**4. Discussion**

Implementation of the HWPP was moderately successful in the grocery store setting as demonstrated by good fidelity to program materials, design team engagement in the IDEAS process, and the number of and uptake of program activities in a relatively short time period. This success can be attributed mostly to the design team's interest in the program and the extra time spent by the facilitator to move the team along; leadership support, including lack of active participation by the store management, was the main barrier to further success. The logic model provided an effective and simple framework for evaluating program implementation and allowed us to better understand the workplace factors necessary for success, as well as challenges or barriers that might be overcome with program modifications or additional resources. The HWPP offers multiple tools that can be used in conjunction with this model including the organizational readiness checklist to evaluate Pre-Implementation and the Management Dashboard and PERS tools to evaluate the Inputs, Activities, and Outputs under Program Implementation.

The program inputs (i.e., HWPP program, design team, steering committee, and facilitator) provided a good structure for the program. The HWPP materials were extremely helpful for the facilitator, although the language was somewhat confusing to the design team. High fidelity to the recruitment criteria led to high engagement and enthusiasm of design team members. The design team's interest in health and improving their store was vital to their success. The design team members had strong and consistent attendance and participation during meetings, yet seemed to lack the skills needed to progress through all steps of the program. They proceeded well with the initial steps to assess the workplace, identify problems, and come up with solutions, but struggled with the subsequent steps required to create a realistic plan to present to the steering committee. It is likely the design team members had not previously had the need nor opportunity to use these skills in their jobs. Employees may develop these skills through their jobs or by participating in employee-management teams for other business reasons. However, teams consisting of employees without these skills may be unable to effectively design and implement workplace changes without additional external support or training [18,19,54,55].

As a result, the team required substantial assistance from the facilitator to organize information and develop plans to complete each step of the process. The time demands on the facilitator far exceeded our expectations. It is possible that the steering committee or store management could have assisted the design team with some steps. We were careful to include various levels of leadership (including union representation) on the steering committee; however, there was a discrepancy between the stated support (i.e., help with scheduling design team members and help rolling out solutions) and the actual support received (i.e., design team members often not scheduled to work on meeting days and steering committee took little responsibility for implementing activities). Earlier and more frequent involvement from the steering committee in the design team meetings may have mitigated the need for substantial facilitator resources.

The main program activity, the IDEAS Tool, was delivered as intended. With support from the facilitator, the design team was able to meet, agree on a goal, and develop specific activities for each objective to propose to the steering committee (Steps 1–5A). The team's inability to meet outside of scheduled meetings and the steering committee's prolonged delay in responding to the design team's proposal left no time in the study period to complete IDEAS Step 7 (i.e., Evaluation), or initiate another cycle of the IDEAS process. Without this entire action-feedback cycle, the potential for organizational learning was decreased. This long delay also affected morale and enthusiasm, which resulted in two members leaving the design team. Additionally, the design team's meeting location may have been a problem for some team members. The onsite meeting space was not private; store managers and other employees frequently passed through the meeting space, causing the design team to feel uncomfortable sharing information. Despite these challenges, the program produced worthwhile outputs, demonstrating program success and a positive design team-steering committee collaboration. Overall, the design team had a positive impression of the process noting an increased comradery with team members and healthier behaviors as a result of the intervention. Some team members reported a sense of self-efficacy for continuing the program, while others did not think they could continue without the research team there to facilitate and hold management accountable. Further, data from surveys and interviews showed that store workers were aware of and utilized the workplace activities developed by the design team, indicating relevance to the target audience. Feedback about the methods used for communicating the activities was helpful in explaining possible reasons for non-awareness.

