Google searches for COVID-19 symptoms can anticipate case numbers

Google searches for COVID-19 symptoms can anticipate case numbers

By Dr. Ramya Dwivedi, Ph.D.Nov 16 2020

The ongoing coronavirus 2019 (COVID-19) pandemic continues to disrupt healthcare organizations, economies and societies worldwide. COVID-19, caused by severe respiratory syndrome coronavirus 2 (SARS-CoV-2), has claimed over 1.322 million lives in less than a year. Herd immunity is far from a reality and vaccine candidates awaiting approval are still not on the market.

Efficient and durable immunization will pose unprecedented challenges when it is available for billions of people across the world. Until then, we need to continue along the current path by utilizing non-pharmaceutical interventions (NPIs), like social distancing measures, mask-wearing and lockdowns, for the foreseeable – with adapted lifestyles and with useful predictable outcomes.

Curiously, a research group has found that Google searches may come to the aid in predicting the evolution of the COVID-19 pandemic, which we need to identify the scale of, manage, and, if possible, contain new outbreaks.

Two common symptoms – the loss of taste and loss of smell – are triggered by SARS-CoV-2 infection. Over 50% of SARS-CoV-2 infected people report both of these olfactory and gustatory dysfunctions. It is important to note that these dysfunctions are present over 10-fold in COVID-19 patients than in other acute respiratory infections. This implies that the onset of these two symptoms should raise a clinical red flag for SARS-CoV-2 infection. This will possibly help in accelerated and accurate diagnosis and subsequent isolation and treatment if required.

The group's study, published as a medRxiv* preprint article, uses Google trends to predict the epidemiological trajectory of COVID-19 in the United States for the search terms: taste loss and smell loss.

Study: Google Searches for Taste and Smell Loss Anticipate Covid-19 Epidemiology. Image Credit: Haysekiz / Shutterstock

The study finds a significant association between Google searches for these two symptoms and the concomitant number of new weekly SARS-CoV-2 infections reported during the same week. The authors find the correlation improved over time.

Real-time analysis of web searches of specific symptoms that characterize a particular pathology helps anticipate or identify the epidemiological trajectory of that disease within a specific geographic setting. It may help predict or anticipate SARS-CoV-2 infections in the cohort population or apprehend any resurgence within a community. This also represents a unique opportunity to monitor disease epidemiology.

Related Stories

  • Identifying SARS-CoV-2 spike protein mutations resistant to antibodies
  • Will an effective COVID-19 vaccine return life to normal in the US?
  • 16 new lineages of SARS-CoV-2 emerge in South Africa despite lockdown

The authors performed an electronic search in Google Trends (Google Inc. Mountain View, CA, US), between January 20 and November 1, 2020. They retrieved the number of new weekly cases of COVID-19 in the US from the official website of the Centre for Disease Control (CDC). The authors analyzed the data with Spearman’s correlation and used receiver operating characteristics (ROC) curves for diagnostic performance.

The authors found a nearly perfect correlation between the Google Trends scores of taste loss and smell loss.

They found the highest correlation for the search terms at the index week and the number of new weekly SARS-CoV-2 infections reported in the US two weeks later.

It is important to note that an even stronger correlation was found when the sum of the Google Trends scores for search terms at the index week was compared with newly diagnosed weekly SARS-CoV-2 infections two weeks afterwards.

The application of this study is extremely important in a low resource setting, where widespread diagnostic molecular testing is not possible. For example, in the Northern Hemisphere, as it heads into the winter months, concomitant with peak cold and flu season, it is useful to be able to distinguish and identify the SARS-CoV-2 infection beforehand.

While common flu symptoms may not prove helpful in predicting the COVID-19 evolution with the onset of winter, the onset of olfactory and gustatory dysfunctions, especially in combination, is rare in patients – it has a high likelihood of an ongoing or recent SARS-CoV-2 infection.

This is a new branch of epidemiology – “infodemiology,” defined as a science-based approach using electronic Web-based information to inform and guide public health policies. It is increasingly being used to support conventional epidemiology in preventing, containing and managing COVID-19 around the world.

This study suggests that the volume of Google searches for olfactory and gustatory dysfunctions may help predict the epidemiological trajectory of COVID-19 before changes in official reporting. A positive variation in the number of Google searches for taste or smell loss is characterized in this study for predicting an increasing trend of new SARS-CoV-2 weekly infections two weeks after the date of Google searches.

The authors strongly recommended that the health organizations and governments consider using the number of Google searches as an adjunct tool for planning future strategies to prevent, contain, and manage outbreaks of the ongoing COVID-19 pandemic.

*Important Notice

medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.


Leave a Reply

Your email address will not be published. Required fields are marked *