Ghana Financial Incentives Trial Wave II: Spillover and Tuberculosis Screening

How to cite

Duch, Raymond. 2024. "Ghana Financial Incentives Trial Wave II: Spillover and Tuberculosis Screening." AEA RCT Registry. AEARCTR-0013055. https://doi.org/10.1257/rct.13055-1.3

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

Raymond Duch*

* Primary investigator

Registry ID
AEARCTR-0013055
DOI
https://doi.org/10.1257/rct.13055-1.3
First registered
Last updated
Status
Completed
Trial dates
2024-02-19 – 2024-03-29
Countries
Ghana (Cape Coast Metro and Akwapim South)
JEL codes
I12, D9
Registry record
http://www.socialscienceregistry.org/trials/13055

Abstract

The Wave I Ghana Financial Incentive trial confirmed that financial incentives have a positive effect on COVID-19 vaccine intentions, reported vaccination status and verified vaccination status. Wave II provides an opportunity to understand whether this financial incentive effect generalizes to other types of health behavior: Does a similar financial incentive design with tuberculosis screening produce similar positive results? The Wave I trial was designed to identify any spillover effects across treatment arms or on untreated individuals within treated communities. We found no evidence of negative spillover effects of cash incentives on individuals who received no cash compensations. An unanswered question is whether the financial incentives affect within-subject behavior. Wave II examines the willingness of those who received financial incentives in the initial TB screening treatment to adopt similar preventative health behaviors six months later. Wave II is also designed to evaluate the relative impact of financial incentives compared to simple informational reminders.

Intervention

We will implement a pilot for the full study. The pilot will be a scaled-down version of the design described in the attached pdf document. We will test for fewer treatment effects. The reduced sample size will be sufficiently large such that we will be appropriately powered to detect the hypothesized treatment effects.
There are three treatment arms in the pilot study:
TB Health: A 45 second TB promotional and information video (based on the English NHS videos);

TB Health + Text: A 45 second TB promotional and information video (based on the English NHS videos) followed (three days before the screening) with a reminder voice/text message;

TB Health + \$3.00: A 45 second TB promotional and information video (based on the English NHS videos) that includes a 15 second message offering \$3.00 if the respondent receives the TB screening scheduled for their village.

Experimental design

Enumerators prepare a complete itemization of adult members of the household. With a random number generator they select one adult to participate in the study. They record the name of the study participant.
Participants are presented with the purposes of the study and an informed consent form that they must agree to before proceeding to the survey.
Subjects are then presented with one of the short 45-second treatment videos.
Subjects in the Cash Incentive treatment arms are informed that they will receive their $3.00 TB testing compensation via cell phone if show up for the TB testing to be conducted the following week in their village.
Subjects in both are asked a set of questions regarding their intention to get the TB testing the following week in their village.
Subjects receive their participation compensation of \$1.00 in cash.
Working with each of the Health Districts, we will organize a two-day TB screening clinic in each of the villages within two-week of the treatment interventions.

Design details

Randomization method
Working with our partners in each district, we will identify the list of District villages that could feasibly be enumerated by our enumeration team (the primary consideration here was either road access or the quality of the road access). Then, within each District, we will generate a complete list of all of the villages that are candidates for enumeration. The villages were then ranked according to their population size (population statistics provide by the 2020 Ghana census). We then formed groups of five villages by putting five consecutive villages on these lists in the same quintuple. In a typical district we would have approximately 50 quintuples. Note this randomisation method was modified slightly for the Pilot study since we only have three treatments. A more detailed description is available in the attached pdf document.
Randomization unit
Village clusters are the initial unit of randomisation
Clusters
For the Pilot we plan to have a total of 78 villages
Observations
2028 individuals
Arms
676 individuals in each treatment arm.
Primary outcomes
There are two primary outcomes: 1) an expressed intention to get a TB screening after the subjects view one of the three treatment videos; 20 whether or not the subject gets a TB screening within two weeks of the intervention. Three months later the outcome variable of interests is 1) expressed intention to get a hypertension/blood pressure screening; and 2) whether they actually get a hypertension/blood pressure screening two weeks after the intervention.
Secondary outcomes
Whether members of the treated individuals' households get TB screenings.

Materials

Cited by

No citation data available for this registration yet.

Keywords