Covid-19: What you need to know today

India’s apex medical research body ICMR and the Union health ministry need to investigate why data for the past few weeks has been inconsistent — not tout Covid-19 recovery percentages as a sign of success.

The number of Covid-19 tests carried out in India continues to fluctuate wildly. On Saturday, only 987,861 tests were carried out in the country, 26% lower than the 1.34 million tests carried out the previous day. A day before, on September 24, India carried out 1.49 million tests; on September 23, the number was 1.15 million; and on September 22, 953,683.

As Dispatch 168 pointed out on September 26, this doesn’t help anybody — simply because there has been no pattern in India’s until then highly predictable Covid-19 numbers since September 10. That column added that because of this, the fact that the seven-day average of daily new cases fell by around 7,000 in the week to September 24 meant nothing.

On Saturday, Kerala’s positivity rate was 11.8%. The positivity rate is the number of people testing positive expressed as a percentage of those being tested. Karnataka’s positivity rate the same day was 12.9%, Madhya Pradesh’s 10.04%, Chandigarh’s 10.22%, Chhattisgarh’s 20.6%, Goa’s 25.3%, Maharashtra’s 21.3%, and Puducherry’s 10.8%. Together, these states and one Union territory accounted for almost one in every two cases of Covid-19 recorded on Saturday. Add Andhra Pradesh, which had a positivity rate of 9.6% on Saturday, to this group, and it emerges that almost 60% of the daily cases in India that day came from states with relatively high positivity rates.

This is another indicator that the infection isn’t really slowing in India (as much as I wish it were).

Some of these states (which are following the right testing strategy) are on a long plateau of positivity rates (more on this shortly), which means they are unlikely to see a sudden dip in cases, although, eventually, they will see a fall.

This is why I used the term relatively high to describe a positivity rate of around 10% — because a rate between 8% and 12% can actually be a good thing in states testing adequately and following the correct testing protocol. For instance, for weeks together, Tamil Nadu’s positivity rate ranged between 8% and 12%. The state was testing heavily and adequately, and its dependence on the highly unreliable rapid antigen tests was low to non-existent, yet its positivity rate didn’t fall. The state’s positivity rate trajectory was on a long plateau that’s typically seen in the case of cities, provinces and countries following the correct testing approach. It is only in recent weeks that the state’s positivity rate has fallen. On Saturday, for instance, it was 6%.

But other states (which are not following the right strategy) are not testing adequately. Positivity rates are complex and need to be understood in the correct context — in some cases, a moderately high positivity rate (10%) may be a sign of adequate testing; in other cases, a low rate or a high one may be sign of inadequate testing; and in still others, a low rate may well mean the infection in the region is in control. To resort to one of this writer’s oft-used refrains, it’s important to track the trajectory of positivity rates to understand them. Once again, Tamil Nadu’s trajectory is a text-book pattern — a rise with increased testing, then a plateau, and finally, a fall.

But positivity rates can also be skewed by the choice of tests — an inaccurate one, such as the rapid antigen test (it identifies many infected individuals as uninfected) can artificially push down the positivity rate. Dispatch 156, published on September 12, details the math of how this happens.

The worrying thing in Saturday’s numbers, from this writer’s perspective, is that Bihar’s positivity rate was less than 1% (actually 0.97%), and Uttar Pradesh’s 2.74%. Both states have high populations, poor public health systems and social indicators, and depend heavily on rapid antigen tests.

October and November is when India’s annual festive season peaks. It is a time for celebration (and people tired from months of lockdowns and restrictions are waiting to exhale) and consumption (something the flagging Indian economy badly needs).

Simultaneously, with half the school-year over, there is a growing clamour in certain quarters to reopen schools.

Decisions on what kind of celebrations can be allowed at this time, and whether schools should reopen need to be made on the basis of data and science — and there are far too many inconsistencies in recent data.

India’s apex medical research body ICMR and the Union health ministry need to investigate why data for the past few weeks has been inconsistent — not tout recovery percentages as a sign of success. After all, with a case fatality rate of 1.6%, ultimately 98.4% of people infected by the SARS-CoV-2 virus will recover. It’s just math.