Health Education Research Advance Access published online on January 23, 2009
Health Education Research, doi:10.1093/her/cyn065
Do intervention fidelity and dose influence outcomes? Results from the Move to Improve worksite physical activity program
1 Department of Health Promotion and Behavior, University of Georgia, Athens, GA 30602, USA
2 School of Public Health Sciences and Professions, Ohio University, Athens, OH 45701, USA
3 Department of Psychology
4 Department of Management
5 Department of Kinesiology, University of Georgia, Athens, GA 30602, USA
Correspondence to: * M. G. Wilson. E-mail: mwilson{at}uga.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
The purpose of this paper is to evaluate the implementation of the Move to Improve worksite physical activity program using a four step framework that includes the following: (i) defining the active ingredients, (ii) using good methods to measure implementation, (iii) monitoring implementation and (iv) relating implementation to outcomes. The intervention active ingredients consisted of a goal setting behavior change program, a team competition and environmental supports. Intervention fidelity and dose were measured by surveys administered to site co-ordinators, team captains and employees. Implementation was monitored by the use of biweekly assessments that tracked individual physical activity levels and through weekly reports of the project director and site co-ordinators. Latent growth modeling was conducted to determine whether intervention outcomes were affected by site implementation (i.e. fidelity) and/or participation by employees (i.e. dose). Results showed high levels of intervention fidelity, moderate to high levels of intervention dose delivered and moderate levels of the intervention dose received. Level of implementation affected the degree of change in vigorous physical activity (Mean = 5.4 versus 2.2;
2 = 4.9, df = 1), otherwise outcome measures were unaffected by fidelity and dose. These findings suggest that practitioners should focus more energy assuring that the core components are fully implemented and be less concerned about the level of participation. |
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
Despite widespread attempts to increase physical activity levels in the general population of the United States, only a third of adults engage in recommended levels of either moderate or vigorous physical activity [1] and a fourth of adults engage in no leisure-time physical activity [2]. Workplaces offer unique opportunities to implement interventions to increase leisure physical activity, and national public health objectives in the United States have identified workplaces as premier settings for physical activity programs [3]. However, the efficacy and effectiveness of such interventions have been modest [4], possibly because implementation of key intervention components was incomplete or unknown [5]. Here, we report on the fidelity and dose of Move to Improve, a workplace intervention that successfully increased employee's leisure-time physical activity [6].
Flay et al. [7] have proposed standards for determining intervention effectiveness, many of which involve the careful documentation and implementation of the intervention. Research has shown that the level of implementation of an intervention significantly impacts the intervention effects to the extent that higher levels of (or better) implementation are associated with more positive outcomes [8]. Threshold effects have also been reported in the literature, in which positive program effects occur only after a certain level of implementation has been attained [8]. Clearly, documenting the implementation of an intervention by means of process evaluation in an effectiveness study is essential to fully understanding the effects of that intervention and ultimately facilitating its translation into practice [9].
Researchers have discussed the importance of documenting implementation and have included implementation as a component of their evaluation frameworks [10]. Glasgow [10] includes implementation as one of the dimensions of his RE-AIM evaluation framework which includes reach, efficacy/effectiveness, adoption, implementation and maintenance and recommends that implementation be included as part of the CONSORT reporting guidelines. Linnan and Steckler [11] include implementation among their key components of process evaluation including context, reach, dose delivered, dose received, fidelity, implementation and recruitment. Although both authors agree that implementation is the extent to which the intervention has been implemented as intended, Linnan [11] argues that implementation should be a composite of reach, dose and fidelity believing that the extent to which the intervention has been received by the audience is part of the formula. Regardless of the importance of implementation as part of the evaluation process, neither author provides a process by which to evaluate implementation, a critical step in the intervention research process. Durlak [8] has outlined four steps that provide a framework from which to study the implementation of an intervention. These steps include the following: (i) defining the active ingredients, (ii) using good methods to measure implementation, (iii) monitoring implementation and (iv) relating implementation to outcomes.
