BUD: A Mobile Application to Improve Procurement and Distribution of Healthful Foods & Beverages in Baltimore City
Study Details
Study Description
Brief Summary
Low-income urban communities have many small food stores, but poor access to healthier foods and beverages. The investigators will develop, implement and evaluate the feasibility of a Baltimore Urban food Distribution (BUD) web-based application (app) to improve access to affordable, healthier products from local producers/wholesalers in 38 urban corner stores in low-income Baltimore neighborhoods, using a randomized controlled trial design and assess its impact on store stocking and sales. The R34 will provide a developed and tested version of the BUD app, which will resolve challenges related to affordability and delivery of healthful foods and beverages to small food stores, permit development of new instruments, assess potential impacts at the consumer level, permitting power and sample size estimates for the full-scale clinical trial, and demonstrate the investigators' ability to recruit and retain large numbers of wholesalers, producers, and corner stores in low-income urban settings.
| Condition or Disease | Intervention/Treatment | Phase |
|---|---|---|
|
N/A |
Detailed Description
The overarching goal of this application is to develop and pilot test a web-based application (app) that will increase access to healthier foods and beverages in low-income urban communities in the United States. Small retail food stores are ubiquitous in low-income urban settings throughout the US and present a unique opportunity to supply surrounding neighborhoods with healthful food options. However, these small stores usually carry few or no foods that are both healthy and affordable. A primary barrier to stocking healthy, affordable foods in small urban food stores is the lack of an adequate distribution network; small store owners generally need to travel on their own to obtain healthy foods and beverages for their stores. Low access to healthy food and high access to food with low nutritional value have been associated with poor diet quality, obesity and chronic disease in many studies.
The study team has worked for more than 17 years in Baltimore to develop, implement, and evaluate chronic disease prevention programs by improving the food environment in low-income communities. The investigators' preliminary formative research assessed the initial acceptability of a mobile app that will enable small urban food store owners to access a range of healthy foods from local wholesalers and producers, and facilitate affordable delivery to their stores. The study team found high acceptability for an app that would leverage the collective purchasing power of digitally-networked small food stores and introduce cost efficiencies into food delivery. For this NHLBI Clinical Trial Pilot Study (R34), the investigators propose to develop a working web-based Baltimore Urban food Distribution (BUD) app, pilot the app, and evaluate its feasibility and impact on the availability, prices and distribution of healthful foods and beverages in East Baltimore, with the following primary aims: 1) To develop and optimize a technically stable and functional digital strategy to overcome small retail food system constraints common in low-income urban food settings; 2) To pilot the BUD app with Baltimore-based producers/wholesalers and corner stores, and assess its feasibility (i.e., acceptability, operability, perceived sustainability, user satisfaction); and 3) To evaluate the impact of the BUD app on corner store stocking (availability, timeliness, quality), prices, and sales of healthy and unhealthy foods and beverages in a pilot study employing a randomized controlled trial design of 38 corner stores. Secondary aims will examine impact on consumers and a cost-benefit analysis for participating retailers and producers.
Findings will permit the investigators to: 1) produce a functional and acceptable web-based app, 2) provide preliminary data needed for power calculations for the full-scale trial, 3) generate and refine process evaluation instruments and set standards for implementation, and 4) establish protocols and demonstrate the study team's ability to recruit and retain large numbers of wholesalers, producers, corner stores and consumers. The study team will assess generalizability of the app by conducting feasibility assessments of the developed app with small store owners and suppliers in other urban settings. The findings from this R34 application are essential to support a full-scale clinical trial, which will test a multi-city deployment of the BUD app and assess its impact on obesity and diet.
Study Design
Arms and Interventions
| Arm | Intervention/Treatment |
|---|---|
| Experimental: Intervention We will pilot the BUD app in 19 intervention corner stores over an 8-month period in East Baltimore. During this time, we will collect data from corner store owners, producers, whole salers, and consumers. |
Behavioral: Web-based application connecting small food store owners and suppliers of healthier foods and beverages
The primary intervention is a web-based app that connects small food store owners in low income Baltimore with suppliers of healthier foods and beverages. To reduce costs associated with small purchasing quantities by corner stores, and high delivery charges, the BUD app uses collective purchasing and shared delivery strategies. BUD will be implemented in four stages, where each stage promotes different food/beverage items and introduces new features. The app will be bundled with a small subsidy in stages 1-2 to encourage initial use, increase familiarity with the app and reduce risk. Trainings in the use of the app will take place at the beginning of each phase. BUD will use collective purchasing at stage 2 of implementation (BuddyUp!). The BuddyLift! feature will start in stage 3, enabling small store owners to deliver BuddyUp! deals to other stores for an additional discount. Participating stores and wholesalers will receive point of purchase materials to promote BUD products.