We encountered several challenges during the program that are best described and understood in the pre-implementation and context elements of the logic model. Most importantly, this pilot project grew out of an existing collaboration with a union and three regional grocers. During the planning phases, because only one grocer volunteered to participate and then offered only one store as the test site, we did not have the opportunity to assess organizational readiness at the corporate or the store levels, nor were we able to choose a site that demonstrated readiness to change. While the initial store manager was enthusiastic, he was transferred to another store early in the study and the manager who replaced him was not as invested. The new store manager's lack of interest in the program filtered down to the design team who felt that their efforts were not appreciated. Over time, the design team's level of enthusiasm and engagement in the process decreased. Many previous studies have shown that lack of organizational readiness and leadership support are critical factors to program success [18,19,28,55–57]. The HWPP program materials describe the importance of organizational readiness but do not provide guidance on how to prevent or remediate diminishing leadership support during the course of implementing the program. In our study, we found that the steering committee and store management were less supportive of interventions that focused on addressing workplace problems (e.g., supply order process and communication) and had fewer concerns about those that focused on changing individual behaviors (e.g., walking program). It is possible that the steering committee did not fully understand the purpose of the program and therefore were less willing to support the design team's ideas. Assessing organizational and leadership knowledge of the Total Worker Health approach may be an important part of determining program readiness and the need for education or training before and during program implementation.

We also faced obstacles related to the labor-management structure and agreements and differences between the different unions. The design team was challenged to find activities that applied to employees from the various unions, since the health benefits varied between different unions. This made it difficult for the design team to promote or build upon existing health resources. Due to labor contracts, design team members were not allowed to meet on paid work time. The research team addressed this by paying team members for their time to attend meetings; we do not know if the team's attendance and engagement would have been different had they been allowed to meet on paid work time. Scheduling design team members to work on meeting days also proved to be difficult, which meant that design team members were asked to come in on their days off. These payment and scheduling challenges made some design team members question management's support and willingness to follow through on proposed activities. The issue of paid time to participate on a design team is a problem when trying to implement a participatory program in hourly-paid workers. Management support should include compensating design team members for their time, and ensuring protected time for team members to develop their ideas.

Research has demonstrated a clear link between worker health and productivity, and investing in employee health has become a popular strategy for improving business outcomes [19,55,56,58]; however, many organizations struggle with supporting worker health initiatives when they compete with business objectives [59]. The design team members in this project recognized the need to fit their ideas into the broader business purpose and were thoughtful in creating activities that capitalized

on existing resources or that could be marketed to retail customers in addition to store workers (e.g., premade healthy meals, healthy items near the checkout). While some activities were initially supported by the steering committee, they were not maintained over time because other initiatives, such as holiday product placement, took priority. Additionally, management put little effort into making the existing healthy options for customers more accessible to employees, suggesting that business needs were more important than worker health. This issue of competing interests between business and health is an important contextual factor to consider in interpreting the outcomes of TWH interventions and programs. Other contextual factors that we encountered in this study included seasonality of the work, skill level of employees, rotating employee schedules, and need to put customers first. All of these factors likely influenced the result of the participatory process used in this study, and may impact health and safety initiatives in the retail industry.

Our research study had several limitations. As described, workers were not able to attend meetings during work time and therefore were paid by the research team to attend. The collection of data for the process evaluation may have had an impact on the program's delivery. It is not known if the successful delivery of the program in one store will be generalizable to other retail locations, with different workers, management, facilities, and culture. In addition, we have limited data on whether the observed program implementation had an effect on the health behaviors of workers.

There were also several strengths to the study, including our relationship with the store that allowed us access to employees and support for the research, in addition to the facilitator's strong rapport with the design team. The detailed process measures allowed us to evaluate the fidelity of the program implementation and note which components were problematic and should be improved in future trials. The HWPP materials provided a useful structure and guide to make decisions throughout the process.

Participatory methods like those used in the HWPP may be useful in developing TWH interventions that address a variety of work factors that affect worker health. Our recommendations for those who may choose to use this program are: (1) Assess organizational knowledge, readiness, resources, and commitment; build in time prior to implementation to educate leadership and ensure that they understand the program goals, processes, and expectations; (2) Include and budget for a knowledgeable facilitator who has good communication, planning, and organizational skills; (3) Choose team members who are enthusiastic and have good communication, planning, and organizational skills (or ensure that the steering committee can assist); (4) Schedule in-person meeting time to complete the activities for each step (rather than assume the team will complete things outside of meetings); (5) Customize the worksheets for the audience and add materials as necessary to aid the team through the process; (6) Involve the steering committee early in the process and ask them at the onset of the program to play an active role in planning and implementing solutions; and (7) Build in time and resources for periodic evaluation and modifications that may result from the evaluation.
