Seminar Summary – Convenient Access and Invitations: Increasing COVID-19 Vaccinations in Kenya

By Naomi Frim-Abrams

On March 20, 2024, the Spring 2024 Human Capital Initiative Research Seminar Series hosted Elisa Maffioli, Assistant Professor of Health Management and Policy at the University of Michigan, to discuss her forthcoming study on the effect of social influence on vaccination behavior during COVID-19 in Kenya. The study evaluates a campaign where healthcare providers would visit homes and invite adults to be vaccinated at a site nearby, randomizing the announcement of the visit ahead of time.

COVID-19 vaccines are a vital public health intervention, but are not yet widespread globally, especially in the study area of Kenya. At the time of the study, in Summer 2022, there were 337,000 confirmed cases and 5,670 deaths from COVID-19 in Kenya, and 56 percent of the study sample population had never received a dose of a COVID-19 vaccine.

There are positive externalities to vaccines, and social influence can play a role in vaccine uptake. Maffioli and her research partners focused their study on two research questions: Can convenient access and invitations increase vaccination uptake, and does social influence promote vaccination?

The literature review examined supply and demand barriers to vaccine uptake, as well as the role of social pressure with a specific focus on the influence of healthcare providers. Prior research explored obstacles such as vaccine hesitancy and access; Maffioli’s project builds on these ideas by providing both convenience of vaccination and personal invitations through healthcare workers.

Elisa and her fellow researchers decided to follow the Kenyan government’s footsteps and mimicked their national COVID-19 vaccine deployment campaign. The researchers tried to address issues of both supply (setting up local vaccination sites) and demand (nudges by sending healthcare workers door-to-door) to improve vaccine uptake. They utilized a one-day vaccination campaign intervention to see the effects of the healthcare workers’ home visits more clearly, as the researchers could control for the spread of information and make sure that the campaign was being promoted by the healthcare workers alone, rather than news spreading through other community members. Researchers met with village elders to determine ideal sites for vaccinations that were public and identifiable such as churches, schools and community meeting spaces. The sample included randomly selected residents within a 15-minute walking radius of these sites.

The treatment used in this study, whether a healthcare worker’s visits were announced or unannounced, was randomized, and participants were not aware of which group they were part. Participants in the “announced” group were informed that a healthcare worker would visit their home within 30 minutes to ask if they wanted to receive a COVID-19 vaccine at the nearby site. Participants in the “unannounced” group received the health worker visit without being informed ahead of time.

Overall, the authors found an increase in dose uptake at the day of the campaign, and the effects were consistent after three months since the intervention. However, they found no evidence of social pressure by healthcare workers: those in the announced group were instead more likely to go and get vaccinated. Maffioli also discussed cost-effectiveness and policy implications. The researchers calculated that the marginal cost per marginal dose of their intervention, based on their data on recorded costs and the number of doses, was $41.76; for context, the most similar study on COVID-19 vaccinations in Sierra Leone estimated the implementation cost to be $33 per additional person vaccinated. In terms of policy, she spoke to the need to test alternative approaches to find even more cost-effective strategies, such as training community health volunteers or using alternative public transportation.

Throughout the presentation, the audience raised various questions about both the study elements and results. Maffioli first addressed comments as to whether historically there has been backlash against vaccination campaigns in Kenya, and if there is mistrust of medical personnel; she responded that there are actually high levels of trust in medical interventions, however, this was not entirely relevant to her research, as her team was addressing social pressure in uptake through the “announced” treatment. Others inquired about the privacy of the healthcare worker visit, and whether news of the intervention spread through community members overhearing or seeing the healthcare worker. She acknowledged that this was a strong point of feedback for the study, and they had not asked the healthcare workers to record whether they noticed others unintentionally witnessing their interactions with the participant. In terms of study design, an audience member asked about the randomization of the vaccination sites, as the locations were non-randomized and chosen after discussions with community elders.

Maffioli’s research speaks to the health challenges of today, examining how an interdisciplinary approach can aid understanding of social barriers to health interventions. She commented in her conclusion that the delivery and acceptance of new vaccines should be a top global priority. A mode of vaccine delivery, such as the one explored in the study, that addresses barriers to vaccine supply and demand, and receives no backlash if provided by healthcare workers, remains a cost-effective strategy to increase vaccinations.

*