Process evaluation of an environmental walking and healthy eating pilot in small rural worksites

Abstract

Small Steps are Easier Together (SS) was a pilot environmental intervention in small rural worksites in Upstate New York in collaboration with Extension educators. Worksite leaders teamed with co-workers to select and implement environmental changes to increase walking steps over individual baseline and to choose healthy eating options over 10 weeks. Participants were 226 primarily white, women employees in 5 sites. A mixed methods process evaluation, conducted to identify determinants of intervention effectiveness and to explain differences in outcomes across worksites, included surveys, self-reports of walking and eating, interviews, focus groups, and an intervention log. The evaluation assessed reach, characteristics of recruited participants, dose delivered, dose received, and context and compared sites on walking and eating outcomes. Emergent elements of participant-reported dose received included: active leadership, visible environmental changes, critical mass of participants, public display of accomplishments, accountability to co-workers, and group decision making. Participants at sites with high reach and dose were significantly more likely than sites with low reach and dose to achieve intervention goals. Although this small pilot needs replication, these findings describe how these evaluation methods can be applied and analyzed in an environmental intervention and provide information on trends in the data.

Introduction

About two-thirds of U.S. women are overweight or obese, and rural low income women have some of the highest rates of obesity (Eberhardt and Pamuk, 2004, Flegal et al., 2010, Patterson et al., 2004). Adult weight gain, physical inactivity, and obesity stand out as important risk factors for breast and other cancers (IARC, 2002).

Ecological approaches that consider individual physical activity and dietary behaviors in the context of multiple social, physical, policy, and economic environments have been suggested as promising approaches to obesity prevention on a population level (Kremers et al., 2006, Sallis et al., 2006, Swinburn and Egger, 2002). Worksite interventions that combine healthy eating with a structured approach to the promotion of physical activity can reach large numbers of employed women over extended periods of time using combined environmental and behavioral approaches (CDC, 2005, Engbers et al., 2005, Tessaro et al., 2002). The results of multi-component worksite environmental interventions are beginning to be reported (French et al., 2010, Lemon and Pratt, 2010). However, such multi-level interventions are complex to deliver and implementers may have difficulty in achieving adequate dose and reach (Johnson et al., 2010, Weiner et al., 2009). Well designed process evaluations are needed to determine key elements of worksite environmental interventions.

In spite of the fact that 85% of all firms have fewer than 100 employees (Census Bureau, 2002), most environmental worksite interventions have been tested with large employers in urban settings e.g. (Pratt et al., 2007). In rural areas three-quarters of firms have fewer than 20 employees (McDaniel, 2001). Challenges lie in the path of effective implementation of health promotion interventions in small worksites in rural areas (Goetzel, Liss-Levinson, Goodman, & Kennedy, 2009). Owners and/or managers in small rural worksites may not believe that they have a role in employee obesity prevention because of privacy concerns, limited space or resources, or lack of on-site expertise (How Employers Wage War On Workplace Obesity, 2008, Linnan et al., 2008, McPeck et al., 2009). There is a need for simple, easy to implement, environmental obesity prevention strategies for small worksites in rural areas.

Small Steps are Easier Together (SS) was designed as a leader-ready, environmental worksite intervention for small rural worksites. Evidence for SS effectiveness in increased walking steps has been reported (Warren, Maley, Wells, Sugarwala, & Devine, 2010). This paper reports on the process evaluation of a pilot test of the SS intervention in five small to medium rural worksites in Upstate New York. Process evaluation is especially important for community-based public health interventions because it sheds light on the development of effective strategies that can be sustained in a variety of real-life community settings (Linnan & Steckler, 2002). The goals of the current process evaluation were to identify determinants of intervention effectiveness and to explain differences in intervention outcomes across worksites. Process data were used to test the main hypothesis that participants in sites with high mean scores on reach, dose delivered, and dose received would attain walking and healthy eating goals significantly more often than participants in sites with lower scores.

This mixed methods process evaluation was designed to use qualitative research methods and findings to shape quantitative data analysis. Specifically the evaluation aimed to use both a priori measures of dose delivered and participant reported elements of dose received to understand intervention elements associated with the behavioral outcomes of this environmental intervention. Dose received is a key element of process evaluations. It is often assessed by asking study participants to report their exposure to a list of essential elements which are determined a priori by investigators. These elements are derived from the theory on which the intervention is based (Linnan & Steckler, 2002). The experience of an intervention by participants, including their own explanation of what made it effective, may differ from that of investigators in important ways (e.g. priorities, motivation) (Weiner et al., 2009). In the current study, assessment of dose received using emergent elements offered a way to incorporate participants’ experiences of environmental interventions in the evaluation. This approach seemed especially appropriate for an environmental intervention in which intervention strategies were tailored to specific sites as part of the implementation design.