Measures for Children at High Risk for Obesity

Case Study 1

How to involve communities and stakeholders in your research and evaluation project?


Integrating communities and stakeholders in research, specifically in measurement selection, is critical to producing evidence that can be effectively translated into programs and policies that promote health equity. This is because communities and stakeholders bring critical expertise in areas such as the local culture and context and the types of data that are needed to effectively change policy and promote scalability of programs. Ultimately, such a community-engaged research approach is important for the rigor, generalizability, and translation of research.

This case study describes a partnership between researchers and multiple stakeholder groups to evaluate a primary care-based program to address food insecurity in conjunction with health behaviors and health outcomes in a primarily low-income Latino community. The program integrates a “food pharmacy” to address food insecurity with a “behavioral pharmacy” to address behavior change for chronic disease prevention. Health care providers prescribe produce vouchers that can be redeemed at an onsite farmers market in the clinic. The behavioral pharmacy is a multi-component theory-based 16-week program that is delivered to adults in a group medical visit format. Adults from food insecure households are eligible to participate in both the food pharmacy and behavioral pharmacy.

When working with a group of diverse stakeholders, building the partnership is the first step in the research process. In this study, key partners include patients, county government, local federally qualified health centers, a non-profit that facilitates the behavioral intervention, and an urban farm. These stakeholders have a vested interest in addressing food insecurity and can use evidence to facilitate policy change. Aligning diverse interests is challenging, and so it is important to dedicate time to building the partnership at the beginning. The stakeholders develop a written partnership agreement that describes the goals, values, responsibilities, and other defining features of the partnership. The topics covered in the agreement can be tailored to the needs of the partnership at a given time and can be amended as the partnership grows and changes. This process strengthens trust among partners, which is critical for defining research questions, identifying measures, implementing the research, and disseminating findings. The team also uses co-learning, a strategy that supports integrating communities and stakeholders in the research process. In co-learning, the community educates researchers about the local context, and researchers educate the community about relevant methodologies. This facilitates active participation of all partners in the research process.


Considerations and challenges

How do you balance the research interests of multiple stakeholders and funders?

In community-engaged research, it is important that each partner benefits from their participation in the partnership. For many partners, generating evidence of the effectiveness of their programs can benefit their programming and future funding. In addition, funders often require common metrics to align measurement across multiple grantees to support comparability and generalizability. The stakeholders must also consider the complexity of the intervention. In this particular intervention, multiple strategies are required to successfully implement the food pharmacy and the behavioral pharmacy in primary care. Strategies include physician and staff training, universal screening for food insecurity, distribution of produce in the clinic, and the group medical visits. The partnership seeks to monitor and evaluate the various strategies as well as determine the intervention effectiveness among specific patient subgroups (e.g., patients with chronic conditions, older adults, parents with young children).

How do you balance scientific rigor with considerations of staff and participant burden?

Given that this program is being implemented in busy primary care settings, it is important to consider data that are being routinely collected such as through the Electronic Health Record, the feasibility of adding measures in the busy work flow of a community health center, the literacy and language abilities of patients, and the fit of measures with intervention targets. For that reason, the partnership decided to collect a limited number of measures on a large number of patients and a more extensive set of measures on a smaller subset.


Measure selection

The research team dedicated time to building the partnership upfront, which facilitated the measurement selection process. The partnership prioritized the following measures: household food insecurity, fruit and vegetable intake, physical activity, social connection, and depressive symptoms. For the purpose of this case study, measurement of household food insecurity and fruit and vegetable intake are the main focus. As described in the considerations, for both household food insecurity and fruit and vegetable intake, the partnership designed the evaluation to collect a limited number of measures with a large number of patients and a more extensive set of measures with a smaller subset. This balanced measurement priorities with patient and clinic burden, which was important to the partnership.

To measure household food insecurity, the partnership looked for previously validated instruments such as the Using previously validated instruments enables comparison with other research efforts nationally and globally. The 2-item Hunger Vital Sign was chosen for all patients because it is recommended for use in primary care, is low-burden for clinic staff and patients, and is easy to incorporate into the busy clinic flow. However, given the importance of addressing food insecurity in the intervention design, the team chose to administer the 6-item USDA survey to a smaller subset. The 18-item was deemed too burdensome, given that clinic staff will be administering the questionnaire in the course of usual clinic workflow.

To assess diet, the partnership considered brief screeners as well as more extensive gold-standard measures such as 24-hour recalls. Similar to household food insecurity, the partnership decided to use a brief screener for all participants and to conduct more extensive dietary assessments with a smaller subset of participants. It was important that the brief screener assess fruit and vegetable intake, given the intervention’s focus on produce. It also had to be easy to self-administer or to be administered by clinic staff with limited dietary assessment training. The partnership also identified instruments that have been validated in Latino populations. The University of California Cooperative Extension Fruit and Vegetable Checklist in Spanish was ideal for assessing fruit and vegetable intake among Spanish-speaking Latinos. It has been validated in similar populations and uses culturally relevant photos that make it easy for all patients, including those with low literacy, to complete. For the gold-standard measure, the partnership decided to use three 24-hour recalls with a smaller subset of patients.


Lessons Learned

  • Dedicate time to building the partnership at the outset through strategies such as developing written partnership agreements.
  • Consider co-learning, especially around the local cultural context and measurement methodology to ensure that all partners can take an active role in study design. Ideally, co-learning topics are balanced between community and academic expertise.
  • If the local context makes it difficult to use more extensive measures, consider using brief measures with a larger population and more extensive measures with a smaller subset of the population.