Data takes centre stage in virus-ravaged world

Line graphs and histograms have replaced memes and the pandemic has prompted people to talk the language of data scientists, rattling off statistics and epidemiological terms with ease

The world, as we knew it until 2019, does not exist anymore. In the last six months, every aspect of our life has changed. And along with the international health crisis and economic disruption it has brought, the coronavirus disease (Covid-19) has also taken over the conversations of our daily lives.

Perhaps one of the most prominent changes it has brought is a newfound obsession with data where line graphs and histograms have replaced memes and other forwards in our WhatsApp conversations and Twitter feeds.

People of all ages and from all professions now talk the language of data scientists, and everyone is mouthing terms such as reproductive rates, infection rate, mortality rate, doubling rate, flattening the curve, and exponential growth curve.

We take a look at some of the most important data concepts that have become common conversation topics since the spread of the pandemic.

FLATTENING OF THE CURVE

It is perhaps the most commonly used data term in conversations on the coronavirus disease and frequently mentioned in government briefings and media headlines and articles. It refers to measures that are meant to isolate people across a region in order to limit the daily or new infections of any disease to a level that can be handled by the local health care systems. The concept suggests slowing down the spread of the virus by introducing non-medical intervention measures like shutting down schools or introducing social distancing norms to ensure that fewer people need hospitalisation at any given point to time, and thus minimise the number of deaths.

The curve referred to is the number of active cases (determined by subtracting the deaths and recovered patients from the total infections). This is then overlaid with the health care region’s capacity to show whether the health care system is overburdened (see chart above).

One of the most widely accepted predictions based on this concept was prepared by a group of 30 epidemiologists from Imperial College in London, who ran simulation models to get an idea of what impact different forms of non-medical interventions would have on the spread of the disease. When they ran the model on the US and the UK, they found that the two countries were looking at around 2.2 million and half-a-million deaths, respectively, in the absence of any government intervention.

One of the strategies studied by the model was mitigation, wherein people who are suspected to have been infected or exposed to Sars-CoV-2 would be quarantined. However, they saw that even then, hundreds of thousands of deaths were likely. So they suggested a more extreme strategy termed suppression, under which everyone would be required to remain socially distant, which they said would save millions of lives. This meant shutting down all spots of public interaction such as schools, universities, and markets and quarantine of anyone who tests positive, along with friends and families.

DOUBLING RATE

The doubling rate is one of the Covid-19 data terms that has been introduced in the daily lexicon because it is so commonly used by the both the Central and state governments in daily media briefings. Loosely explained, it is the duration in which any number — for instance the fatality rate — doubles and is generally used to calculate exponential growth such as population rates, inflation and compound interest.

Although a precise calculation is complex, an approximate doubling rate can be calculated by using the “rule of 72,” wherein you divide 72 by the growth percentage and the result is an estimated doubling rate in periods.

In terms of the coronavirus pandemic, the doubling rate refers to the number of days it would take for any region to double its case or death count. So if there were 1,000 cases in a region on a particular day and the doubling rate at the time was five days, it would mean that the cases would touch 2,000 five days later (average daily increase of 14.9%).

On March 25, when the lockdown started, the doubling time cases in India based on the previous week’s rate of new cases was 3.4 days. Since then, it has increased to 18.3 days on June 13. One shortcoming of comparing the doubling rates is that it may not truly reflect the volume of cases that are looming. For instance, when the doubling rate was 3.4 days on March 25 (see chart), we were adding around 150 cases a day. While the doubling rate is more than five times to 18.3 on June 13, the caseload now is now more than 320,000, which means that we are adding 12,500 cases every day.

COMPARING OUTBREAKS

To account for the exponential nature of a pandemic in which cases multiply, instead of increasing in a linear manner, data analysts look at charts in a “logarithmic scale,” which makes changes in rates of case growth more prominent. This provides an avenue to compare rates of change between regions with vastly different caseloads.

A key characteristic of the chart is that the Y-axis (vertical axis) is plotted in a logarithmic scale which enables us to display data over a very wide range in a compact way.

In order to keep comparisons uniform, this chart takes Day 1 as the first day more than 500 cases were reported (see chart). Thus, we can compare countries going through advanced stages (such as the US) to countries that started reporting cases far later (India, Brazil) even though they may have a vast difference in the number of cases or deaths.

HAMMER AND DANCE

This analysis was made famous by Silicon Valley businessman Tomas Pueyo, who argued that the world needs to crackdown hard in a strategy that can be likened to a “hammer and dance” approach – with a first intense “hammer” move to stop the outbreak from growing followed by a sustained “dance” of lighter measures to contain the virus until a vaccine is found.

The analysis, posted on blogging website Medium in March, takes into account several projections and infection trends that show how the spread of the disease can be interrupted by containment efforts such as social distancing. If governments fight hard at the start, they will curb deaths and relieve hospitals.

During the Hammer phase, the goal is to get R number (the reproduction number) as close to zero, as fast as possible, he wrote. The idea is to fight the spread of the disease with the most intense of measures right when it starts spreading so once it moves to the Dance phase, restrictions can be removed and applied as needed in order to keep R number as close to 1 as possible, wrote Pueyo.