Produce Rx: A randomized trial to evaluate vouchers to increase the consumption of fruits and vegetables and promote weight loss among food insecure individuals

How to cite

Zeitlin, Andrew; Alva, Maria. 2019. "Produce Rx: A randomized trial to evaluate vouchers to increase the consumption of fruits and vegetables and promote weight loss among food insecure individuals." AEA RCT Registry. AEARCTR-0004074. https://doi.org/10.1257/rct.4074-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Andrew Zeitlin* , Maria Alva

* Primary investigator

Registry ID
AEARCTR-0004074
DOI
https://doi.org/10.1257/rct.4074-1.0
First registered
Last updated
Status
In development
Trial dates
2019-04-01 – 2019-12-31
Countries
Private
JEL codes
H51, I12, I38
Registry record
http://www.socialscienceregistry.org/trials/4074

Abstract

Washington, DC’s low-income communities – specifically Wards 5, 7, and 8 – are on the wrong side of nearly every documented social determinant of health. Of the more than 220,000 individuals in these three wards, 108,000 rely on SNAP, nearly 40% are obese, and 35% are unemployed. Today, there is a 15-year difference in life expectancy between residents of Ward 8 and residents of Ward 3, less than 10 miles away. The top four leading causes of death in DC are now diet-related chronic illnesses. Residents of these predominantly African American neighborhoods have acutely limited access to the resources essential to daily life, including the key need for affordable, healthy food. In clinical settings, patients often cite long distances to grocery stores and limited budgets for food as barriers in taking control of their diets and health.

Individuals that consume less expensive and more calorie-dense foods, which typically have lower nutritional value, are in a vicious cycle of hunger and over-consumption that impairs management of Type-2 diabetes, among other diet-related chronic conditions. Because obesity is a significant risk factor for chronic diseases, effective strategies for weight management are required.

This study will conduct a randomized controlled trial to evaluate the Produce Rx Grocery Store Model. The Produce Rx program aims to improve the nutritional status of Medicaid-enrolled adults diagnosed with at least one diet-related health condition in DC. Produce Rx specifically targets individuals who are currently experiencing or at risk for diet-related chronic diseases. This program combines vouchers for fruit and vegetables redeemable at a local supermarket with the offer of nutritional classes and an incentive for regular primary care visits.

We have designed an experiment to answer three primary research questions:
1. Does Produce Rx increase expenditure on fruit and vegetables?
2. Does participation in Produce Rx improve biomarkers among individuals with prediabetes, diabetes or hypertension?
3. Does Produce Rx decrease total health care expenditures and utilization (inpatient hospital and emergency department visits) among Medicaid managed care members?

We also address (as secondary hypotheses) questions of healthcare utilization and participant satisfaction, which have implications for the sustainability of the program.

4. What are the characteristics of those individuals with high voucher redemption rates versus those with lower rates? Do these differ by subgroups?
5. How do participants perceive the program? What barriers to participation do people report? How do they perceive food and nutrition resources in their neighborhoods and the relationship with their healthcare team compared to controls, after participation?

Intervention

The Produce Rx prescription is equivalent to $20 per week per participant. This represents the total estimated amount a person would need to purchase the recommended amount of daily fruits and vegetables per USDA standards. USDA research using data from 2013 showed that it was possible for a person on a 2,000-calorie diet to eat enough quantity and variety of fruits and vegetables for about $2.10 to $2.60 per day, corresponding to $2.27 - $2.81 a day in 2018 dollars.

In addition, Produce Rx enrollees will be offered nutritional counselling services made available at the point of grocery sale.

These prescriptions must be redeemed every three months by a visit to the participant’s primary care physician.

Experimental design

Medicaid-enrolled participants with chronic diet-related health conditions, as measured in insurer data, will be individually (block) randomized into either the Produce Rx program or a control group. Analysis will be by intent-to-treat.

Design details

Randomization method
Randomization was done in office by computer.
Randomization unit
Randomization is at the individual level. Individuals are randomly assigned to either the offer of the Produce Rx program or to a control group. As distance to the supermarket is an important determinant of access, and as three clinics are anticipated to participated in the study, randomization was blocked by zip code of residence and by the clinic in which the participant is enrolled.
Clusters
6,271 individuals.
Observations
6,271 individuals.
Arms
3,152 individuals assigned to treatment; 3,119 individuals assigned to control.
Primary outcomes
The primary outcomes for the study are as follows: 1. Health biomarkers 2. Fruit and vegetable expenditure 3. Medical costs
Secondary outcomes
We will conduct two types of secondary analyses: those that look at outcomes not defined as primary, and those that look at subgroup heterogeneity in treatment effects, with regard to primary outcomes. The former includes: - Self-reported total fruits and vegetables expenditure (one-week recall), using survey data collected at endline. Subgroup analyses planned for the primary outcomes include heterogeneity by: - Age - Gender - Severity of underlying medical condition based on the following categories: (i) pre-diabetes in the 12-month pre-trial period (5.7 %<HbA1c < 6.5% or 100 mg/dl < FPG<126 mg/dl or 140 mg/dl <OGTT< 200 mg/dl);. (ii) diagnosed with type-2 diabetes mellitus in the 12-month pre-trial period (HbA1c ≥ 6.5% or FPG ≥126 mg/dl or OGTT ≥ 200 mg/dl) (iii) hypertension 130/80 but no other conditions (iv) both diabetes type 2 and hypertension at enrollment.

Cited by

No citation data available for this registration yet.

Keywords