Defining the active ingredients includes using theory or past research to delineate the intervention's active ingredients in clear operational terms, which should be guided by beliefs explaining why they should be successful [8]. The most common means of translating the program ingredients into practice is through the use of program manuals or software programs in the case of online or computer-based programs. These are also the easiest means of documenting the program's active ingredients. Using good methods to measure implementation involves developing an accurate and valid system for assessing implementation [8]. This should include assessing both the fidelity and dose. Monitoring implementation entails assessing the program's active ingredients throughout implementation. Monitoring implementation can provide greater confidence in program results and provide feedback for improving the program the next time it is implemented [8]. Finally, relating implementation to outcomes involves using implementation data to better understand program effects. This may include conducting separate analyses for sites that implemented the program well compared with sites that did not or comparing participants who received a full dose of the program with those who did not. The advantage of adopting Durlak's steps is that they can be included as part of most evaluation frameworks, including those cited above. The purpose of this paper is to show how these steps could be used to document the implementation of an intervention, the Move to Improve worksite physical activity program. Careful evaluation of the implementation of an intervention may provide valuable information for future implementation and/or translation to other settings and populations.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
Participants and setting
The overall study was conducted at 16 worksites of a large retail organization in the United States and Canada. Using matched-pair random assignment, eight of the sites received the intervention (n = 885) and eight served as control sites (n = 557). Sixty-nine percent of the participants were female, 60% were white, 25% were African-American or Black and 7% were Hispanic or Latino with an average age of 36. All workers were office workers in administrative/support positions. Participation was voluntary and all workers at each location were eligible to participate, with the exception of one very large site where we capped participation. Participation at baseline and final cohort were 41 and 61%, respectively. The study was reviewed and approved by the University of Georgia Institutional Review Board. More details on participant recruitment and assignment to conditions have been reported elsewhere [6].
Define the active ingredients—intervention
Conceptual foundation
The active ingredients of the Move to Improve intervention were based on the conceptual foundation of ‘goal setting’ and ‘social–ecological theory’. Goal setting theory proposes that goals influence behavior via the direction, mobilization and persistence of effort and the adoption of new behavioral strategies to achieve the goal [12]. Ecological theory suggests that interventions should be multilevel and integrate personal and environmental resources (e.g. social support) to increase physical activity [13, 14]. Social–ecological theory when applied to the workplace seeks to utilize the social and communication linkages that exist in the workplace, as well as the structural and management systems that form the operational context. Hence, the intervention had two components: (i) a goal setting component that focused on setting and attaining personal and team goals and (ii) an organizational action component that was designed to activate the social and physical environment to promote physical activity. Within this structure, three active ingredients were evaluated, which included an individual goal setting process, a team goal setting process and environmental supports for physical activity.
Active ingredients
An individual goal setting process was established in order to encourage participants to set weekly goals for 10-min blocks of exercise and/or number of steps taken daily (as measured by a pedometer given to participants upon enrolling). Personal goals were targeted toward meeting or exceeding established public health recommendations for physical activity. These included accumulating 
150 min each week of moderate-to-vigorous physical activity [15, 16] and/or 10 000 pedometer steps each day [17]. In order to help participants set goals that were realistic and attainable, each participant was given an intervention manual, which discussed the nature and importance of setting goals as well as how to overcome obstacles to being physically active, reducing temptations to not exercise, how to avoid relapse, how to stay motivated with competing demands and how to remain physically active beyond the end of the program. The purpose of the intervention manual was to help the participant select and commit to a goal, monitor progress toward that goal and identify and overcome obstacles to goal attainment.
The team goal setting process revolved around a ‘team competition’. The team competition was chosen because it could activate peer social support and take advantage of the identities and competitive spirits of the various work units within the organization. Participants organized themselves into teams which were headed by a team captain. Team captains volunteered for the role or were informally elected by team members. Team captains provided support and encouragement for physical activity and helped track the team goal attainment. Members of the team were encouraged to exercise together, help their teammates overcome obstacles and provide encouragement for continued activity. Goals were also introduced into the team's activities. Teams publicly reported their goal attainment levels on a poster located in a high-traffic area at the worksite (break/lunch room) and competed for awards based on the percent of members attaining their goals over the course of the intervention. Each member of every team that had 75% of its members reach the goal attainment target (accumulation of 150 min of moderate and vigorous physical activity or 10 000 steps per day at least 9 of the 12 intervention weeks) received an incentive.
Finally, ‘environmental supports’ including development of a steering committee, demonstration of management support and environmental prompts were put in place to support physical activity. The steering committee helped co-ordinate program activities and facilitate management support. The need for management support of worksite health promotion programs is well established [18, 19]. This support allowed for increased flexibility among supervisors to enable the team captains to plan, promote and co-ordinate the intervention and participants to attend intervention-related activities. Environmental prompts have been found to be effective in previous worksite studies [20–24]. Environmental prompts in this intervention were designed to serve as a continual reminder of the goals (i.e. 10 000 steps a day) and ways to be physically active at work (i.e. taking the stairs, parking farther from the building).
