Increasing Engagement With a Healthy Food Benefit
- Conditions
- Purchasing BehaviorsNutrition
- Interventions
- Other: MessagingOther: Financial incentives
- Registration Number
- NCT02486588
- Lead Sponsor
- University of Pennsylvania
- Brief Summary
In collaboration with Vitality/Discovery in South Africa, an RCT amongst adult enrolled in the lowest tier of Vitality's HealthyFood benefit will be conducted. We will compare the effect of various messaging, and financial incentive strategies on healthy food purchasing behaviors in this population.
- Detailed Description
The Vitality HealthyFood benefit is a three-tiered incentive program designed to encourage healthier food choices amongst members of Discovery Health's Vitality program. Once enrolled in the benefit, participating members can receive 10%, 15%, or 25% cash-back on the healthy foods they purchase at selected grocers. And, while overall enrolment in the program is high, there was been difficulty documenting change in purchasing behaviors; healthy foods have remained a consistent 25-30% of food spend, and unhealthy foods at around 15%. In the current study we would like to test whether differing types of messages and financial incentive structures may be more salient and motivating and have a greater impact on food-purchasing decisions. The proposed study will be a collaboration between Discovery Health/Vitality and researchers at the Perelman School of Medicine at the University of Pennsylvania. In the proposed RCT, Vitality members currently enrolled in the lowest tier (10% level) of the HealthyFood will be randomized to various messaging and financial incentive structures. We will then compare the effectiveness of these different interventions at increasing healthy food purchases among enrolled members.
This study will be a randomized field trial of the combination of several theoretically-motivated interventions: varying weekly messaging, financial incentives, and monthly messaging.
Messaging and financial incentives
Weekly Messaging strategies There are three weekly messaging strategies we want to test. The goal of the messages is to increase the salience of the HealthyFood programme for enrolled members. The three strategies were chosen to enable us to compare the impact of no weekly message vs. a general weekly message vs. a tailored weekly message. The messages will be sent via SMS to all subjects and will be delivered on the same day of the week.
No weekly message-A subset of study subjects will receive no weekly SMS message. This is consistent with the current benefit structure.
General weekly message-A subset of study subjects will receive a weekly SMS that provides general information on the HealthyFood programme and on eating a healthier diet.
Personalized (tailored) weekly message-A subset of study subjects will receive a weekly SMS that includes details about their individual purchasing behavior over the past week. This message will highlight both the percentage and item breakdown of their purchases that were healthy vs. unhealthy.
Financial Incentive Structures
There are three different incentive structures we will test in this study. The goal of the financial incentives is to increase members' motivation to purchase healthy foods. Two of the strategies are currently part of the HealthyFood programme, the third is a new structure which introduces a economic disincentive on the purchase of unhealthy foods.
10% cash-back: This is the current cash-back level for all subjects at the start of on the study. This group is included as a comparison for the other incentive structures.
25% cash back: A subset of study subjects will experience an increase in their cash-back percentage from 10% to 25% of healthy food purchases. The purpose of this group is to assess whether the increase from a 10% level to a 25% level of cash-back is sufficiently large to influence purchasing behaviors.
10% + 15% (healthy-unhealthy): A subset of study subjects will become eligible for additional 15% cash-back above their current 10% level. This additional 15% cash-back will be based on the net monetary amount between their healthy and unhealthy purchases. With this structure, an individual will only receive additional cash-back (above the 10%) if they spend more on healthy foods than on unhealthy foods, and will be able to maximize their cash-back rewards by purchasing only healthy foods.
Monthly Messaging Strategies
Currently all Vitality members enrolled in the HealthyFood programme receive a monthly SMS message informing them of their cash-back deposit.
Bundled deposit SMS (standard)-This message will be the same message that Vitality currently sends to members enrolled in the program me.
Unbundled deposit SMS-Given the introduction of a new incentive structure which includes a penalty for unhealthy purchases, we wanted to test a message strategy which increases the salience of these potential financial losses.
By combining these different weekly messages, incentive structures, and monthly message, the study team determined the aforementioned 6 study arms of interest.
Recruitment & Eligibility
- Status
- COMPLETED
- Sex
- All
- Target Recruitment
- 7314
-
For this study, we will focus on the following individuals:
- Adult members of Vitality Health who have activated their HealthyFood benefit in 2014 but have not completed the Vitality Age calculation or Vitality Healthy Check, and therefore are on a 10% cash- back reward level on healthy food purchases.
