Coronavirus | Opeds and editorials

Viral outrage only spikes the data

As the number of cases of COVID-19 in Karnataka and Telangana begins to surge , there is a corresponding urge to affix blame to a cause. This has happened earlier too when there was a rise in cases in other States — evidence that governments had failed and had shielded numbers finally appeared to be here. But the knee-jerk response to bad numbers is actually causing great harm.

Affects lesson learning

When we react with outrage to the rising numbers alone, we get no closer to finding out or fixing what is wrong; if we were to react with horror to bad processes instead, we would have some hope of mitigating the worst of the crisis.

COVID-19 | Interactive map of confirmed coronavirus cases in India

This is particularly important in the context of a disease that, in most of the country, can only be identified when a doctor prescribes a test based on limited criteria; reported new cases are a direct consequence of the decision and ability of State and city administrations to expand the parameters of testing, and numbers need to be viewed in that context. Roughly 5,000 COVID-19 cases are what separate Telangana from Andhra Pradesh, but it would be a mistake to see this as an indication that the two States are in a similar situation. Telangana discovered these 34,000 cases by testing just 1.7 lakh tests, while Andhra Pradesh discovered its 30,000 cases by conducting 11.5 lakh tests. Then there is the question of what type of test is being conducted. Delhi now conducts more than twice as many tests each day than it did a month only to discover fewer cases. But well over half of its tests are the far less sensitive and reliable antigen tests.

What all of this means is that the number of new cases reported each day in a city or State is not a purely mathematical fact. It is a combination of not just the city’s disease environment but also its political economy. Reacting to its numbers without accounting for this will lead us to the wrong lessons. We have seen this tendency to view higher numbers as a lack of governance than a success of reporting, and it has had similarly deleterious consequences. The publication of India’s annual crime statistics is met every year with headlines about relatively developed States having among the highest rates of reported crime, particularly against women. But until we reach the point of “full registration” — which means a country that is confident it is capturing the full extent of crime — more reported crime is for the most part a good thing. Jacob Punnoose, former Director General of Police of Kerala, put it this way to this writer: “If the police of a State [are] able to significantly increase the registration of crimes in which grievous physical injury does not take place, molestation, for instance, or overspeeding, I will congratulate that police commissioner.”

Instead, the outrage following “bad” new numbers is already creating perverse incentives against honest reporting. While investigating the working of committees set up by State governments to audit COVID-19 deaths, this writer found that not all deaths of those who were COVID-19 positive were being recorded as COVID-19 deaths. What was more worrying was that Chennai, New Delhi and Mumbai had recorded zero “suspected or probable” COVID-19 deaths — deaths of people who presented like COVID-19 patients but had not had a positive test prior to death — despite clear guidelines by the World Health Organization and the Indian Council of Medical Research. This means that such cities could be missing 20-30% or more of true COVID-19 deaths, according to Dr. Prabhat Jha, professor at the University of Toronto and an expert on mortality who led the ‘Million Death Study in India’.

Data | With COVID-19 cases rising steadily, no bend in sight for India's 'new infections' curve

In each of these three cities, health administration officials said reporting more deaths would lead to their cities and States being heavily criticised, and unfairly so, they thought. An adviser to the Delhi government said, “If we start adding suspected deaths, no one’s going to praise us for it. All we’ll get is ‘Deaths shoot up in Delhi’ sort of headlines.”

The delay factor

Such concerns have been borne out by recent media coverage of similar attempts. Reconciliation exercises aimed at fixing delayed data in the three cities and were reported as if hidden deaths had been “exposed”, even though delayed reporting and reconciliation are a feature of epidemic reporting the world over, and the new numbers were the result of these audit committees doing the job they were set up to do. When missing deaths are discovered, they are slowly added to the total over the next few days by administrations that fear people becoming angry or panic-stricken if they read about or see a sudden spike.

From a public health perspective, the attempts of State governments to restrict the numbers of deaths being reported so as not to show a higher burden than other States is a loss of valuable knowledge. In fact, Dr. Jha recommends that governments adopt the most liberal definition of a COVID-19 death as possible; leaving in a few extra deaths of COVID-19 positive persons who died in road accidents, or untested people who died of a respiratory illness that was not COVID-19 is a small price to pay, given all of the potentially missed cases such a liberal attitude will bring into the system, he said.

Data | How do States fare on reporting COVID-19 data?

It would of course be a mistake to celebrate every increase in numbers, particularly in the case of a pandemic that is causing such devastation. But parsing the numbers to see what is driving the increase, and whether the underlying processes are working well is key. Without that, all that our outrage would do would be to create incentives to suppress numbers. That is a decline in data that we should not celebrate.

Rukmini S. is an independent journalist based in Chennai

Our code of editorial values

This article is closed for comments.
Please Email the Editor

Printable version | Jul 1, 2022 1:57:19 am |