|
| No Intervention: Control We will collect data from 19 control corner stores over the same 8-month period. They will not receive any form of intervention or delay intervention. |
Outcome Measures
Primary Outcome Measures
- Change in stocking of healthy and unhealthy foods [The Pre measure will be made in the two months prior to starting the intervention; the Post measure will be made in the two month after the end of the intervention]
Stocking of healthy and unhealthy foods will be assessed from Pre to Post intervention at participating corner stores. A Store Impact Questionnaire will capture the number of promoted foods and beverages stocked during each visit, based on direct structured observation of corner store shelves, as well as stocking of unhealthy options (foods and beverages high in sodium, fat and/or sugar). We will create healthy food availability scores (range 0-30) and unhealthy food availability scores (range 0-30). We will calculate change scores, by subtracting each pre measure from each post measure. We have used this approach in multiple studies of the food environment to assess stocking, and changes in stocking associated with interventions.
- Change in sales of healthy and unhealthy foods [The Pre measure will be made in the two months prior to starting the intervention; the Post measure will be made in the two month after the end of the intervention. Measures collected during the trial will occur at the end of each stage of intervention.]
Change in unit sales of promoted healthy and selected unhealthy foods will be assessed from Pre to Post intervention at participating corner stores (recorded over a seven day period). Store owners will be provided with an Android tablet PC uploaded with simple point of sale (POS) software to track sales of targeted foods and beverages, and trained in its use. Sales information collected using the software is time stamped and stored in a secure, password-protected cloud system which can then be accessed and exported in an Excel file by the research team. For each food a unit sales change score will be calculated by subtracting the pre sales number from the post sales number. Overall healthy and unhealthy change in unit sales scores will be calculated. We have tested this POS of sale system in several Baltimore corner stores and found it to be highly accepted by small store owners.
Secondary Outcome Measures
- Change in purchasing of healthy foods by consumers [The Pre measure will be made in the two months prior to starting the intervention; the Post measure will be made in the two month after the end of the intervention]
Adult consumer purchasing of healthy promoted foods and beverages will be captured by the adult impact questionnaire (AIQ) which will be conducted at baseline and post-treatment in a sample of 190 regular customers (5 customers/store in 19 intervention and 19 control stores). Customers will be requested to report how often they purchased specific healthy promoted foods and beverages in the past 7 days, and where these foods were purchased. A score called the "Healthy Food Purchasing Score" will be calculated and have a possible score value ranging from 0-38 points. We will subtract the pre score from the baseline score to calculate change in purchasing of healthy foods by consumers.
- Change in consumption of Healthy Eating Index by consumers [The Pre-measure will be made in the two months prior to starting the intervention; the Post-measure will be made in the two month after the end of the intervention]
Adult customer consumption of healthy and unhealthy foods is captured by the semi-quantitative Block Food Frequency Questionnaire (FFQ) for the entire consumer sample. This validated questionnaire contains questions regarding the frequency and amount of consumption of over 50 commonly consumed foods. The Block FFQ has been validated for use in low income African American urban populations. Investigators will calculate a Healthy Eating Index (scoring ranging from 0-100) using the consumed food frequencies and portion sizes reported on the FFQ. The investigators will assess change in the consumption of healthy foods by subtracting the pre HEI score from the post HEI score.
- Estimated changes (reduction) in operating costs [This measure will be made in the two months after the end of the intervention only via interview.]
Data collection will include a Supplier Fixed and Variable Costs spreadsheet completed by the store, wholesaler, and farm owners which will provide information for the cost-benefit analysis of the treatment for each sector of the Baltimore food system participating in the BUD study. Respondents will be asked to estimate changes in these operating costs from before use of the app to the present. The team will compute savings in operating costs and the unit of measurement will be U.S $. It will consider the fixed and variable costs. Fixed costs will include item costs that will not change with the level of activity (rent of space; interest in long-term loans;). Variable costs will include item costs that change with level of activity (labor expenses; material expenses, transportation costs, utilities).
- Estimated changes (savings) in acquisition prices [This measure will be made in the two months after the end of the intervention via interview only.]
The Supplier Fixed and Variable Costs spreadsheet completed by the store, wholesaler, and farm owners will provide information for the cost-benefit analysis of using the BUD app for each participating sector of the Baltimore food system. The quantity and price of acquisition will be observed and used to compute the total savings. The unit of measurement will be US dollars.
- Estimated total financial expenses [This measure will be made in the two months after the end of the intervention only via interview.]
The Supplier Fixed and Variable Costs spreadsheet completed by the store, wholesaler, and farm owners using the BUD app will provide information for estimated changes in total financial expenses. This data collection will include the financial costs of acquiring and implementing the BUD application. It will consider the fixed and variable costs. Fixed costs will include item costs that will not change with the level of activity (rent of space; cost items that need to be paid even if the producer close). Variable costs will include item costs that change with level of activity (labor expenses; material expenses, transportation costs, utilities). The unit of measurement will be U.S $ monthly, with adjustment (e.g., by total products).
