Information Constraints and Preventive Behavior amongst COVID-19 Risk Groups in Khyber Pakhtunkhwa, Pakistan

How to cite

Rogge, Lisa; Landmann, Andreas; Vollmer, Sebastian; Noon, Muhammad Jawad; Khan, Zohaib; Khan, Sheraz. 2020. "Information Constraints and Preventive Behavior amongst COVID-19 Risk Groups in Khyber Pakhtunkhwa, Pakistan." AEA RCT Registry. AEARCTR-0006307. https://doi.org/10.1257/rct.6307-2.0

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

Lisa Rogge* , Andreas Landmann , Sebastian Vollmer , Muhammad Jawad Noon , Zohaib Khan , Sheraz Khan

* Primary investigator

Registry ID
AEARCTR-0006307
DOI
https://doi.org/10.1257/rct.6307-2.0
First registered
Last updated
Status
Completed
Trial dates
2020-08-20 – 2020-11-30
Countries
Pakistan (Khyber Pakhtunkhwa)
Registry record
http://www.socialscienceregistry.org/trials/6307

Abstract

As the COVID-19 pandemic progresses, strict lockdown measures are eased and individual preventive behavior might decrease. At the same time, neither effective treatment nor vaccination are available, so that it remains crucial to shield in particular the population groups that are at higher risk to experience a complicated disease course from an infection. Such individual preventive behavior is particularly relevant in resource-constrained settings, where access to adequate healthcare services is limited, such as in our study location in Pakistan. To motivate the adoption of such measures, the government of Pakistan has issued detailed recommendations for preventive actions for its population and spread this message through various channels. Data from our telephone survey with low-income households shows though that gaps in knowledge and practice prevail. We are therefore evaluating whether a more targeted and personalized information campaign of the local public health service (Social Health Protection Initiative) can contribute to narrowing this gap, especially for the at-risk population.
The expected outcome of this trial is twofold: first, we test whether the SMS-based information campaign can increase the knowledge and adoption of recommended actions in both the general and the at-risk population. Beyond the effectiveness in both groups, we test whether light personalization based on the risk group increases this effect. Secondly, through the combination of the interview and health insurance enrollment as well as claims data, we plan to demonstrate the scale-up potential of such an intervention. This will on the one hand shed light on a low-cost complement in the control of a pandemic and showcase the possibility to use administrative data to effectively target the most vulnerable when resources are limited.

Intervention

The intervention consists of six text messages informing the recipients of the risk groups for a complicated COVID-19 disease course as well as preventive measures against a COVID-19 infection.

Experimental design

Study participants are sampled from the list of 1.5 million households in the Pakistani province Khyber Pakhtunkhwa who are enrolled in the free public health insurance. A randomly selected subset was already part of our telephone survey on attitudes, practice and knowledge regarding COVID-19 between April and July 2020, so that we know that around 70% of households have at least one member who is at increased risk for a complicated COVID-19 infection based on age and preconditions. This subset forms the basis of the intervention sample: Within the risk group, 2/3 are allocated to the treatment group, of which half receives the same generic messages as the non-risk group and for the other half the information on the risk groups is personalized to the condition(s) that apply to their household. Similarly, two thirds of the non-risk group receive the intervention, while the remainign third serves as the control group and receives no messages. To reach the desired sample size, an additional sample is drawn from the household list, of which two thirds are randomly allocated to the treatment and one third to the control group. Outcomes will be measured one week after completion of the intervention through a telephone survey and over the course of 2 months through calls to a telemedicine helpline.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Household
Clusters
NA
Observations
Intervention sample: 21500; Interview target: 2500
Arms
Interviewed sample: 418 personalized messages, 418 generic messages and 418 control in risk group; 343 generic messages, 172 control in non-risk group Additional sample (exact distribution across risk groups and personalization will only be known after the post-intervention interviews): 13,333 treatment; 6,667 control
Primary outcomes
Uptake of preventive practices
Secondary outcomes
Knowledge of risk groups, knowledge of preventive practices

Materials

Papers from this trial

  1. Rogge, Lisa; Khan, Sheraz Ahmad; Khan, Zohaib; Noon, Muhammad Jawad; Landmann, Andreas; Vollmer, Sebastian; 2021, "The Effect of Personalized Health Information on Preventive Behavior amongst COVID-19 Risk Groups: a Randomized Experiment in Pakistan", Courant Research Centre Discussion Papers No. 289 http://www2.vwl.wiso.uni-goettingen.de/courant-papers/CRC-PEG_DP_289.pdf

Cited by

No citation data available for this registration yet.

Keywords