What Works to Get the Elderly Vaccinated Against COVID-19? Experimental Evidence from India

How to cite

Duflo, Esther; Vaidyanathan, Girija; Sharma, Garima; Schilbach, Frank; Grela, Erin; McKelway, Madeline; Banerjee, Abhijit. 2021. "What Works to Get the Elderly Vaccinated Against COVID-19? Experimental Evidence from India." AEA RCT Registry. AEARCTR-0007470. https://doi.org/10.1257/rct.7470-2.0

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

Esther Duflo* , Girija Vaidyanathan , Garima Sharma , Frank Schilbach , Erin Grela , Madeline McKelway , Abhijit Banerjee

* Primary investigator

Registry ID
AEARCTR-0007470
DOI
https://doi.org/10.1257/rct.7470-2.0
First registered
Last updated
Status
On going
Trial dates
2019-02-04 – 2021-08-15
Countries
India (Tamil Nadu)
Secondary IDs
ClinicalTrials.gov ID: NCT04870593
Registry record
http://www.socialscienceregistry.org/trials/7470

Abstract

COVID-19 poses a substantial risk to elderly populations; understanding how to get elders vaccinated against the virus is therefore a policy priority. We will conduct an experiment with elders in Tamil Nadu, India to evaluate interventions policymakers might use to raise vaccination rates amongst the elderly. The particular interventions we will consider are: calling elders to inform them about vaccination; encouragement of a "buddy system," whereby elders are accompanied to the vaccination site by another adult who could also get the vaccine; and seeding vaccination information with "gossips," individuals who elders identify as being good at spreading information in the elders' communities. The interventions will be phone-based, and can be implemented quickly and at low cost. This makes them promising strategies the Tamil Nadu government and other governments could use to vaccinate elderly populations against COVID-19.

Experimental design

The experiment will be conducted in a sample of elders living in Tamil Nadu, India. These elders are a subsample of elders from an ongoing panel study on aging in Tamil Nadu.

We will assign elders' communities to one of four information campaigns. A subsequent household-level randomization will determine whether individual elders are informed directly. The four information campaigns communities may be assigned are:

(1) Control: we will call the elders assigned to be given information directly, inform them that they are eligible for the vaccine, tell them where they could get it, and address common concerns they might have.

(2) Buddies: we will call the elders assigned to be given information directly and deliver all of the information given in condition (1). In addition, we will encourage a "buddy system," whereby elders are accompanied to the vaccination site by another adult who could also get the vaccine. We will suggest to elders that they have another eligible adult accompany them to the vaccination site, ask elders to think of such an adult, and then call those adults to inform them about the vaccine and encourage them to help the elder get to the vaccination site so they could both be vaccinated.

(3) Gossips: we will call the elders assigned to be given information directly and deliver all of the information given in condition (1). We will also ask elders who in their communities is good at spreading information. We will then call these "gossips" and ask them to tell their communities that elders are eligible to be vaccinated, where vaccinations are given, and other details to address common concerns about the vaccine.

(4) Buddies and gossips: we will call the elders assigned to be given information directly and deliver all of the information given in condition (1). We will then do the same encouragement of the buddy system as in condition (2). In addition, we will ask elders who in their communities is good at spreading information. We will then call these "gossips," ask them to spread the information gossips in condition (3) were asked to spread, and also ask them to encourage their communities to use the buddy system to get elders vaccinated.

As mentioned above, elders in randomly selected households will not be given information directly.

We will call all elders, both those given information directly and those not, in the subsequent weeks to ask if they have been vaccinated.

Design details

Randomization method
Randomization will be done in Stata.
Randomization unit
Assignment to one of the four information campaigns will be at the community level. Communities will be villages in rural areas and urban blocks in urban areas. The assignment of elders to receive vaccination information directly or not will be at the household level.
Clusters
385 communities and 2,275 households
Observations
3,006 elders
Arms
One-fourth of communities will be assigned to each of the four information campaigns. One-third of households will be assigned to receive no direct information and two-thirds will be assigned to be directly informed.
Primary outcomes
Vaccination against COVID-19

Cited by

No citation data available for this registration yet.

Keywords