Methods to measure implementation
The process evaluation measures were primarily developed to assess implementation of the active ingredients, specifically documenting intervention fidelity (degree to which the protocol was implemented as planned), dose delivered (the amount of the intervention delivered) and dose received (the amount of intervention received by the participants) [25, 26]. Table I delineates how these were operationalized and measured in this study.
|
The project director met with each site co-ordinator weekly to discuss what was implemented and the extent of implementation. This was primarily used as a measure of fidelity and dose delivered. For example, did a steering committee exist at that site (fidelity)? Were they actively involved in the intervention, performing their duties as planned (dose delivered)? In addition, three questionnaires were developed to collect data from ‘site co-ordinators’, ‘team captains’ and ‘employees’. Each of these questionnaires asked questions about intervention participation, barriers, support and implementation. The ‘site co-ordinator’ questionnaire included 17 quantitative items, which used a five-point Likert response format [not at all (1), a little, some, a good bit and a great deal (5)]. Site co-ordinators were asked to rate (i) participation levels for team captains and employees (e.g. ‘To what extent did the team captains participate in the intervention activities?’); (ii) support levels for site management, team captains and employees (e.g. ‘To what extent did site management support the intervention activities?’) and (iii) nine potential barriers to implementation (e.g. unclear goals or lack of direction). The ‘team captain’ questionnaire was the same as the site co-ordinator, with the addition of two questions inquiring about site co-ordinator participation and support. The ‘employee’ questionnaire included 12 dichotomous (yes/no) items. Three items inquired about participation (e.g. ‘Did you read the manual?’), four items about team activities (e.g. ‘Did your team set group goals to be more active?’), two items about environmental supports (e.g. ‘Did you see signs posted to encourage you to take the stairs?’) and three about implementation (e.g. ‘Did you receive your incentive in a timely manner?’). These questionnaires were the primary measures of the dose received.
All three questionnaires were administered at the end of the intervention. Every site co-ordinator and team captain was surveyed at their respective worksite, while employees who completed the intervention were randomly selected to complete the questionnaire via phone. Employees were called at their work number and each interview lasted 
5 min and was conducted by one trained interviewer who followed a standardized protocol. If the participant was not reached the first time, a message was left stating they would be called back. After three unsuccessful calls, they were not contacted further. Participation rates for the questionnaires were 100% for the site co-ordinators (8 of 8), 62% for the team captains (59 of 95) and 50% for employees (35 of 70 called—includes those not reached). For reporting purposes, team captain and site co-ordinator responses were divided into high (‘a great deal’ and ‘a good bit’), moderate (‘some’) and low (‘a little’ and ‘not at all’) categories. These data were analyzed using Statistical Package for the Social Sciences (SPSS) Version 15.0.
Monitoring implementation
As part of the goal setting intervention, participants were asked to complete and turn into their team captain biweekly goal assessment sheets that detailed their level of physical activity (pedometer steps and 10-min blocks of time) over the past 2 weeks and their goal attainment. The assessment sheets allowed each participant to self-monitor their level of physical activity. At the same time, participants were asked to set new goals for the next 2 weeks as well as to complete questions designed to reinforce the benefits of exercise and to overcome barriers to exercise performance. In addition, the site co-ordinator for each site and the research team project director completed weekly reports of project activities primarily focusing on implementation of the team competition and environmental supports. Issues and/or problems were discussed and solutions were devised as needed.
Relate implementation to outcomes
A key to linking implementation to outcomes is to conduct outcome assessments so that the influence of different levels of implementation can be assessed [8]. Because this project had an individual behavior change (goal setting) component, a team competition and an environmental support component, we examined whether incomplete implementation of these aspects affected the outcomes. The primary outcome measure for the study was physical activity which was assessed using the International Physical Activity Questionnaire (IPAQ) short form [27] and expressed as hourly participation each week in activities rated according to multiples of metabolic equivalents (MET-hour x week–1). Studies assessing the reliability and validity of the IPAQ have been conducted in 14 centers in 12 countries on six continents using standardized methods. The results demonstrate reasonable reliability and validity, with broad applicability to a wide range of countries and cultures. In this study, physical activity was assessed at baseline, midpoint and end of the 12-week intervention period.