- Changes in prices of healthy foods [The Pre measure will be made in the two months prior to starting the intervention or at the time of introduction; the Post measure will be made in the two month after the end of the intervention.]
Prices of healthy promoted foods and beverages will be collected 5x, as well as of a subset of less healthy alternatives. The SIQ and process evaluation instruments will capture the price of each product during each visit, based on corner store owner report. Most corner stores do not label individual foods with the price, nor are shelves labeled. We have successfully collected pricing information using this method for previous studies in Baltimore corner stores. We will examine changes in prices from the point the food is sold at the store, until the end of the intervention. A change score will be calculated by subtracting the pre price from the post price of each healthy food. We will use this information to compare prices of these healthier foods with those of unhealthy alternatives to help ensure acceptability to customers.
- Feasibility Metrics: User Satisfaction Acceptability, operability, perceived sustainability and user satisfaction with the BUD app Acceptability, operability, perceived sustainability and user satisfaction with the BUD app [Measures collected during the trial will occur at pre-intervention, during the intervention and immediately after the intervention]
During intervention delivery, the 19 intervention corner stores, wholesalers and producers using the BUD app will receive regular requests to rate the app on multiple dimensions (e.g., ease of use, timeliness of delivery, etc.), including open-ended questions. These requests will occur while they are using the BUD app. Ratings will primarily be in the form of close-ended 5 point Likert scale questions (e.g., I am finding the BUD app easy to use.). Respondents will be invited to provide textual information on why scores were given. The investigators will calculate topical and overall user satisfaction scores, and track change in these scores over time.
- Process Metrics: Reach [Measures collected during the trial will occur at pre-intervention, during the intervention and immediately after the intervention]
During intervention delivery, the investigators will assess multiple process evaluation metrics. These will be recorded via in-person data collection, or through the BUD app itself. The first process measures is reach which is measured by the # intervention store owners who actively participate in trial (# receive BUD training, # who access BUD app at least once).
- Process Metrics: Dose Delivered [Measures collected during the trial will occur at pre-intervention, during the intervention and immediately after the intervention]
During intervention delivery, the investigators will assess multiple process evaluation metrics. These will be recorded via in-person data collection, or through the BUD app itself. The second process measure is dose delivered which includes # trainings completed by JHU team/# planned, # promotional materials posted/# planned, and # subsidies provided/# planned.
- Process Metrics: Fidelity [Measures collected during the trial will occur at pre-intervention, during the intervention and immediately after the intervention]
During intervention delivery, the investigators will assess multiple process evaluation metrics. These will be recorded via in-person data collection, or through the BUD app itself. The third process measure is fidelity which includes such measurements as # store owners actively using the BUD app, # times they use the app, and # using materials/subsidies provided.
Eligibility Criteria
Criteria
Inclusion Criteria:
- Inclusion criteria for stores:
Store owner/manager willing and able to order food through a smart phone or other internet-enabled device
Store owner/manager willing to attend in-store trainings in the use of the BUD App
Store located in a low-income neighborhood considered as a Healthy Food Priority Area by the Johns Hopkins Center for a Livable Future111 in East Baltimore
Store located >0.25 miles from a supermarket
Store classified as a small food store (< 4 aisles, < 2 cash registers)
Store owner/manager is English, Korean, Spanish or Mandarin-speaking for first language
Inclusion criteria for wholesalers and producers:
Provide service to Baltimore City (e.g., for producers, this could mean participating in Baltimore City-based farmers markets)
Willing to use the BUD app, including posting and maintaining data on a minimum number of products
Willing to participate with delivery services arranged
Inclusion criteria for consumers (community members):
-
Regular customers of the store (purchase food items at least once a week in the store) identified by the small food store owner/manager enrolled in the study
-
Adult (between 21 years old and 75 years old)
-
Live/work within a 1/2 mile radius from one of the 38 small food stores participating in the study
-
Live in a household of at least 2 persons (criteria intended to provide a more stable sample, to reduce loss to follow-up)
Exclusion Criteria:
-
Anticipate moving out of Baltimore City in the next 12 months
-
Pregnant (due to changes in diet, weight and body composition)
Contacts and Locations
Locations
| Site | City | State | Country | Postal Code | |
|---|---|---|---|---|---|
| 1 | Johns Hopkins University | Baltimore | Maryland | United States | 21218 |
Sponsors and Collaborators
- Johns Hopkins Bloomberg School of Public Health
- National Heart, Lung, and Blood Institute (NHLBI)
Investigators
None specified.Study Documents (Full-Text)
None provided.More Information
Publications
None provided.- 5R34HL145368
- R34HL145368