The impact analytics can have on COVID-19

The effective use of data analytics can not only help to contain the spread of the disease, but also ensure medical supplies are sent to the right place and make contact tracing that much simpler.

Johannesburg, 01 Sep 2020
Read time 4min 30sec

Widely recognised as a solution that delivers problem-solving and predictive prowess, analytics today should also be considered an essential navigational tool in the uncharted waters of the current pandemic. After all, predictive data models are essential tools for ensuring the next potential hotspot of infection has the necessary number of beds, medical staff, personal protective equipment and ventilators to hand.

More than merely predicting the numbers outlined above, data analytics can also project the impact of policy measures – such as stay-at-home orders, the closure of non-essential businesses, or mask mandates – on attempts to flatten the curve.

According to Steve Bennett, director, public sector and financial services practice at SAS, who was speaking at SAS Global Forum 2020, there are a multitude of examples indicating how analytics can assist government and healthcare providers around the world with COVID-19. For one thing, the implementation of analytics will assist public health entities to make better decisions that will help to slow the spread of the disease, as well as speed up the economic recovery.

“Moreover, analytics can assist at both ends of the spectrum. On one hand, it can assist with tracking and analysing the movement of large numbers of people at the national level, while at the same time helping to improve contact tracing at the local level, to help limit disease spread in communities,” he says.

“At the macro level, the easiest way to understand the movement of large groups is by analysing cell and tower data in order to track not only how people are moving, but how these movements change with the implementation of different policies. It is important to note that the data we are talking about here has nothing to do with geolocating individuals, but is rather aggregated data designed to enable population scale analysis, which will support high-level decisions.”

In the end, says Bennett, there are three key sets of decisions that this sort of information can help government or public healthcare to make. The first is around determining where to implement the relevant social containment policies, and then measuring the effect of these.

“The second area analytics can assist with is in understanding the geographic spread of the disease in order to ensure that medical resources are optimised in the right areas. Finally, this data can also play a role in helping to minimise the impact of COVID-19 on the economy, by understanding when and where different segments of the economy can be safely opened up again from restrictions.

“While the data is high level, authorities can still drill down to specific areas in order to determine who is moving where and how often they are doing so and which transport routes they are using. This is necessary in order to take knowledgeable decisions about how to prevent the spread of the disease – it goes without saying that clearly understanding how the public moves can be vital in mitigating the impacts of the pandemic.”

Furthermore, these analytics can also assist in predicting where the disease might go next, thanks to future forecasting undertaken by machine learning that has been trained on existing cases. This is ideal for providing government health organisations with a broad view of what regions are most at risk and allows for better policy implementation, while also ensuring that medical care and supplies are delivered to where these are most needed.

Then, at the other end of the spectrum, he continues, contact tracing is vital to track the spread of infected or potentially infected people. Contact tracing is still generally done the same way as it was a century ago, suggests Bennett, who adds that this simply isn’t good enough for the challenges posed by a fast-moving disease.

“With analytics, we are able to update the manner in which this is done, using a simple, mobile system that enables the user to, for example, not only search for Patient X, but when you have their information, to also call up all relevant data captured in a consistent and understandable manner. In addition, it is easy to add new contacts and update this in a master record in the field, in real-time. This solution means authorities can develop a wide range of connective information around Patient X that can be further enriched over time, with data like travel, transportation and location information.

“The real difference – particularly when combating a highly contagious disease – is that whereas making these connections from the traditional contact tracing method could take days or weeks, here it only requires a few clicks to see potential transmission tracks and to understand other drivers of risk. Essentially, what analytics offers here is much quicker access to much deeper levels of understanding. And in a pandemic, speed and knowledge saves lives,” he concludes.