number of events; inclusion in the Health Survey of question about awareness of WPAD would improve the evaluation of WPAD results. These improvements will yield more reliable results and facilitate future evaluation of PAFES effectiveness.