Information from Google searches as well as COVID-19 data on vaccination and previous hospitalizations can help create a real-time, predictive calculator of future hospitalizations, according to a June 11 study published in Nature.
The research team looked at Google search terms that related to COVID-19 on a state level, such as symptoms or diagnoses. They combined these terms with vaccination rates, past COVID-19 cases and hospitalizations to produce predictions on a state and national level.
They produced a retrospective prediction between January and December 2021 that accurately captured trends, both at a national and state level. The model was also able to overcome delays by using the COVID-19 Google searches as indication of mass cases before a surge.
"The state-level real-time hospitalizations predictions made by [the prediction tool] could help local public health officials make timely allocation decisions of healthcare resources, such as ventilator, ICU beds, personal protective equipment, personnel, etc., as well as account and promptly prepare for future surges of COVID-19 pandemics caused by new virus variants," wrote the authors.