Harnessing Social Media for Public Health: A Case Study of Malaria Prevention in India
The digital revolution has brought unprecedented levels of connectivity to even the most remote corners of the globe. In low- and middle-income countries, internet access has experienced exponential growth, with a staggering 62% of the population now online. This widespread connectivity has opened up new avenues for disseminating vital health information, prompting governments, researchers, and development organizations to explore the potential of social media platforms for public health campaigns. However, the inherent nature of social media algorithms, which prioritize engagement and active users, raises concerns about the equitable reach of these campaigns. Do these efforts effectively target and impact the most vulnerable populations, particularly those at highest risk of contracting diseases like malaria?
A recent World Bank working paper sought to answer these critical questions by evaluating a malaria prevention ad campaign conducted on social media platforms in India. The research focused on understanding the campaign’s effectiveness in promoting malaria prevention behaviors, encouraging timely medical attention for fever symptoms, and, crucially, determining whether the campaign reached those most vulnerable to malaria exposure. The study identified dwelling type – specifically, whether homes were constructed with solid or non-solid materials – as the most significant household-level predictor of malaria risk. Individuals residing in non-solid dwellings were deemed to be at higher risk of contracting malaria.
The campaign, spearheaded by Malaria No More, an international non-governmental organization, ran for six months across 22 states in India during 2020-2021. Employing a targeted approach, the ads were tailored for specific audience personas to maximize their resonance and impact. To rigorously assess the campaign’s effectiveness, a cluster-randomized control trial was implemented across three states: Uttar Pradesh, Chhattisgarh, and Jharkhand. Eighty districts were randomly assigned to either a treatment group, exposed to the malaria prevention ads, or a control group, which did not receive the ads. This rigorous methodology allowed researchers to isolate the campaign’s impact and draw meaningful conclusions about its effectiveness.
An innovative aspect of this study was the use of multiple data sources to measure behavior change. Data was collected through chatbot surveys, enabling direct interaction with social media users, and from health facility records through India’s Health Management Information System (HMIS). This combination of methods provided a comprehensive view of the campaign’s impact, both from the perspective of individuals and from objective health data. The use of HMIS represents a significant advancement in evaluating social media campaigns in developing countries, providing a more robust and reliable measure of real-world impact. Data collection occurred both during the campaign to capture immediate effects and up to three months after its conclusion to assess longer-term impacts. The analysis focused on intent-to-treat (ITT) effects, acknowledging that individuals within treatment districts may not have directly seen the ads, offering a more conservative estimate of the campaign’s true reach.
The study’s findings revealed a nuanced picture of the campaign’s effectiveness. While the campaign demonstrably increased bed net usage overall, its impact varied significantly based on dwelling type. During the campaign period, bed net usage increased by 3 percentage points, representing a 4.5% rise compared to the control group. However, a deeper dive into the data revealed that the increase was primarily concentrated among users living in solid dwellings, where bed net usage rose by 9%–11%. No significant impact was observed among users residing in non-solid dwellings, highlighting a disparity in the campaign’s reach and effectiveness among different vulnerability groups. Furthermore, individuals in solid dwellings were 13% more likely to seek medical care within 24 hours of experiencing fever symptoms, further reinforcing the campaign’s disproportionate impact on lower-risk populations.
Longitudinal data collected one to three months after the campaign’s conclusion provided further insights into the campaign’s lasting effects. Continued exposure to the campaign messages resulted in a measurable reduction in malaria incidence, again primarily among individuals living in solid dwellings. Self-reported malaria cases in this group decreased by a substantial 44% at the extensive margin (whether any household member contracted malaria) and by 53% at the intensive margin (the proportion of household members who contracted malaria) compared to the control group. Interestingly, the reduction was most pronounced among households already owning bed nets, suggesting that the campaign served as a reminder to use existing nets rather than prompting new purchases. This finding underscores the potential of social media campaigns to reinforce existing healthy behaviors. The HMIS data corroborated these findings, revealing a decrease in malaria incidence in urban areas, where solid housing is more prevalent. Monthly malaria incidence fell by 6.2 cases per million people, equivalent to a significant 30% reduction from the pre-campaign monthly rate, over the nine months following the campaign launch. However, no significant change in malaria incidence was observed in rural areas, where non-solid dwellings are more common, further emphasizing the uneven impact across different vulnerability groups.
This study highlights the potential and the challenges of utilizing social media for public health interventions. While the malaria prevention campaign demonstrated positive impacts on bed net usage and malaria incidence, these benefits were largely concentrated among lower-risk populations residing in solid dwellings. The lack of significant impact on those living in non-solid dwellings, a group identified as being at higher risk of malaria, raises crucial questions about the equitable reach of social media campaigns. Did the campaign messages fail to resonate with lower-income users, or did the inherent biases of social media algorithms, which prioritize engagement and activity, limit the campaign’s reach among these vulnerable populations? Further research is needed to unpack these questions and develop strategies to ensure that social media campaigns can effectively reach and impact those who stand to benefit the most. This study serves as a valuable contribution to the growing body of knowledge on utilizing digital platforms for public health, highlighting the need for careful consideration of equity and access in the design and implementation of these initiatives. As internet access continues to expand globally, understanding how to leverage social media platforms to effectively reach and engage vulnerable populations will be crucial in achieving equitable health outcomes for all.