How Stanford scholars collaborated with county health officials and a local community to allocate COVID-19 resources, address health disparities

The article can be found here.

Date Published: August 27, 2021

Summary of Article

During the COVID-19 pandemic, allocating limited supplies has been a real challenge, especially in underserved, minority communities that have been affected disproportionately by COVID-19. Stanford researchers helped public health officials in Santa Clara County by using machine learning to prototype a new way to distribute COVID-19 diagnostic tests that were important to better understanding the transmission of the virus. The researchers’ findings showed that machine learning and collaboration with community health workers broadened testing capacity, decreased demographic disparities in testing and caught clusters of infections early on. Daniel Ho, senior author of the study, and his team began working with public health officials to investigate how they could assist with their pandemic response. Dr. Analilia Garcia, the racial and health equity senior manager for the Santa Clara County public health department, was concerned by how COVID-19 was disproportionately impacting the county’s Latinx residents. While this demographic makes up just over one quarter (25.8 percent) of the county’s population, they have accounted for more than half of all its COVID-19 cases (50.3 percent). Garcia believed to combat this problem, programs that provided at-home testing to residents were necessary. Ho and the RegLab team helped Garcia and officials determine where testers should go by using simple insight from machine learning called “uncertainty sampling”: Go where there is the most uncertainty about COVID-19 transmission. The team used upper confidence bound sampling to determine where the data is most unclear about the highest positivity rates. This approach allowed the team to navigate “the explore-exploit tradeoff”: How to allocate resources that can be used to explore risk (unknown outbreak areas) and exploit known risk (known outbreak areas) that would guide health workers to where they were needed most. With this model, Ho and his team created detailed maps and assigned community health workers to specific neighborhoods in real-time based on daily intake of cases. To deliver door-to-door tests, Spanish-speaking community healthcare workers were recruited who were able to serve as trusted advocates between Latinx residents and the healthcare system. Healthcare workers expanded testing resources in East San Jose neighborhoods between 60-90 percent relative to the baseline over two months. The results of this collaboration demonstrate what can happen when academia, local government, and the community come together to tackle a problem.

Author Information

Daniel E. Ho

Daniel E. Ho is the senior author of the study and the William Benjamin Scott and Luna M. Professor of Law at Stanford Law School. He is a Professor of Political Science and Senior Fellow at Stanford Institute for Economic Policy Research. Ho is the Associate Director for Stanford Institute for Human-Centered Artificial Intelligence as well as the Director for the Regulation, Evaluation, and Governance Lab (RegLab). He is the founder of the RegLab, where he and other Stanford researchers work with government agencies to adapt and leverage advances in data science and machine learning for public policy. Other authors on the paper include first authors Ben Chugg, Lisa Lu, and Derek Ouyang, all affiliated with Stanford. Benjamin Anderson and Raymond Ha, from Stanford’s RegLab, contributed. Alexis D’Agostino, Anandi Sujeer, Sarah L. Rudman, and Analilia Garcia from the county of Santa Clara Public Health Department are also co-authors on the paper.

Areas of Application

The main areas of application from this article are health and wellness, public health and policy, and racial equity. The work Ho and his team did directly impacted the Santa Clara County community. Public officials recognized a problem and utilized data science practices to find a solution. The research utilized health data and implemented practices that benefited the community. Ho and his team worked closely with public health officials, impacting public health decisions for the community. Because the Latinx community was disproportionately impacted by COVID-19 in these areas, the work Ho and his team did helped with accessibility and equity for this vulnerable population. This article demonstrates how data science can positively impact public policy and the well-being of individuals, including health and racial equity.

Reflection

I found this article to be very informative and insightful. As someone who is interested in working in the healthcare field as a data scientist, reading about how data science is making a positive impact on the well-being and health of individuals is very inspiring to me. I recall hearing about how minority groups were disproportionately affected by the COVID-19 pandemic and was interested to see what government and public health officials were doing to combat this issue. Santa Clara County is a good example of how officials recognized an issue and worked to find a solution. I think it is very important for health officials to utilize all resources and information available to them, including meaningful data insights and practices.

Similar Articles

The first article describes how Google supported organizations around the world to aid in the COVID-19 response. Google recognized how critical it was to monitor and forecast disease spread, improve health equity, slow transmission, and support healthcare workers. Google supported the development of data platforms to help model disease. This article addresses some of the issues that were addressed in the original article about Santa Clara County, such as disease modeling and improving health equity. Both articles demonstrate how data and machine learning are essential tools in combating the COVID-19 pandemic. The second article describing how to fight COVID-19 with machine learning further emphasized the importance of machine learning in combating COVID-19 and its impacts on the world. This article mentions how machine learning is helping to screen patients and diagnose COVID-19, something that the original article focused on in their research. The third article is very relevant to the original article as it discusses the impact of COVID-19 on Latinx communities. Latinx people account for 18% of the U.S. population, but make up 33% of the nation’s COVID-19 cases. The AMA Center for Health Equity’s report identified several drivers of inequity including structural drivers such as anti-immigration policies and rhetoric and restrictive health care insurance access as well as social determinants such as lack of Spanish-language COVID-19 resources and increased police and immigration enforcement in Latinx communities amid the pandemic. Many people in the Latinx community are reluctant to go to the doctor because of distrust, and there are not very many Latinx physicians to provide culturally responsive care and education. This article by the AMA is very relevant to the original as it addresses some of the reasons why the Latinx community was disproportionately affected by COVID-19 in Santa Clara County. It is important that data science is used for the greater good including racial equity and health equity, like the work of Ho and his team.

Summary Table for Beaches Dataset

library(readr)
beaches <- read_csv("~/Desktop/DS3001/DS-3001/data/beaches.csv")
DT::datatable(beaches)

Plot for Beaches Dataset

beaches <- na.omit(beaches)
totals <- beaches %>% 
  group_by(season_name) %>% 
  summarize(rainfall = sum(rainfall))


beaches_plot <- ggplot(totals, aes(x=season_name,y=rainfall)) + geom_bar(stat='identity', position='dodge') + labs(x='Season Name', y='Rainfall Total', title = 'Rainfall Totals for Each Season')

ggplotly(beaches_plot)