We call this 'biosurveillance.' It's a growing field in medicine and public health, which looks for patterns in online doctors' searches related to the signs and symptoms of a disease. Analysis of these searches allows us to be alerted when doctors in a particular region suddenly begin to see the same symptoms - symptoms that could be the early signs of a potentially deadly outbreak. The reason we can see these patterns is that many doctors around the world are using digital tools to search for clinical information and evidence. As they do, it becomes possible to identify spikes in their searches.
As a doctor and infectious disease specialist, I'm fascinated by the potential of health information technology to be used as a new weapon in public health. Although traditional techniques for detecting and monitoring outbreaks continue to be important for protecting public health, these approaches can be slow and don't always detect all cases.
I came across the possibility of using digital medical information for biosurveillance through my company, which makes one of the world's most widely used online medical references, called UpToDate. More than a million doctors in over 180 countries use UpToDate to support their decisions while caring for patients. They can look up advice on the best approach to the diagnosis and treatment of different conditions and back up their decisions with the latest evidence. Because UpToDate is so heavily consulted - with over 35 million topic views per month - we wanted to see if our users' searches could enable us to identify infectious disease outbreaks.
Based on the research we have done to date, the answer is yes. Take the spike in doctors' searches we found in 2014 and 2015 in Jeddah and Riyadh, Saudi Arabia, for information relating to MERS. MERS is a potentially deadly viral infection that emerged in the Middle East in 2012 and continues to cause outbreaks. The increase in UpToDate searches for MERS and its symptoms in cities in Saudi Arabia correlated with three different outbreaks in hospitals there. The individual doctors probably couldn't see the pattern emerge on a case-by-case basis. But stepping back and allowing the data to paint the big picture showed the pattern of the outbreaks.
We have also analysed UpToDate search activity in the setting of several infectious disease outbreaks in Arizona and found similar results. In some cases, we even found that clinicians were searching UpToDate for information about a disease before the earliest cases were reported to public health authorities.
Analysing global search activity on health topics has been done before. Google saw the potential when it developed Google Flu Trends, using patterns in search queries from the public to predict or detect influenza outbreaks. There is also a lot of interest in using content in social media apps to detect outbreaks.
However, analysing what clinicians themselves are searching for, based on their consultations with patients, is more likely to relate to actual cases (rather than other factors such as the public's fears about a possible outbreak reported in the media). And since these tools only need the clinician to have a smartphone and an internet connection, it could be possible to identify patterns suggestive of an outbreak from most places in the world. All that is needed is a focused effort to analyse the information that the tools generate.
The stakes are high. As we saw with Ebola, public health authorities have been caught off guard by outbreaks around the world. The wide use of digital information tools in medicine opens up the possibility for biosurveillance to help us detect outbreaks earlier. By doing so, public health authorities would be better able to halt such outbreaks, preventing new infections and their complications, and even saving lives.
The opinions expressed in this article are solely my own and do not reflect the views and opinions of Brigham and Women's Hospital.
-- This feed and its contents are the property of The Huffington Post UK, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.