First, an overall rating of the level of implementation of the environmental strategies at each site was devised by the research team. Sites were ranked from highest to lowest based on site co-ordinator, team captains and employees responses to questions about (i) implementation barriers, (ii) participation levels, (iii) support for intervention activities and (iv) overall implementation. These four rankings were averaged to derive an overall ranking for each site. Based on this ranking, sites were split at the median into two categories: high and low implementation (four sites per category). Second, program participants were split into two groups: those who completed 100% of the program as evidenced by completion and returning of the biweekly goal sheets (full participation—six of six sheets returned) compared with those who did not (partial participation—five or less returned).
Latent growth modeling (LGM) [28] was performed using Mplus 4.2 [29] to test the difference in change in physical activity over time between high-implementation sites and low-implementation sites as well as between full participation and partial participation groups. Stage one LGMs were conducted for each group separately for each physical activity level outcome (vigorous or moderate physical activity or walking). For the stage one models, two latent variables were specified for each LGM, initial status (i.e. latent means at baseline) and change (i.e. the slope or trajectory of change across the three time points of the study). Physical activity was expected to increase equally over the three time points, therefore the change structure was specified as linear. The stage one models were combined for stage two LGM tests of differences between the high- versus low-implementation sites and between the full versus partial participation groups on initial status and change in the physical activity variables. In the baseline model, the stage one LGM models were combined and no equivalent constraints were imposed. Next, equivalence constraints were imposed on the initial status variables across the two groups and then on the change variable across the two groups (i.e. high versus low implementation or full versus partial participation). Significance was based on 2 difference tests between the baseline model in which those parameters were freely estimated in each group and the nested models in which the initial status and/or the change variables were constrained to be equal between the groups. A worsening in model fit of the nested models relative to the baseline model indicated that parameters were different between groups. If the initial status variables across the two groups were non-equivalent, then equivalent constraints were removed when testing the equivalence of the change variable.
The chi-square statistic assessed absolute fit of the model to the data and ideally should be statistically non-significant. However, the test is sensitive to sample size, so other fit indices were also used [30]. Values of the comparative fit index (CFI) and the Tucker-Lewis Index 0.90 and 0.95 were used to indicate acceptable and good fit. Values of the root mean square error of approximation 
0.08 and 0.06 were used to represent acceptable and close fit. Values 0.96 for CFI in combination with values of the standardized root mean square residual <0.10 results in the least sum of Type I and Type II error rates, especially in sample sizes 250 [30].
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
Intervention implementation
Personal goal setting
Because the goal setting component was included in a program manual that was given to participants, we are confident that this part of the intervention was delivered in the same manner to each participant (fidelity). According to reports from the site co-ordinator and project director, each employee who participated in the program received a manual, indicating a complete (100%) dose delivered. Of the employees surveyed, 100% reported that they had read the manual and 94% thought the manual was helpful, indicating a high level of dose received of the goal setting component. For purposes of this analysis, a high level of fidelity or dose would be values in the top quartile (>75%), moderate are values in the second quartile (50–74%) and low are values in the third quartile (25–49%).
Team competition
Based on the site co-ordinator and project director reports, 100% of sites had established teams and implemented the competition component of the intervention. This was verified by the team captain survey, again affirming a high level of intervention fidelity. Approximately one-half of participants (51%) returned all the goal sheets, 46% reported meeting as a team during the intervention, 60% indicated their team set goals to be more active and 89% reported they were aware of the activities of other teams at their site. Together, these results suggest a moderate level of success in delivering the intervention dose. Finally, 61% of team captains reported that the other team captains participated at a high level (a good bit or a great deal) with 59% indicating a high level of support for the intervention among team captains. Site co-ordinators, on the other hand, reported lower levels of team captain participation (63% not at all or a little) but conversely reported that 100% of team captains supported the intervention at a high level.
Environmental supports
Steering committee
Six of eight site co-ordinators reported they were able to form a steering committee and five of those six reported that the committee met regularly throughout the intervention, indicating a high level of intervention fidelity and dose.
Management support
The study was approved by management at each site as a prerequisite to conducting the study. Hence, 100% of the sites had upper management support for the program. However, since the program was conducted as part of the ongoing health promotion programs offered to employees, there was no written communication from upper management to document that support. On the surveys, 75% of site co-ordinators and 51% of team captains reported a high level of support from site management.
