Despite continued efforts to improve health systems worldwide, emerging epidemics remain a major public health concern. Effective response to such outbreaks relies on timely intervention, ideally informed by all available sources of data. 

The increasing complexity of outbreak data has led to the rise of outbreak analytics — a data science designed to inform outbreak response. 

So how does it work? There are four major goals:

• Collecting multi-source data. Here, it’s important to remember that data collection is not merely about picking some structured data from clinical documents. Raw data sources are also involved. Therefore, outbreak data analytics requires noise removal, data cleaning, and normalization to enable the data to generate actionable insights. 
• Providing situational awareness. In the pandemic, the times of uncertainty, it’s critical to know when the risk is manageable and when it’s most likely to get hazardous. Situational awareness monitored via diverse public sources helps mitigate and control this risk. 
• Shaping effective response. At this point, data visualization steps in. It presents the analytical results in a clear-cut and user-friendly form (via dashboards, charts, etc.), helping decision-makers design effective response measures without delays.
• Training and capacity building. Once a crisis is resolved, the decision-makers look back at the situation to analyze the accepted response scenarios and select successful strategies. Then they make relevant amendments to professional training and health emergency plans to get better prepared to face another health crisis. 

The COVID-19 pandemic is not the only one in the 21st century or even in the past decade. Not long ago, Ebola and Zika wreaked havoc in West Africa and the Americas. Rapid response and its adaptability, surveillance mechanisms, and management strategies were what helped researchers and clinicians back then. But are they sufficient now? Not really. 

Given the global reach of the pandemic, an effective response is hard to design without the key parties’ extensive cooperation, and it’s not only about cross-field research within outbreak analytics. With the virus tormenting over 200 countries, an adequate response strategy should rely on best practices from each region. 

Luckily, clinical researchers and medical professionals worldwide have joined their efforts in the fight against the virus. Looking for the best crisis resolution strategies, they have developed an elaborate data sharing culture. They now open-source their developments and constantly work on improving them. Thanks to this approach, we saw unique diagnostics projects, such as DarwinAI (Canada). With this computer vision tool in place, it’s possible to diagnose COVID-19 by chest radiography scans only. Before, the only medical-imaging COVID-19 diagnostic method was computer tomography (CT).

The cooperative approach includes outbreak analytics, too, which has fueled real-time and prevention analytics. These two types of analytics are at the core of containing the virus dissemination.

Real-Time Data

In a critical situation, this type of analytics drives rapid data-based decision-making and tailors the process to each situation separately. Such analytics is of special importance to local medical professionals that are on the frontline of the virus fight.

Local real-time analytics tools are developed with regard to the needs of key actors. For example, Kinetica and Disaster Tech deployed a dynamic AI-powered analytical platform to assist the U.S. crisis responders to track coronavirus-related data on the fly. The solution allows emergency services to visualize live statistical data on personal protective equipment (PPE) availability, hospital capacity, the number of test kits, equipment availability, and more to choose the most suitable location for a patient and save time.

Real-time analytics is not limited to solving operational tasks locally. Our World in Data, an Oxford-based project designed for addressing global problems, deployed a massive publication dedicated to COVID-19 statistics worldwide to power continuous real-time analytics. The source provides access to a bulk of coronavirus-related data, from new cases and mortality rate to policy response for each country living through the coronavirus crisis. All charts, reports, and other interactive data visualizations are daily updated and available for download.

Analytics for Prevention

Real-time analytics helped the South Korean government to power up prevention strategy design and Covid-positive patient surveillance. It uses the data from IoT and AI solutions underlying the live smart cities networks and personal information provided by confirmed patients. This allows researchers to track the patients’ movements, identify their contacts, and predict the potential outbreak scale in a given region with the help of big data analytics. The data is also used for drafting preventive measures and instructions.

Taiwan managed to leverage real-time analytics for timely prevention, too. Just before the Lunar New Year celebration that marks the Asian holiday season, the country integrated the national health insurance database with those for immigration and customs. When infected travelers started to arrive, a big data analytics solution combed through the integrated databases and established the connection between visitors’ travel histories and the symptoms they experienced. The solution sent real-time alerts during a hospital visit to assist with case identification. It also enabled travelers’ classification by infection risks relying on the flight origin and travel history over the last 14 days.  

In a situation of uncertainty and turmoil like the present pandemic, insightful data is the king. And it’s relevant analytical solutions that can harness it and leverage adequate response and prevention measures without delays. In this respect, outbreak analytics works well.

At the same time, the global spread of the virus brings another aspect to the spotlight — international cooperation. To develop effective management strategies, key stakeholders and decision-makers from all across the globe should review and improve the developed solutions and polish the response measures relying on real-time data. Thus, uniting efforts in all fields, from analytics to diagnostics and treatment, is the only way to stop the pandemic and get better prepared for any other health emergency. 

Yaroslav Kuflinski is an artificial intelligence/machine learning observer with Iflexion.