Fighting coronavirus: 4 ways analytics is making a difference
Jeff Alford, SAS Insights Editor
It's a pitched battle right now. Us, humanity, against a virus that has spread (and continues to spread) at a rate that many of us find difficult to comprehend.
The new coronavirus and the disease it causes (COVID-19) have affected nearly every societal activity around the planet. Critical supply disruptions, job and financial insecurities, and shortages in medical supplies, food and essential services are forcing us to adapt. The virus has separated us from each other and from the embrace of our cultural and religious communities, where many are accustomed to finding great comfort.
Unfortunately, isolation is part of the solution.
Luckily, some of the smartest and most experienced science and medical teams around the world are tackling the daunting work of understanding the mechanisms behind the spread of this pandemic. Analytics is providing new insights based on massive amounts of data to stem the uptick in new cases and help meet society’s needs.
Here are four ways people are using analytics to improve responses to the coronavirus outbreak.
Using data visualization to track the coronavirus outbreak
Being able to visualize the spread of the virus can help raise awareness, reveal its impact, and ultimately assist in prevention efforts. Learn more in our report that depicts the status, location, spread and trend analysis of the coronavirus.
Medical resource optimization
For years, even decades, many health policy experts advocated a lean health care system that discouraged unused beds and underused facilities. The question of the day is: How can health care systems rapidly expand capacity when faced with a pandemic? And since resources are in short supply, what critical assets are needed in each location?
In anticipation of an onslaught of patients, hospitals around the globe cleared as many beds as possible. Canceling elective surgeries was a relatively easy step. But the vital and much more complex task is determining where and when beds will be needed as COVID-19 moves through various populations.
By using analytical methods to more efficiently allocate available bed space and hospital resources, even to neighboring health systems, health officials hope to reduce mortality rates in areas hardest hit by the virus. These analytical methods include:
- Applying and refining epidemiological models to project COVID-19 infections within a region.
- Predicting the potential numbers of infected people that will require medical intervention and what the level of care might look like.
- Forecasting the number of required caregivers based on scenario modeling.
The insights generated by analytics could lead health care systems to take several actions, or a combination of them:
- Repurposing beds for a higher level of care.
- Using day surgery/outpatient centers as new bed space.
- Exploring the possibility of reopening shuttered facilities.
- Reallocating clinicians, vital equipment and supplies to where they’re needed most.
In an industry where human lives hang in the balance every day, there is arguably no more important time than now to make use of all available resources.
Ensuring demand planning stability
Demand forecasts for everything are topsy-turvy at present – from hospital patient flows to goods and services. What can data and analytics teams do?
Improving existing forecasts of ongoing or expected COVID-19 disruptions with a broad range of forecasting methods – including data patterns as the virus moves across regions – may help to improve short- and midterm forecasts.
At a recent Institute for Business Forecasting virtual town hall, SAS forecasting expert Mike Gilliland noted some of the key points made by his fellow forecasters:
- Macro forecasting is hard right now, but micro forecasting (down to the product level) is even harder.
- There are many interventions going on beyond what is normal, such as government stimulus and falling oil prices, all adding to the uncertainty and complexity.
- Distinguish the passive work of observing and collecting data from the proactive work of driving demand away from shortages and toward substitute (available) products.
- Try to use additional streams of external data, not just your own internal data.
With social network analytics, contact tracing investigations for public health become much easier. With available data, 'epidemic detectives' can begin building networks of contacts. The more data, the better the chances of identifying points of vulnerability.
Contact tracing investigations for public health
Contact lists can grow quickly, depending on the rate of infection and environmental factors that may encourage its spread. We've all seen how seasonal illnesses (like colds and flu) can spread in a close-contact environment, like a classroom or a city bus. And, of course, the more social interactions, the more work is involved to identify all those on the contact list.
One of the downsides of contact tracing is that it relies on the memory of interviewees to identify where they've been and who they've come in contact with. Few people can accurately retrace their steps and actions over the course of a typical two to three day period.
With social network analytics, contact tracing investigations for public health become much easier. With available data, "epidemic detectives" can begin building networks of contacts. The more data, the better the chances of identifying points of vulnerability. Then, using data visualization techniques, health officials can begin identifying common contact points and building a robust, reliable picture of an infected person’s social interactions.
We are creating an entire methodology to evaluate people’s movements over time and correlate them to the spread of the virus. By knowing key locations based on their role in the network (that is, how central to speed the spread of the virus, how well-positioned to control the spread flow, or how well-connected to other locations, etc.) we can help government agencies in defining social distancing measures more accurately, according to SAS Principal Data Scientist Carlos Pinheiro.
Network analysis outcomes give us a better understanding of people’s flow throughout locations and how specific lockdowns can substantially affect the spread. It’s not just locations with a high number of positives that affect the risk, but also key network locations that regulate the flow of people – and thus the spread of the virus.
Analytics and AI can find answers and recommend ways to handle this pandemic – faster and with more accuracy.
Situational awareness and critical response analysis
In the early days of the coronavirus outbreak, SAS began to use the fast-flowing current of data initially from China, then from around the globe, to visualize the spread and effects of COVID-19 on the world's population.
The result is our evolving coronavirus dashboard. The dashboard's utility lies in its ability to quickly access and organize data in easy-to-understand data visualizations to improve situational awareness. Organizations in many different industries are using it to explore the relationship between outbreak hot spots and their business, including critical supply chains, distributed workforces and emergency response preparedness.
“In the early days and weeks of any widespread global health concern, particularly in a fast-moving outbreak like the coronavirus, there are many unknowns,” says Mark Lambrecht, Director of the Global Health and Life Sciences Practice at SAS. “Data visualization can be a good starting point to understand trends and piece data points together into a meaningful story. The ability to visualize the spread of the virus can help raise awareness, understand its impact and, ultimately, assist in prevention efforts.”
SAS experts are ready to help you get a dashboard up and running fast. Contact us using this short online form.
Recommended reading
- Operationalizing analytics: 4 ways banks are conquering the infamous ‘last mile’Here are four examples across the banking industry that show how these leading organizations followed a clearly defined path to conquer the infamous 'last mile' of analytics.
- ModelOps: How to operationalize the model life cycleModelOps is where analytical models are cycled from the data science team to the IT production team in a regular cadence of deployment and updates. In the race to realizing value from AI models, it’s a winning ingredient that only a few companies are using.
- The untapped potential in unstructured textText is the largest human-generated data source. It grows every day as we post on social media, interact with chatbots and digital assistants, send emails, conduct business online, generate reports and essentially document our daily thoughts and activities using computers and mobile devices.
Ready to subscribe to Insights now?