In the mid-1800s, London witnessed a cholera outbreak that killed more than 10,000 people. Local authorities were puzzled over the source of the disease. It wasn’t until Dr John Snow mapped the cases geographically that they traced the root of the disease to a contaminated water pump. The pump was shut down, and cases trickled to a stop.
When you’re up against a global pandemic, you need all the information you can get. Data analytics has the potential to help public agencies and medical teams understand how a disease spreads, and how to limit its impact. “It’s about being better informed and being able to inform better,” says Remco den Heijer, Vice President of ASEAN at analytics company SAS.
GovInsider spoke to their team of analytics experts to learn how data can help combat global pandemics – by helping public officials and doctors understand its spread, closing the gaps in contact tracing, and sieving out fake news.
Tracking disease spread with social media
Social media and news sources share new information on COVID-19 every day. Having a data dashboard to pull all this data together can help authorities better understand how the disease is spreading.
One factor to look at would be an increase in Google searches of related symptoms. When there is a spike in searches for symptoms such as high fever and coughing, it can “give a very strong clue” that a disease might be spreading in a particular area. “The healthcare authorities should be able to reach out quickly to give them free clinics and treatments,” says Jason Loh, Head of Analytics and Artificial Intelligence in Asia Pacific, Global Tech Practice at SAS.
Online news can also help to trace the spread of the disease. In China, authorities are combing through news reports of confirmed cases from media outlets all across the country. The details from these reports help medical teams understand how and how quickly the disease spreads. This may lead to more targeted treatments, says Den Heijer. “Instead of a generalised approach to treatments, you make treatments based on what you know what will or will not work,” he explains.
A good quarantine system is crucial when tackling highly contagious diseases. Data processing tools can help close the loops in contact tracing and make it more efficient.
Traditionally, “contact tracing is a very manual process”, explains Loh. After interviewing an at-risk person, police officers have to draw out a map of his or her movements and identify who he or she would have met.
This tedious process can be shortened with an automated system that combs through the interview transcript. SAS’s software picks out the people an at-risk person could have come in contact with. “When I have 500 officers interviewing some 10,000 people, this can be developed quickly so they can contact trace faster in order to stop the spread,” says Loh.
Understanding fake news
Mass panic has made dealing with the spread of COVID-19 more challenging. It’s crucial for governments to understand what people are worried about, and address those concerns. “When you understand the ground sentiment, you can tackle the situation better,” says Den Heijer.
Analytics are vital for refining epidemiological models, and helping national healthcare systems respond. SAS has helped predict different curves in infection rates, ensuring that different methods from lockdowns to masks can be deployed to ensure that hospitals can cope.
This data is vital for supply chains, particularly when managing limited resources like ICU beds, ventilators and personal protective equipment. Is a new emergency quarantine facility required? Does a nation need to boost its PPE manufacturing capacity? Analytics predicts this requirement in advance.
Good healthcare management requires good people. Analytics can forecast the potential supply of caregivers, using multiple datasets including retired staff availability, shift patterns, and regional resourcing contrasted with infection numbers in case nations need to redeploy staff.
Not all healthcare takes place in hospitals, however. Telehealth plays a key part, especially when combined with symptom checking tools. SAS Analytics can automate the workflow for these tasks, ensuring available staff to assist, and data to demonstrate risk areas and the potential for a future spike. This is even more powerful when combined with data on population density to identify possible clusters or locations that require a lockdown.
SAS’s analytics software is used by more than 3000 healthcare sites across nearly 50 countries. It helps doctors detect conditions early and provide better treatments.
With Covid-19, data is particularly crucial, and we are already beginning to see the impact it can have. “Today, the world knows a lot more about Covid-19 than SARS,” says Loh.