Environmental prompts
The environmental prompts were produced and printed by the research team and distributed to each study location. As a result, we are confident that the information conveyed was the same across all sites. Eighty-eight percent of the site co-ordinators (seven of eight) reported they were able to post the prompts as planned. Employees reported seeing signs encouraging activity (77%) and checking their team's progress on the poster (80%), indicating a high level of intervention fidelity, dose delivered and dose received.
Implementation and physical activity
Site implementation
There was a significant difference between the models for vigorous physical activity (2 = 5.2, df = 1) but not for moderate physical activity or walking (Table II). Low-implementation sites had a higher initial level (MET-hour x week–1) of vigorous physical activity (Mean = 16.9) than the high-implementation sites (Mean = 12.2), but initial levels of moderate physical activity or walking did not differ according to implementation level (Table III). Because initial status for vigorous physical activity differed between the implementation groups, equivalence constraints of the initial status variable were removed from the model that tested equivalence of change. The high-implementation sites had a greater increase in vigorous physical activity (Mean = 5.4) over the three time points than did the low-implementation sites (Mean = 2.2) (2 = 4.9, df = 1), but there was no difference between implementation groups in the change in moderate physical activity or walking (see Tables II and III).
|
|
Employee participation
There was no significant difference between the models for vigorous and moderate physical activity, however, there was a significant difference for walking physical activity (
2 = 5.2, df = 1) (Table II). Participants who returned all six goal sheets had a higher initial physical activity level (full participation) (Mean = 11.0) than those participants who did not return all of them (partial participation) (Mean = 6.4) (Table III). The change variables for both groups were constrained to be equal and included in the best fitting model for each of the physical activity outcomes. Comparisons of this model to the model with no constraints yield non-significant results for all the physical activity outcomes. Therefore, whether an employee had full or partial participation had no effect on the change in physical activity over the three time points. |
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
This study evaluated the implementation of a worksite physical activity intervention using the steps outlined by Durlak [8]. We were primarily interested in determining whether the level of implementation (fidelity and dose) affected the primary study outcome—physical activity. Although process evaluations are increasing in frequency in the literature [31–34], few have addressed fidelity and dose and fewer have examined their affect on the primary outcome [35]. Our results are intended to provide feedback for researchers and practitioners who may refine future implementation and facilitate translation of this intervention to other settings and organizations. Overall, study sites reported a relatively high level of implementation fidelity (degree to which the protocol is implemented as planned), moderate to high levels of dose delivered (the amount of the intervention delivered) and moderate levels of dose received (the amount of intervention received by the participants). With the exception of vigorous physical activity, which was positively influenced by the level of implementation, neither the implementation of the intervention (fidelity) nor employee participation in the intervention (dose received) affected the primary outcome. Because all physical activity outcomes were increased by the Move to Improve intervention when compared with a health education control condition [6], it appears that lower levels of implementation fidelity were sufficient to increase moderate intensity physical activity and walking (i.e. a threshold effect) but that higher levels of fidelity are associated with larger increases in vigorous physical activity. Since implementation fidelity was largely in control of the site co-ordinator, a dynamic site co-ordinator who closely followed the intervention protocols could have significantly influenced this outcome. No other site variable seemed to be able to explain this affect.
Because of increased demand for health promotion programs and sustained or decreased levels of resources to conduct the programs, health promotion professionals are often required to seek alternative means for delivering worksite programs beyond the traditional one-on-one or small group formats. This has resulted in increased interest in and use of other means of implementation that do not require large amounts of direct contact by health educators or other highly trained professionals. As such, interventions or programs that make use of self-help manuals, web-based materials or peer leaders or coaches have become increasingly attractive. Worksite health promotion professionals, and most likely community professionals as well, are becoming increasingly accustomed to conducting programs by ‘remote control’.
When health promotion professionals implement a program directly, they can be confident it is being implemented exactly as intended. When that professional gives it to someone else to implement, such as a peer leader or coach, then there is the possibility that, even if carefully trained on proper implementation techniques, the implementation would be modified in some slight way either by omission or by commission. That one degree of separation from the health promotion professional increases variability in the implementation process. If the lay health worker in turn gives the program to the participant to follow through, another degree of separation occurs as the participant may choose to not follow through with parts of the intervention for a variety of reasons. So, as one would suspect, the farther the degree of separation from the original source (in this case the health promotion professional) the greater the variability in implementation.
