Targeting and incentives for prevention
How to cite
Lagarde, Mylene; Ruimallo-Herl, Carlos. 2019. "Targeting and incentives for prevention." AEA RCT Registry. AEARCTR-0004898. https://doi.org/10.1257/rct.4898-1.4000000000000001
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Incentivizing the take-up of preventive health services can be effective to increase access in low-income countries. Yet even when care is incentivized, some individuals may not be sensitive to price because they still fail to see the benefits of using preventive services on their health. This may be particularly the case when individuals suffer from diseases whose symptoms are not salient, or when the hassle costs of setting up an appointment may outweigh its perceived benefits. Hence, policy-makers may be concerned that universal financial incentives for preventive services that fail to target the most at-risk individuals can generate wastage. In this study, we compare three ways of providing a preventive medical check-up for cardiovascular diseases (CVD): (i) free provision of the medical check-up (which includes a blood test and a consultation with a doctor); (ii) receiving a small payment for completing the check-up; (iii) receiving a very small, medium or large financial reward depending on whether the patient is diagnosed, at the check-up, to have respectively a small, medium or large risk of suffering from CVDs in the next 10 years. Comparing take-up of preventive care and behavior change at follow-up across these three groups allows us to test whether varying the level of the incentive by risk status is a cost-effective way to better screen and reach individuals the most at risk.
Intervention
Across all arms, eligible individuals will be given basic information about CVDs, personalized information about their risk factors, and a voucher for a free preventive check-up for CVDs, including a blood test and a medical consultation.
The voucher will come with varying benefits depending on the treatment arm to which individuals are randomly assigned:
1. Individual 'universal' incentive: all individuals who do the preventive check-up receive a small monetary payment.
2. Risk-based incentive: individuals who do the check-up receive a monetary payment which is determined by the outcome of the check-up: if they are found to suffer from a low, medium or high risk of developing a CVD, they respectively receive a low, medium or high payment. While neither the individuals nor the healthcare providers know this risk ex-ante, individuals know whether they have some or many of the risk factors associated with increased risk level.
3. Control. Individuals who do the check-up receive do not receive any additional benefit.
Experimental design
The experiment will be conducted between November and December 2019 in San Salvador, the capital city of El Salvador. Study participants are recruited amongst the clients of a microfinance institution that seeks to improve awareness and prevention of cardiovascular diseases amongst its clientele and offers free access to a medical clinic.
To be eligible, an individual has to be at least 40 years old. After they consent to take part in the study, eligible individuals complete a baseline survey, receive information about cardiovascular diseases and obtain a voucher for a free preventive check-up.
Individuals are randomized by the enumerator to one of the study arms following a randomization sequence generated prior to the start of the fieldwork, using a web-based randomization tool. To ensure a high degree of balance and obtain the desired proportions across treatment groups within each stratum, we stratify the randomisation by a measure of CVD risk (low risk vs. medium or high risk) calculated based on the self-reported information and health outcomes (BMI, blood pressure) collected in the baseline survey (while this measure is expected to be correlated with the measure computed during the check-up it is not necessarily identical).
Design details
Cited by
No citation data available for this registration yet.