Past research has used internet search data to gain insights into the spread of infectious disease. This new study is among the first to focus on searches relevant to heart disease, including terms that may signal symptoms of a heart attack.
‘Strong correlation lies between the people’s search queries for heart disease symptoms and geographical location of the heart disease hospitalizations.’
"On an individual level, this study shows that people are not just searching for routine symptoms but symptoms that can be life-threatening," said Conor Senecal, MD, resident physician at Mayo Clinic and the study's lead author. "It suggests that what's presented to them online can have real-world impacts on their health."
The study focused on search trends and hospitalization data related to coronary heart disease, the most common type of heart disease and the leading cause of death in the U.S. Caused by the buildup of plaque in the arteries that supply blood to the heart, coronary heart disease can lead to chronic chest pain, heart attacks and other heart problems.
To track search queries, scientists analyzed how frequently people in the U.S., United Kingdom and Australia used Google to search for chest pain, chest discomfort, chest pressure or angina (a medical term for chest pain) between 2012 and 2017. State-by-state Medicare beneficiary data available from the U.S. Centers for Disease Control and Prevention was used to track coronary heart disease hospitalizations.
The study focused on two questions. The first was whether the geographic distribution of relevant search queries reflected the geographic distribution of heart disease hospitalizations. On that question, scientists found a strong correlation (0.8 on a scale from -1 to 1, with scores above 0.7 being considered a strong correlation) between the search queries and hospitalizations on a state-by-state basis. In general, this indicates that people living in states with higher rates of heart disease hospitalizations search for symptoms more frequently, scientists said.
The second question focused on whether the timing of relevant search queries reflected the seasonal trends in heart disease hospitalizations. Past research has shown that coronary heart disease hospitalizations peak in the winter months and drop in the summer. In this study, scientists found relevant search queries tracked relatively well with seasonal trends, with significant spikes in the winter and dips in the summer in the U.S. and U.K. In Australia, search queries peaked in the Southern Hemisphere autumn and dipped in the spring, aligning with previous research on Australia's seasonal heart disease trends.
According to scientists, the results show that the use of the internet to investigate heart symptoms is widespread across many regions and suggest that search trends can be a useful resource for tracking heart disease over time and in different regions.
"It's very expensive and burdensomei to do well-done epidemiological studies," Senecal said. "Search data is an imperfect tool, but it may be able to provide some hints or clues into issues that could be further researched with more rigorous methods. It basically provides a cheap and easy way to take a first look at interesting questions."
The study did not investigate whether internet searches led more people to visit the hospital after experiencing heart symptoms.
"It's crucial moving forward that large search providers and public health officials work together to provide accurate information for these conditions through internet-based search," Senecal said. "Patients should seek out reputable resources and, if they have any concerns, contact a physician."
Senecal will present the study, "Temporal and Geographic Correlation between Coronary Artery Disease Prevalence and Search Engine Queries," on Saturday, March 10 at 3:45 p.m. ET in Poster Hall A/B.