To an extent that was evident in the Move to Improve study. Intervention fidelity, as defined in this study, was largely within the control of the project team. As the control was moved one degree away to the site co-ordinators and team captains, we noticed greater variability in the way it was implemented (dose delivered). When it was moved an additional degree away, even greater variability existed in the way the participants actually followed the program (dose received). Hence, the process evaluation documented high levels of fidelity, moderate to high levels of dose delivered and moderate levels of dose received.
If this in fact is a phenomena that is likely to occur as a normal part of the implementation of an intervention, then it is essential to understand what impact this has on the outcome of that intervention and, as a result, its impact on translation of that intervention to another setting, population or problem. In the case of this study, there was no significant relationship between the level of participation on the part of the employees and the outcome. Thus it is logical to conclude that participants do not need to completely follow every step to benefit from the program. On the other hand, there was some evidence that the level of intervention implementation did affect the outcome, particularly for vigorous physical activity. This would lead one to surmise that the intervention should be completely implemented in order to be fully effective. Hence, it is recommended that health promotion professionals implementing this intervention to increase vigorous physical activity spend the bulk of their energy making sure the core components are fully implemented and worry less about participants completing the program exactly as directed.
As with most process evaluations, the findings are limited by the data collection methods and samples drawn. Without a continuous presence on site (intervention by remote control), we had to rely on self-reports by the site co-ordinators and team captains. As such, it suffers from the limitations inherent in self-reported data. The measures used were largely single-item measures, making it impossible to determine reliability. Although we attempted to collect data from all site co-ordinators and team captains, we had to take a random sample of participants and recognize that non-participation could have affected the results. Finally, we were unable to collect primary data from some sources, site management in this case, and had to rely on secondary sources (e.g. reports from site co-ordinators and team captains).
In conclusion, when translating interventions, it is important to have a clear understanding of the level of implementation and its impact on outcomes. This study was presented as an example of how intervention implementation could be documented and used to draw conclusions that would facilitate future implementation and translation. Future studies on the Move to Improve program and other physical activity interventions are needed to understand the relationship between implementation level and physical activity outcomes and whether or not there are threshold effects.
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
Health Protection Research Initiative (DP 000111) from the Centers for Disease Control and Prevention.
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementReferences |
|---|
1. Barnes P. Physical Activity among Adults: United States, 2000 and 2005 (2007) Hyattsville, MD: US Department of Heath and Human Services, CDC.
2. Centers for Disease Control and Prevention. Trends in leisure-time physical inactivity by age, sex, and race/ethnicity—United States, 1994–2004. MMWR Morb Mortal Wkly Rep (2005) 54:991–4.[Medline]
3. U.S. Department of Health and Human Services. Healthy People 2010 (2000) Washington, DC: DHHS.
4. Dishman RK, Oldenburg B, O'Neal H, et al. Worksite physical activity interventions. Am J Prev Med (1998) 15:344–61.[CrossRef][Web of Science][Medline]
5. Marcus BH, Williams DM, Dubbert PM, et al. Physical activity intervention studies: what we know and what we need to know: a scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity); Council on Cardiovascular Disease in the Young; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research. Circulation (2006) 114:2739–52.
6. Dishman RK, DeJoy DM, Wilson MG, et al. Move to improve—a randomized workplace trial to increase physical activity. Am J Prev Med (2009) 36:133–41.[Medline]
7. Flay BR, Biglan A, Boruch RF, et al. Standards of evidence: criteria for efficacy, effectiveness and dissemination. Prev Sci (2005) 6:151–75.[CrossRef][Web of Science][Medline]
8. Durlak JA. Why program implementation is important. In: Program Implementation in Preventive Trials—Durlak J, Ferrari J, eds. (1998) Philadelphia, PA: Haworth Press. 5–18.
9. Bull SS, Gillette C, Glasgow RE, et al. Work site health promotion research: to what extent can we generalize the results and what is needed to translate research to practice? Health Educ Behav (2003) 30:537–49.
10. Glasgow RE, Lichtenstein E, Marcus AC. Why don't we see more translation of health promotion research to practice? Rethinking the efficacy-to-effectiveness transition. Am J Public Health (2003) 93:1261–7.
