REDEEM trial: The effect of individual and mixed REwards in DiabEtEs Management, a randomised controlled trial

Studies in developed countries have demonstrated benefits to preventive care (diet modifications, exercise) among patients with diabetes. Despite this knowledge, patients often don't manage their diabetes properly. In a study that we conducted in Lima, 71% of subjects with diabetes were aware of their disease, 40% were receiving treatment, and only 7% had achieved therapeutic goals. In Latin America, diabetes has a great impact on the health system due to high costs of disease control (US$10.7 billion) and indirect costs (US$54.5 billion) related to loss of productivity, disability, and premature death.

Payment for achieving therapeutic goals, or "rewards," have been tested as a way to improve health. Several studies have shown the promise of this approach in encouraging and maintaining weight loss, which is an important way to control diabetes. One study found that paying people to lose weight led to significantly greater weight loss. In another study in which the payments were linked to the success of multiple members of a group achieving their weight loss goals, the participants not only lost more weight, but also maintained the weight loss. These results are intriguing, and suggest that the study of rewards in self-management of diabetes requires further research.

The approach of paying people to achieve health-related goals has also worked in developing countries. For example, Juntos is a program in Peru that pays pregnant women and mothers to utilize certain health services. On study found that Juntos increased the probability that children in households met their weight-height goals. In addition, families that were part of Juntos were more likely to seek professional health assistance in case of any disease, and the pregnant women were more likely to have their delivery attended by a skilled professional. This type of program takes advantage of the strong family structures in Latin America, and brings an element of group responsibility to the improvement of health. The individual responsible for fulfilling the requirements is not only encouraged to be responsible because of an individual reward, but to improve the situation of the family. This highlights the potential to use responsibility to others to design more effective policies.

In this project, our goal is to determine the effectiveness of two different economic reward programs in promoting preventive behaviors and improving the health of individuals with Type II diabetes in Peru. In addition to rewarding patients for their improved health outcomes, we will go one step further and evaluate a cash incentive for a family or friend supporter, an approach we believe has great potential and may be more effective. In our study, we will enroll diabetes patients and randomly assign them to one of three groups. In one group, patients will receive cash rewards if they achieve certain goals related to management of their diabetes. In a second group, patients and a family or friend supporter will both receive a cash reward if the patient achieves their diabetes management goals. This will allow us to test whether or not the element of responsibility to others improves the ability of the patient to achieve their goals. Finally, a third group will not receive any cash reward. This is the "control" group to which the other groups will be compared to measure success of our approaches.

We expect that after nine months of rewards, patients will effectively improve the preventive behaviors. Six months later, we will also assess if these improvements have been maintained without the cash reward. If our study is successful, it could be a way to encourage people with diabetes to better manage their disease. While there are costs associated with this approach, it could save money in the long run by reducing the complications of diabetes that occur when people don't manage their disease.

There is strong evidence of benefits of preventive care among patients with diabetes, but adherence to guidelines worldwide is quite low and this constitutes a major problem across healthcare systems. Cash rewards have been tested in adherence to treatment for substance abusers, and to promote weight loss and smoking cessation. Less is known about the effect on behaviour of pairing individual cash with group cash rewards or cash rewards to a family member or friend who provides support in the process.

This study will explore the ability of rewards to change the behaviour of individuals with type II diabetes in a randomised controlled trial. Two different nine-month reward strategies will be tested against usual care (control group). One arm will receive individual-only cash rewards. A separate arm will receive mixed rewards, which consist of individual cash reward plus a cash reward to a supportive individual chosen by the study participant. The primary outcome will be metabolic control rates as per HbA1c criteria.

Research Question 1 will compare the effect of each of the intervention arms against the control group on the primary study outcomes at nine months.

Research Question 2 will allow a comparison between the two reward interventions in terms of behavioural-related goals.

Research Question 3 will explore, within each of the intervention groups, if the intrinsic motivation sustains six months after the removal of the reward.

Fourth, throughout the study qualitative data will be collected. If the trial is successful, e.g. if Research Questions 1 and 2 do show an effect, we will be well positioned to gain an understanding about the pathways on which the intervention operates to achieve the observed differences. In addition, we will complete a cost-effectiveness analysis to demonstrate the feasibility of the proposed approach, if our intervention shows a benefit.

One of the main beneficiaries of this research are diabetes patients, as our intervention could greatly improve their health outcomes. Clinical trials in developed countries and population-based studies have indicated that there are major benefits to be achieved by preventive care among patients with diabetes. However, despite the medical consensus on the effectiveness of preventive measures, achievement of control targets by patients worldwide is quite low. In this application, we aim to explore the role of rewards to change the preventive behaviour of individuals with Type II diabetes in a randomised controlled trial in Peru. If we are successful and our intervention is implemented, it could have a significant impact on the health outcomes of patients with diabetes. This will benefit them personally, and will also benefit their families by extension, both emotionally and financially.

In addition, policymakers can benefit from this research, as it can help them to design improved interventions for the management of diabetes patients. For example, Peru's Ministry of Inclusion and Social Development (MIDIS) already recognizes the value of economic rewards and have implemented them in other health-related areas. They will be very interested in the results of our study, and could support and benefit from its implementation. Our intervention could also benefit the financial state of the health care system by reducing the costs of morbidity and mortality associated with diabetes. For example, ESSALUD, the social health insurance in Peru, could benefit from our intervention, as it could lead to reduced morbidity and mortality in patients with diabetes, leading to a reduction in health care costs. We have included letters of support from both MIDIS and ESSALUD, demonstrating their interest and potential involvement in the future of our intervention.

Finally, academic researchers will also benefit from our research. Mixed rewards have not been explored in a health care context. These types of rewards may improve performance when the agent has low knowledge about their ability and cost of their action, which is of particular relevance in the case of diabetes management. Peer pressure may elicit higher individual effort than pure cash rewards, and sentiments of trust and group belonging may create additional pressure for individuals to perform better in mixed reward than in pure cash rewards. If our results are promising, they could lead to studies of this type of intervention in other areas of health care.