- Members who have chosen Woolworths as their preferred HealthyFood partner.
- Members with an available email address and SMS-capable telephone
- Members with meeting specified spending cutoff (R1000 total spend in the previous month, at least 90% of which was spent at Woolworth's.)
- Anyone who does not meet all of the above inclusion criteria
Study & Design
- Study Type
- INTERVENTIONAL
- Study Design
- PARALLEL
- Arm && Interventions
Group Intervention Description General Weekly Message Financial incentives 10% cash back, general weekly message, standard monthly message Personalized Weekly Message Messaging 10 % cash back, personalized weekly message, standard monthly message 25 percent cash back Messaging 25% cash back, personal weekly message, standard monthly message Unbundled monthly message Messaging 10%+15% cash back, personal weekly message, unbundled monthly message Unbundled monthly message Financial incentives 10%+15% cash back, personal weekly message, unbundled monthly message General Weekly Message Messaging 10% cash back, general weekly message, standard monthly message Standard monthly message Messaging 10%+15% cash back, personal weekly message, standard monthly message 25 percent cash back Financial incentives 25% cash back, personal weekly message, standard monthly message Personalized Weekly Message Financial incentives 10 % cash back, personalized weekly message, standard monthly message Standard monthly message Financial incentives 10%+15% cash back, personal weekly message, standard monthly message
- Primary Outcome Measures
Name Time Method Monthly % Healthy Food Expenditure Monthly over 6 month intervention period The difference in monthly healthy food expenditures between arms. This measure is defined as the monthly amount spent on healthy foods divided by the total monthly amount spent on food, multiplied by 100.
For those individuals who had no recorded expenditure in a given month, they will be assigned 0% healthy food expenditure that month; in this way, this outcome can be considered a composite score of being engaged in the healthy food program (0%=not engaged) and the level of engagement (if monthly expenditure \>0). Each individual's data will be averaged across the intervention period.
- Secondary Outcome Measures
Name Time Method Monthly % Unhealthy Items Monthly over 6 month intervention period The difference in monthly unhealthy food items purchased between arms. This measure is defined as the monthly number of unhealthy food items divided by the total monthly number of food items.
The average over the full intervention months will be calculated within an individual and compared between arms. To be complementary with the primary endpoint analysis, in a month where an individual is not engaged in the program (0 total monthly expenditure) the worst-score (100) is imputed for that month. A similar MI approach will also be conducted as a sensitivity analysis.Percentage Change in Monthly % Healthy Food Expenditures Between Arms During the 6 month intervention period Examining the difference in healthy food expenditure trends between arms where the % health food expenditure is defined as describe for outcome 1 (above)
Percent Change in Monthly %Unhealthy Food Expenditures Between Arms During the 6 month intervention period Examining the difference in unhealthy food expenditure trends between arms where %unhealthy food expenditure is defined as described for Outcome 2 (above).
Monthly % Unhealthy Food Expenditure Monthly over 6 month intervention period The difference in monthly unhealthy food expenditures between arms. This measure is defined as the monthly amount spent on unhealthy foods divided by the total monthly amount spent on food, multiplied by 100.
The average over the full intervention months will be calculated within an individual and compared between arms. To be complementary with the primary endpoint analysis, in a month where an individual is not engaged in the program (0 total monthly expenditure) the worst-score (100) is imputed for that month. A similar MI approach will also be conducted as a sensitivity analysis.Percent Change for Monthly % Healthy Food Items Between Arms During the 6 month intervention period Examining the difference in proportion healthy food item trends between arms where % healthy food items is defined as described for Outcome 3 (above)
Percent Change in Monthly %Unhealthy Food Items Between Arms During the 6 month intervention period Examining the difference in proportion unhealthy food item trends between arms where % unhealthy food items is defined as described for Outcome 5 (above).
Monthly % Healthy Items Monthly over 6 month intervention period The difference in monthly healthy food items purchased between arms. This measure is defined as the monthly number of healthy food items divided by the total monthly number of food items.
The average over the full intervention months will be calculated within an individual and compared between arms. To be complementary with the primary endpoint analysis, in a month where an individual is not engaged in the program (0 total monthly expenditure) the worst-score (0) is imputed for that month. A similar MI approach will also be conducted as a sensitivity analysis.