11. Linnan L, Steckler A. Process evaluation for public health intervention and research. In: Process Evaluation for Public Health Interventions and Research—Steckler A, Linnan L, eds. (2002) San Francisco, CA: Wiley. 1–23.
12. Locke EA, Latham GP. Building a practically useful theory of goal setting and task motivation. Am Psychol (2002) 57:705–17.[CrossRef][Medline]
13. DeJoy DM, Southern DJ. An integrative perspective on worksite health promotion. J Occup Med (1993) 35:1221–30.[Web of Science][Medline]
14. Stokols D. Establishing and maintaining health environments: toward a social ecology of health promotion. Am Psychol (1992) 47:6–22.[CrossRef][Medline]
15. Pate RR, Pratt M, Blair SN, et al. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. J Am Med Assoc (1995) 273:402–7.
16. Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation (2007) 116:1081–93.[CrossRef][Web of Science][Medline]
17. Tudor-Locke C, Basset DR. How many steps/day are enough? Preliminary pedometer indices for public health. Sports Med (2004) 34:1–8.[CrossRef][Web of Science][Medline]
18. DeJoy DM, Wilson DM. Organizational health promotion: broadening the horizon of workplace health promotion. Am J Health Promot (2003) 17:337–41.[Medline]
19. O'Donnell M. Design of workplace health promotion programs. In: Health Promotion in the Workplace—O'Donnell M, ed. (2002) New York: Delmar. 49–77.
20. Webb OJ, Eves FF. Effects of environmental changes in a stair climbing intervention: generalization to stair descent. Am J Health Promot (2007) 22:38–44.[Web of Science][Medline]
21. French SA, Story M, Jeffery RW. Environmental influences on eating and physical activity. Annu Rev Public Health (2001) 22:309–35.[CrossRef][Web of Science][Medline]
22. Russell WD, Dzewaltowski DA, Ryan GJ. The effectiveness of a point-of-decision prompt in deterring sedentary behavior. Am J Health Promot (1999) 13:257–9.[Web of Science][Medline]
23. Andersen RE, Fransckowiak SC, Snyder J, et al. Can inexpensive signs encourage the use of stairs? Results of a community intervention. Ann Intern Med (1998) 129:363–9.
24. Hillsdon M, Foster C, Thorogood M. Interventions for promoting physical activity. Cochrane Database Syst Rev (2005) CD003180.
25. Nicklas TA, O'Neill CE. Process of conducting a 5-a-day intervention with high school students: Gimme 5 (Louisiana). Health Educ Behav (2000) 27:201–12.
26. Dusenbury L, Brannigan R, Falco M. A review of research on fidelity of implementation: implications for drug abuse prevention in school settings. Health Educ Res (2003) 18:237–56.
27. Craig CL, Marshall AL, Sjöström M, et al. International physical activity questionnaire (IPAQ): 12-country reliability and validity. Med Sci Sports Exerc (2003) 35:1381–95.[CrossRef][Web of Science][Medline]
28. Bollen KA, Curran PJ. Latent Curve Models. A Structural Equation Perspective (2006) Hoboken, NJ: Wiley.
29. Muthén LK, Muthén BO. Mplus User's Guide (2006) 4th edition. Los Angeles, CA: Muthén & Muthén. 1998–2006.
30. Hu L, Bentler PM. Cutoff criteria for fit indices in covariance structure analysis: conventional criteria versus new alternatives. Struct Equation Model (1999) 6:1–55.
31. Campbell MK, Resnicow K, Carr C, et al. Process evaluation of an effective church-based diet intervention: Body & Soul. Health Educ Behav (2007) 34:864–80.
32. Steele R, Mummery KW, Dwyer T. Development and process evaluation of an internet-based physical activity behaviour change program. Patient Educ and Couns (2007) 67:127–36.
33. Hunt MK, Barbeau EM, Lederman R, et al. Process evaluation results from the Health Directions-Small Business study. Health Educ Behav (2007) 34:90–107.
34. Hunt MK, Lederman R, Stoddard AM, et al. Process evaluation of an integrated health promotion/occupational health model in Wellworks-2. Health Educ Behav (2005) 32:10–26.
35. Murta SG, Sanderson K, Oldenburg B. Process evaluation in occupational stress management programs: a systematic review. Am J Health Promot (2007) 21:248–54.[Medline]
Received on May 21, 2008; accepted on December 16, 2008
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||