Analyzing Web Search Helps Experts Predict and Respond to COVID-19 Hotspots

Web-based analysis demonstrates its value in predicting the spread of infectious diseases, and a new Mayo Clinic study shows the value of analyzing Google Web searches for keywords related to COVID-19. ..

A study published in Google Trends found a strong correlation between keyword searches on the Internet search engine Google Trends and the outbreak of COVID-19 in parts of the United States. Mayo Clinic Minutes.. These correlations were observed up to 16 days before the first reported cases in some states.

“Our research shows that information exists in Google Trends before it occurs. Predictive analysis uses this data to better allocate resources for testing, personal protective equipment, medications, and more. You can assign it, “says Dr. Mohamad Bydon. A neurosurgeon at the Mayo Clinic and a senior researcher at the Mayo Institute for Neuroinformatics.

“The neuroinformatics team focuses on the analysis of neurological disorders and neuroscience, but the novel Coronavirus Dr. Bydon, the lead author of the study, said: “My team and I have devoted resources to better understanding and tracking the pandemic epidemic. Looking at Google Trends data, we found that we were able to identify hotspot predictors. If you use the keyword, it will appear on the 6-week timeline. “

Several studies have pointed to the role of internet surveillance in early prediction of previous outbreaks such as H1N1 and Middle East respiratory syndrome. There are several advantages to using Internet monitoring methods over traditional methods, and the study states that the combination of the two methods is likely to be the key to effective monitoring.

The survey searched for 10 selected keywords based on their frequency of use and new patterns on the Internet and Google News at the time.

The keywords are:

• COVID symptoms
• Coronavirus symptoms
• Sore throat + shortness of breath + malaise + cough
• Coronavirus testing center
• Loss of odor
• Resol
• antibody
• mask
• Coronavirus vaccine
• COVID stimulation check

Most keywords had a moderate to strong correlation a few days before the first COVID-19 case was reported in a particular area, followed by a decrease in the correlation.

“Each of these keywords had a different degree of correlation with the number of cases,” says Dr. Bydon. “If you look at 100 keywords, the correlation with the case can be even stronger. As the pandemic progresses, people will search for a variety of new information, and so will the search terms.”

The use of web search monitoring data is important as an aid to data science teams trying to predict pandemics and new hotspots. “Information delays can lead to missed opportunities to improve readiness for outbreaks in specific locations,” says Dr. Bydon.

With traditional monitoring including extensive testing public health Reports may lag behind the incidence of infectious diseases. The need for more tests and faster and more accurate tests is paramount. Delayed or incomplete reporting of results can lead to inaccuracies when data is published and public health decisions are made.

“Waiting for hotspots to appear in news media is too late to respond effectively,” says Dr. Bydon. “From a national readiness perspective, this is a great way to help understand where future hotspots will appear.”

The Mayo Clinic recently introduced an interactive COVID-19 tracking tool that reports up-to-date data for all 50 states and all counties in Washington, DC, along with insights on how to assess risk and plan accordingly. “By adding variables such as Google Trends data from Dr. Bydon’s team and other leading indicators, we have significantly improved our ability to predict cases spikes, stagnations, and declines across the country’s regions,” Mayo said. Dr. Henry Tin of the Clinic said. Chief Value Officer.

Dr. Ting worked with Mayo Clinic data scientists to develop content sources, validate information, and correlate tracking tool expertise at the Mayo Clinic COVID-19 Resource Center at mayoclinic.org.

The study was conducted in collaboration with Robert D. of the Mayo Clinic and Patricia E. Khan’s Center for Health Care Provision Science. The author has not reported a conflict of interest.