BYU researchers are using a database of 1918 influenza deaths to draw lessons from the past and make informed decisions going forward.
The university’s Family History Technology Lab teamed up with FamilySearch to create this online interactive database. Researchers used machine learning (the same technology used in self-driving cars) to read death certificates.
From the death certificates, they learn a person’s cause of death and demographic information. The information from these certificates is used to create profiles of the deceased individuals, which are linked to FamilySearch.
Lab co-manager and economics professor Joe Price said the 1918 database includes people who died during that year from influenza in 10 states, with seven states more in progress. The database is available on the lab’s website.
Fellow lab manager and computer science professor Mark Clement said he believes this research can help answer questions about current and future pandemics. Clement helped develop the handwriting recognition software to read the cause of death from the death certificates.
The lab hopes to provide additional data from past pandemics to use with COVID-19 data as more research continues. Clement and Price said they believe this data will help to prepare for future pandemics and show what the consequences are for not taking preventative measures.
Though the database is not yet complete, Price and his fellow researchers are already finding ways to use the data to understand how shutdowns affect future deaths.
Price is working on publishing a paper with BYU student Carver Coleman and University of Notre Dame professor Kasey Buckles, comparing and contrasting the influenza deaths of cities in two states during the 1918 pandemic. The research is focused on what happened in cities that had shutdowns and those that didn’t.
Price and Coleman found in the four cities that decided not to do anything, the death rate was about twice as high as other comparable cities. Price said he hopes this research will help people understand that doing nothing is a costly option.
This project is also unique because its data is tied to individual people. Coleman said previous research done with 1918 pandemic deaths used monthly data from the Centers for Disease Control and Prevention and other sources. Now the lab has access to specific daily death rate data. Researchers can also access demographical data such as a person’s race, age and even occupation.
The individual profiles containing this data are stored on the FamilySearch website but are also accessible from the database. Profiles contain a life summary, a list of family members, a family timeline, photos, memories and activities to teach about their heritage and traditional dress.
Price said this information will be key in another paper focusing on survivors’ families. “We’re looking at what the long-run effects might be for the family members, both in terms of the direct effect of having a family member die, but also if there’s something unique about having family members die during a pandemic.”
Price and Coleman said they want to continue expanding the database for years to come. While right now the database only includes people who died in the year 1918, Coleman said he hopes the database will one day reach 100 years of deaths. This will lead to more research opportunities in other causes of death, such as pollution or natural disasters, he said.
The research the team has done and hopes to do in the future relies heavily on the digital records provided by FamilySearch, Coleman said. “It wouldn’t have been possible without all the work that FamilySearch put into digitizing all the death certificates.”
The lab’s research dreams do not stop there, as Price hopes to soon expand the database into other countries such as South Africa and Brazil.
Price said he knows there will be obstacles since there are places in the world that don’t have high-quality death certificates. However, most deaths were recorded in some way, either on gravestones, newspapers or in personal handwritten records.
“Right now the world is so full of so many records that we can’t search because they’re handwritten. The moment we can unlock handwritten documents, we can make everything in the world searchable,” Price said.